Finding Brands and Losing Your Religion?

Keisha M. CutrightUniversity of Pennsylvania

Tülin ErdemNew York University

Gavan J. FitzsimonsDuke University

Ron ShacharInterdisciplinary Center (IDC) Herzliya

Religion is a powerful force in many people’s lives, impacting decisions about life, death, and everythingin between. It may be difficult, then, to imagine that something as seemingly innocuous as the usage ofbrand name products might influence individuals’ commitment to religion. However, we demonstrateacross 6 studies that when brands are a highly salient tool for self-expression, individuals are less likelyto report and demonstrate strong religious commitment. We suggest that a desire to maintain consistencyamong self-identities is one important driver of this relationship and find that the effect is mitigated whenthe perceived distance between brands and religious values is minimized.

Keywords: religion, brands, self-expression, self-consistency

Supplemental materials: http://dx.doi.org/10.1037/a0037876.supp

Whether you consider it to be a source of deadly extremism orthe pathway to humanity’s highest potential (Pargament, 2002),few will deny the power of religion. In the United States, the mostreligious of developed countries, 71% of people are “absolutelycertain” that God or another universal spirit exists (Pew ResearchCenter, 2010), and 59% say religion plays a “very important” rolein their lives (Pew Research Center, 2002). In many developingnations, religious beliefs are even stronger. Greater than 90% ofpeople in many areas of Africa, Asia, and the Middle East say thatreligion plays a “very important” role in their lives (Pew ResearchCenter, 2002). Still, many wonder whether religion may actuallybe slowly declining in its influence on everyday life. The vastmajority of Americans (77%) believe this to be the case (Newport2013), and research suggests that the worldwide percentage ofpeople who define themselves as “religious” fell from 77% in 2005to 68% in 2012 (Gilani, Shahid, & Zuettel, 2012). Interestingly,although religiosity may be experiencing a downward trend, an-other trend seems to have accelerated in its place—the introductionand advertisement of brand-name products—contributing to whatsome deem to be the “Golden Age of Brands” (e.g., Twitchell1999). As brands increasingly provide consumers with new andintriguing ways to express the self at every corner, are individuals’relationships with brands replacing their relationships with God?

In this research, we explore the notion that when individuals usebrands to express the self, they will exhibit lower levels of reli-gious commitment, defined as the degree to which they adhere toreligious values, beliefs, and practices in daily life (Worthington etal., 2003). Although religious commitment is a broad topic andmultiply determined, we suggest that brands play an interestingrole based on individuals’ need to maintain consistency amongdifferent aspects of the self. Specifically, we suggest that brandsand religion can activate distant, and at times, conflicting aspectsof self. Combining this insight with a long history of researchsuggesting that individuals desire to have a coherent, consistentview of self (e.g., Heider 1958), we expect that individuals willavoid the discomfort of simultaneously expressing these conflict-ing identities by minimizing their religious commitment whenbrands are an active tool for communicating one’s identity.

In sum, we aim to provide the first empirical investigation of theimpact of branding on religious commitment, a relevant topicgiven the ubiquity of brands in everyday life and the esteem withwhich many hold their religious commitment. In doing so, weprovide a novel perspective on the psychological impact of brand-ing and enhance our understanding of when and why religiouscommitment may often be deprioritized in individuals’ lives.

Brands, Religion, and Identity

Research suggests that what we consider to be the “self” actu-ally reflects a constellation of multiple, context-dependent identi-ties that capture a host of diverse roles, relationships, traits, goals,values, and experiences (e.g., Higgins, 1987; Markus & Nurius1986; McConnell, 2011; Roberts & Donahue, 1994). Consumersorganize these identities within an associative network and iden-tities vary in their degree of associations to one another; someidentities are tightly connected with one another, whereas othersare highly distant and dissociated from one another (Amiot, De La

This article was published Online First September 15, 2014.Keisha M. Cutright, The Wharton School, University of Pennsylvania;

Tülin Erdem, Stern School of Business, New York University; Gavan J.Fitzsimons, Fuqua School of Business, Duke University; Ron Shachar,Arison School of Business, Interdisciplinary Center (IDC) Herzliya.

Correspondence concerning this article should be addressed to KeishaM. Cutright, The Wharton School, University of Pennsylvania, Jon M.Huntsman Hall, Suite 700, 3730 Walnut Street, Philadelphia, PA 19104.E-mail: kcut@wharton.upenn.edu

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

Journal of Experimental Psychology: General © 2014 American Psychological Association2014, Vol. 143, No. 6, 2209–2222 0096-3445/14/$12.00 http://dx.doi.org/10.1037/a0037876

2209

Sablonniere, Terry, & Smith, 2007; Greenwald et al., 2002; Lane& Scott 2007; Luna et al., 2008; McConnell, 2011; Roccas &Brewer 2002). Of interest to the investigation at hand, religion andbrand-name products both serve as contributors to an individual’snetwork of identities (although they often represent a fairly disso-ciated pair of identities—as is later discussed).

With respect to religion, research suggests that it is an importantmeans by which many individuals discover and reaffirm who theyare, “whose” they are, and where they belong in the world (Da-mon, 1983; King, 2003; Mol, 1976; Ysseldyk et al., 2010). It notonly provides individuals with a purpose for their existence butalso offers prescriptions for how to live and what goals to pursue(Baumeister, 1991; Pargament, 2001). Religion also allows peopleto experience an identity that is connected to a higher power anda community of believers (Gebauer & Maio, 2012; Granqvist et al.,2010; King, 2003; Krause & Wulff, 2005) and ultimately enhancesindividuals’ feelings of self-worth (Batson & Stocks, 2004;Sedikides & Gebauer, 2010).

As individuals affirm religion as a part of their identity, theyhave many tools for signaling this identity to themselves andothers. For example, attendance at religious services and partici-pation in various religious rituals (e.g., ceremonies celebratingbirth, coming of age, marriage, death, etc.) are ways by whichindividuals affirm and express their identities (Hammond, 1988;Seul, 1999). Moreover, material symbols of religion (e.g., art,literature, dress, etc.) serve as important forms of identity expres-sion (Gaines, 1985; Keenan & Arweck, 2006; Sandikci & Ger,2010). In fact, some researchers have equated religion to an artform given its expressive tendencies (Beit-Hallahmi, 1986;Pruyser, 1976). Others have highlighted how religious languageitself is a powerful and unique form of expression that allowsindividuals, even in public (nonreligious) groups, to communicatetheir place in civic landscapes (Lichterman, 2008).

Importantly, just as religion represents an important aspect ofindividuals’ identities, so do brand-name products. Building onWilliam James’ (1890) contention that possessions can be re-garded as integral aspects of one’s identity, a great deal of researchhas demonstrated that people often construct their self-conceptsand express their identities through their relationships with prod-ucts (Belk, 1988; Berger & Heath, 2007; Chernev, Hamilton, &Gal, 2011; Reed, 2004; Richins, 1994; Sirgy, 1982; Solomon,1983; Wallendorf & Arnould, 1988). Within this tradition, re-search further suggests that brand-name products are especiallywell suited for constructing and expressing aspects of the self dueto the distinctive images and personalities that they possess(Aaker, 1997; Escalas & Bettman, 2003; Fournier, 1998; Gardner& Levy, 1955; Muniz & O’Guinn, 2001; Schouten & McAlexan-der, 1995; Wu, Cutright, & Fitzsimons, 2011). Aaker (1999), forexample, demonstrated that consumers use different brands toexpress their identities, depending on the personalities of thebrands (e.g., ruggedness, excitement) and the particular situations.Relatedly, Escalas and Bettman (2003) demonstrated that individ-uals connect to brands that allow them to express identities that areconsistent with their desired ingroups. Research also suggests thatpeople often spend lavishly on branded products to create and“show off” identities related to financial wealth and particularaspects of culture, style, or taste (Amaldoss & Jain, 2005; Twitch-ell, 2002). It is important to note, however, that brands function notonly in allowing individuals to signal their identity to others but

also in reaffirming who they are to themselves (Bodner & Prelec,2003; Loewenstein, 1999).

Brand and Religion: Distant Aspects of Identity?

Although the prior research suggests that both religion andbrands are important aspects of individuals’ identities, the under-lying goals and values that they communicate may be perceived byconsumers as being quite distant from one another. To illustrate,we turn to value systems research. Schwartz (1992), for example,identified what is considered to be a foundational set of 10 basichuman values and articulated how they relate to one another aseither complements or competitors within a two-dimensional cir-cumplex model. The first dimension of this circumplex focuses onhow values that highlight an “Openness to Change” (i.e., hedo-nism, stimulation, self-direction) directly conflict with values re-lated to “Conservation” (i.e., security, conformity, tradition). It isthe second dimension, however, that is most informative for thepresent research. It suggests that values related to “Self-Enhancement” (i.e., hedonism, power, achievement) conflict withthose focused on “Self-Transcendence” (i.e., benevolence, univer-salism). Or as Burroughs and Rindfleisch (2002) summarized, theself-oriented values conflict with the more communal-orientedvalues, respectively. Important to the present work and discussednext, the findings of prior research imply that a focus on religionbest reflects a self-transcendent, communal orientation, whereas afocus on brands best reflects a self-enhancement, self-focusedorientation.

With respect to religious values, researchers find that a keytheme across the world’s major religions involves encouragingpeople to seek meaning beyond their everyday existence and torevere the fascinating universe (Schwartz & Huismans, 1995).Moreover, in encouraging people to think beyond the self, reli-gions generally highlight the importance of benevolence and ad-vise against self-indulgent tendencies. Accordingly, in a studyinvolving individuals across four major Western religions (Prot-estantism, Roman Catholicism, Greek Orthodoxy, Judaism),Schwartz and Huismans (1995) found that religiosity is positivelycorrelated with benevolence (a key element of the communal-oriented, self-transcendent dimension of values in Schwartz’s cir-cumplex). The authors also found strong negative correlations withthe self-oriented, self-enhancement dimension of Schwartz’s val-ues (power, achievement, hedonism). As discussed next, it iswithin this latter dimension that values associated with brandsseem to most typically reside.

As noted previously, a substantial body of research suggests thatindividuals often use brands for identity purposes. Incorporatingbrands into aspects of their identity, consumers often rely onbrands to complement, bolster, or defend desired aspects of self,that is, to enhance the self. Even when brands are used to expressone’s connection and membership within a desired group (e.g.,Apple community, Harley-Davidson riders), a primary benefit ofaffiliation is the articulation of one’s own identity. As the designagency Fitch (2003) suggested on behalf of the consumer, “par-ticipation in a mass consumer experience is in some way, ironi-cally, a reflection of my discerning individuality.” More specifi-cally, and building on the specific aspects of self-enhancement thatSchwartz (1992) refers to as being particularly distant from self-transcendence motives, brands are often used to enhance feelings

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2210 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR

of power and achievement. For example, MBA students withlower levels of academic achievement have been shown to com-pensate for their mediocrity with fancier watches, shoes, and thelike (Wicklund & Gollwitzer, 1982). Similarly, individuals whofeel low in power, or those (men) seeking to attract mates bysignaling status, are more likely to prefer high-status goods, usu-ally prestigiously branded (Griskevicius et al., 2007; Rucker &Galinsky 2008). Additionally, research suggests that individualswith low control seek to partner with brands for a sense of help andempowerment while pursuing desired goals (e.g., greater health,better academic performance) (Cutright & Samper, 2014).

Importantly, the communal and self-transcendent focus of reli-gion often directly opposes such incorporation of goods into theself for self-enhancement. For example, one of the seven deadlysins in the Judeo-Christian tradition (and often deemed the “dead-liest”) is pride/vanity. Pride, as Belk (1983) summarized, involvesexcessive admiration of self and perceptions of self-sufficiencysuch that the self replaces God and/or other people as the focus ofone’s concern. The risk with brands, then, is that as they enhancethe self and perpetuate pride, God and others will become lesscentral in individuals’ lives.

The idea that individuals may perceive brand-related aspects ofself to be distant and incongruent from religious aspects of self isalso consistent with recent research by Shachar et al. (2011). Theauthors found that when religious people are given a choicebetween branded and nonbranded products, they are less likely tochoose brands than their nonreligious counterparts. This suggeststhat when consumers are able to express themselves with theirreligious identity, brand-related aspects of identity are relegated toa lower priority. Importantly, the results are strongest when reli-gion is considered to be an extension or expression of self, high-lighting the importance of individuals’ identity in the relationshipbetween religion and brands. Moreover, the results hold even aftercontrolling for materialism (measured as the desired acquisitionand expected happiness from material objects; Richins 2004),suggesting that the choice of brands versus nonbrands involvesmore than a desire to acquire (utilitarian or expressive) materialobjects and is more specifically about the degree of identityexpression that is involved. Finally, although this prior research isfocused on the effects of religion on consumption choices and thepresent research focuses instead on the less intuitive effects ofbrands on religion, the demonstration of a basic link betweenreligion and brands is encouraging for the present hypotheses.

The Desire for Consistency

The notion that brands and religion may reflect highly distantand inconsistent aspects of self has interesting implications, par-ticularly in light of research that reveals that consumers seek tomaintain cognitive consistency within their network of self-associations (Greenwald et al., 2002; Heider, 1958; Osgood &Tannenbaum, 1955). Specifically, research suggests that whenindividuals perceive inconsistencies among salient identities, itleads to feelings of dissonance or tension that people typicallyattempt to avoid or quickly resolve (Festinger, 1957). When indi-viduals are unable to successfully maintain or regain perceptionsof consistency across values and other aspects of self, their satis-faction with self and well-being suffers (Rokeach & Ball-Rokeach,1989; Sheldon & Kasser, 1995). As one example, Burroughs and

Rindfleisch (2002) found that individuals who hold high levels ofcollective-oriented values, such as religiosity, in combination witha self-oriented focus on material goods, experience higher levels ofinternal conflict and lower levels of well-being.

Building on this prior research, we suggest that because express-ing the self through religion and brands often highlights distant andconflicting aspects of the self that can create tension when coacti-vated, individuals will avoid expressing themselves with bothbrand-name products and religion. More specifically, we hypoth-esize that when expressing the self through brands, individuals willbe less likely to prioritize religion as an important aspect ofidentity and will thus show lower levels of religious commitment.Importantly, we expect that this will only be the case whenindividuals are using brands to express an aspect of their identityand not when using brands in a functional, nonidentity expressivemanner.

In what follows, we test our basic hypothesis that the salience of(identity-expressive) brands leads to lower levels of religiouscommitment than the lack thereof by using a combination ofimplicit and explicit attitude measures (Studies 1–3) as well as realbehavior (Studies 4 and 5). Additionally, we demonstrate thatmaintaining consistency in one’s expressions of self is one under-lying driver of the proposed relationship (Studies 4–5) by usingindividual differences in the preference for consistency and ma-nipulations that vary perceptions of incongruity between brandsand religion.

Study 1A

Study 1A was designed to test the basic hypothesis that religiouscommitment is lower when brands are a salient means of express-ing one’s identity than when they are not. Participants were giventhe opportunity to make several choices among branded or non-branded items and then reported their levels of religious commit-ment. We used this basic choice procedure to manipulate thesalience of brands versus merely priming brands because “choice”provides a means by which individuals can express themselves andsay who they are (Kim & Drolet, 2003). This is significant becausewe expect the salience of brands to lead to lower religious com-mitment only when brands are incorporated into one’s expressionsof self. In this context, the brands provide the content with whichindividuals express their identity, and the exercise of choice pro-vides the method by which such content is communicated.

Method

Participants were 60 adults1 recruited from a market researchfirm (27 women; ages 21–69; one participant failed to complete

1 Desired sample sizes were identified before each study was run, anddata were not analyzed until all data were collected within a study.Participants volunteered to complete the studies in exchange for mon-etary compensation, and all studies were run in accordance with Insti-tutional Review Board guidelines. The dependent variables for allstudies are reported with the exception of one question asked in Studies1A, 2, 4, and 5: “How important is it to you to express aspects of youridentity through your religious beliefs?” (1 � not at all important, 7 �extremely important). The results for this measure are always signifi-cant and follow the patterns of the Religious Commitment Scale. Givenits redundancy with the scale, it is not discussed in the results here, butis available from the authors.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2211FINDING BRANDS AND LOSING YOUR RELIGION?

the dependent variables). The experiment consisted of twobetween-subject conditions: high versus low brand salience. Uponbeginning the survey, participants read that we were interested inunderstanding individuals’ product preferences. They were thenassigned to either a condition in which brands were highly salientor a condition in which brands were not. In the high-brand saliencecondition (i.e., the “brand” condition), participants chose betweentwo branded products, in four different sets. For example, in onechoice, they decided between a Nike duffle bag and a New Balanceduffle bag. In another, they chose between a Caribou coffee mugand a Starbucks mug. In the low-brand salience condition (i.e., the“nonbrand” condition), participants chose between the same pairsof products except the brand names were removed (see the sup-plemental material).

After making their choices, participants completed a brief fillerexercise to reduce suspicion about the study’s purpose.2 Next,participants were told that we would begin a separate investigation.They were then asked to complete the dependent measures. Be-cause religious commitment is a multidimensional construct thatcan be assessed on the basis of (a) religious beliefs and (b)religious activities (Glock & Stark, 1968; Worthington et al.,2003), we leverage measures that address both of these dimensionsto test our hypotheses. Specifically, the first dependent measurewas a standard summary measure of religious commitment (“Re-ligious Commitment Scale”) that captures the degree to whichindividuals have incorporated religious values, beliefs, and prac-tices into their daily lives (Worthington et al., 2003). Participantsrated their agreement on a 5-point Likert scale to 10 statementssuch as “My religious beliefs lie behind my whole approach tolife” (� � .94, M � 2.92, SD � .97). This scale was chosen notonly because it is a brief, reliable measure of the religious com-mitment construct and leverages several items that have beenuseful in prior research but also because it has been validatedamong several religions. Our second measure of religious com-mitment evaluated the importance of attending religious services.Religious service attendance is considered to be the most commonform of “public” religious commitment and is thought to be thegateway to other forms of religious commitment (Finney, 1978;Payne & Elifson, 1976). Participants indicated how important it isto attend religious services (7-point scale, not at all important—extremely important) and how often they should attend religiousservices (6-point scale: never—more than once a week) (Inglehart,2000). Responses were formed into a standardized religious ser-vice attendance index (� � .91, M � �.02, SD � 1.82).

We then collected a variety of measures to understand whetherour results could be explained by any unexpected differences in themanipulations, such as mood, feelings of uncertainty, or level ofeffort required to process product information. Specifically, par-ticipants indicated how the choice exercise made them feel via thePositive and Negative Affect Schedule (PANAS) (Watson, Clark,& Tellegen, 1988). They also indicated how much they enjoyedmaking the choices (1 � dislike extremely, 7 �like extremely),how much they liked the products that they chose between (1 �dislike extremely, 7 � like extremely), and whether their choicesmade them feel uncertain (1 � strongly disagree, 7 � stronglyagree). Participants also reported how deeply they thought abouttheir choices (1 � not at all, 7 � very deeply), how muchinformation they gathered about their choices from the pictures(1 � none, 7 � extreme amount), and how tired they were after

making their choices (1 � not at all, 7 � extremely tired). Wecollected additional measures to ascertain whether simple percep-tions of wealth or feeling materialistic, considered in lay terms toreflect a thirst for acquiring (usually high-quality) goods (Fournier& Richins, 1991), would be influenced by our manipulations anddrive religious commitment (or whether—as explored in subse-quent studies—our results are better captured by explicitly ac-counting for perceptions of conflict between brands and religion).Specifically, participants were asked to indicate whether theirchoices made them feel materialistic and wealthy (1 � stronglydisagree, 7 � strongly disagree for each item).

Finally, participants reported demographics, including age, gen-der, and religious affiliation. No significant interactions with suchdemographic measures were revealed in this or the remainingstudies.

Results and Discussion

Consistent with our hypothesis, participants in the brand condi-tion reported lower religious commitment (M � 2.66, SD � 1.08)than participants in the nonbrand condition (M � 3.18, SD � .80),F(1, 57) � 4.50, p � .04; d � .55. The brand condition alsoreported lower importance of religious service attendance(M � �.49, SD � 1.99) than the nonbrand condition (M � .43,SD � 1.55), F(1, 57) � 3.96, p � .05; d � .52. No differences inpositive or negative mood emerged as a function of the brandsalience manipulation (Fs � 1) (see the supplemental material forPANAS and other auxiliary variable means for Studies 1–2).Moreover, there were no reported differences in individuals’ en-joyment in making their choices or their overall liking of the items(Fs � 1). Individuals did not report significant differences infeeling wealthy, F(1, 57) � .71, p � .40, or materialistic, F(1,57) � 1.49, p � .23. One can also use these variables as covariatesin our statistical analyses and the prior results hold.3 Of note, thefact that the effects on the dependent variables hold even aftercontrolling for materialism is interesting, as it suggests that it is notmerely a desire for material goods that drives the relationshipbetween brands and religion. Instead, as we later demonstrate, it islikely the conflict that arises when considering both religion andbrands as expressions of identity that best captures the relationship.Finally, participants did not report thinking more deeply abouttheir choices, gathering more information, or feeling more tired ormore uncertain between the brand and nonbrand conditions (Fs �1). Thus, this set of results provides initial support for the hypoth-esis that religious commitment is lower when brands are a salientmeans of expression than when they are not.

Study 1B

Although Study 1A provides early support for our hypothesis,one might wonder whether manipulating brand salience by askingindividuals to choose between items from different brands mayhave caused individuals to focus on differences that are difficult tocontrol for across brands such as brand image, familiarity, pricing,

2 Participants were asked to find two words in a crossword puzzle.3 For example, the effect of brand salience on religious commitment is

significant when “materialistic” (p � .05) and “wealthy” (p � .04) are usedas covariates.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2212 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR

or quality when making their choices. Although we attempted tominimize the plausibility that such factors would influence ourresults by using a variety of brand pairs and by asking participantsto choose between very similar items from the different brands, wereplicated this prior study in Study 1B by asking participants tochoose between items of the same brand. Thus, the expressivecontent of brands would still be salient, but other differencesminimized. Importantly, as one might expect identity expression tobe greater when choosing between different brands, this studyprovides a more conservative test of our hypothesis by demon-strating that even choosing from among the same brands enablesindividuals to express themselves with the traits and personalitiesof the brands strongly enough to influence religious commitment.

Method

Participants were 71 university students (34 women; ages 18–32). The experiment consisted of two between-subject conditions:high versus low brand salience. Upon entering the lab, participantswere assigned to either a condition in which brands were highlysalient or a condition in which brands were not. In the high-brandsalience condition (i.e., the “brand” condition), participants chosebetween two branded products (from the same brand) across 10different pairs. For example, in one choice, they decided betweena red Adidas shirt and a green Adidas shirt. In another, they chosebetween a white Starbucks mug and a brown one. In the low-brandsalience condition (i.e., the “nonbrand” condition), participantschose between the same pairs of products, except the brand nameswere removed (see the supplemental material). Of note, as choos-ing between the same brands may be a more conservative orweaker opportunity for self-expression than choosing betweendifferent brands (Study 1A), we conducted a pretest (see thesupplemental material) to confirm that individuals in the brandcondition felt better able to express their identities with theirchoices than those in the nonbrand condition.

After making their choices, participants completed the depen-dent measures—the Religious Commitment Scale (Worthington etal., 2003) (� � .97, M � 1.73, SD � 1.01) and the importance ofattending religious services index (� � .93, M � �.15, SD �1.79). Participants then indicated how the choice exercise madethem feel via the PANAS scale (Watson et al., 1988). No signif-icant differences in positive or negative mood emerged (Fs � 1).

Results and Discussion

Replicating the results of Study 1A, participants in the brandcondition reported lower religious commitment (M � 1.31, SD �.42) than participants in the nonbrand condition (M � 2.14, SD �1.23), F(1, 69) � 14.14, p � .0004; d � .90. The brand conditionalso reported lower importance of religious service attendance(M � �.67, SD � 1.55) than the nonbrand condition (M � .35,SD � 1.88), F(1, 69) � 6.25, p � .01; d � .60.

Together, Studies 1A and 1B suggest that religious commitmentis lower when brands are salient than when they are not. Impor-tantly, this result has been revealed in a context in which individ-uals are not only exposed to brands but also able to communicatesomething about themselves with these brands because of the“choice” component. This is consistent with our notion that reli-gious commitment is lower when brands are salient than when not,

but only when brands serve an identity expression role. In Study 2,we explicitly investigate the importance of this identity link in therelationship between brands and religious commitment.

Study 2

In Study 2, we sought to provide stronger support for the notionthat religious commitment declines when individuals use brands asa means of expressing aspects of their identity relative to whenthey do not. We did so by asking individuals to think about one oftwo different kinds of brands—either a brand that allows them toexpress some aspect of their identity or one that does not (but ishighly regarded and very functional). We expect that religiouscommitment will be lower among individuals who focus on aself-expressive brand than those who do not. This design, bycomparing conditions that were both focused on brands, not onlyallows us to identify the importance of self-expression in the linkbetween brands and religious commitment but also providesgreater confidence that the shifts in religious commitment thatwere observed in Studies 1A and 1B were driven by the presenceof brands versus the presence of generics (given that generics arenot present in this study). This design also allows us to minimizeconcerns that differences in brands versus nonbrands that are notnecessarily related to identity expression (e.g., perceived quality,price, etc.) were driving the effects.

Method

Forty-one participants were recruited online (21 women, ages20–67) and randomly assigned to one of two conditions. In theidentity expression brand condition, participants were asked to“think of a brand that you really like that is helpful in allowing youto express aspects of your personality” or to “think of a brand thatyou really like that is highly functional, but doesn’t say anythingabout your personality” and to write a statement describing howthe brand serves a self-expressive or functional role in their lives,respectively. Participants were then asked to complete the Reli-gious Commitment Scale and the importance of religious serviceattendance index. Participants were then asked to report their mood(using standard PANAS measures). Additionally, to ensure thatthere were no unexpected differences in the ways that peopleviewed the expressive or functional brand that they selected, weasked participants to indicate the overall enjoyment of writingabout the brand (1 � dislike extremely, 7 � like extremely); theiroverall liking of the brand (1 � dislike extremely, 7 � likeextremely); and the degree to which the brand they selected couldbe considered prestigious, expensive, and of high quality (1 �strongly disagree, 7 � strongly agree, for each item). They alsoindicated whether they felt materialistic, wealthy, and uncertain(1 � strongly disagree, 7 � strongly agree, for each item). Finally,participants completed demographic measures.

Results and Discussion

Consistent with our hypothesis, individuals in the self-expressive brand condition reported lower religious commitmentthan those in the functional brand condition, F(1, 39) � 3.99, p �.05 (Mexpressive brand � 2.83, SD � 1.23; Mfunctional brand � 3.14,SD � 1.20, d � .62). We also see a similar pattern of results on the

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2213FINDING BRANDS AND LOSING YOUR RELIGION?

importance of religious services, although not significant, F(1,39) � 2.26, p � .14 (Mexpressive brand � �.27, SD � 1.74;Mfunctional brand � .66, SD � 1.99, d � .50). As with the priorchoice manipulation, there were no differences in positive mood(F � 1) or negative mood, F(1, 39) � 2.62, p � .11. There werealso no differences in enjoyment (F � 1) or liking (F � 1). Finally,the items that individuals chose to write about did not differ interms of being prestigious, expensive, or high quality (Fs � 1).And, individuals did not report differences on feeling materialistic(F � 1), wealthy, F(1, 39) � 2.16, p � .15, or uncertain, F(1,39) � 1.89, p � .18. These results build on our prior findings bysuggesting that it is not simply being primed with brands thatimpacts religious commitment, but it is more specifically related tothe degree to which one can express one’s identity with the brand.When individuals discussed expressive brands, religious commit-ment declined relative to those who discussed functional brands.

Study 3

In Studies 1–2, we investigated the relationship between brandsand religious commitment by using direct, explicit measures ofreligious commitment. In Study 3, we sought to provide conver-gent validity by using a more implicit proxy for religious commit-ment. By leveraging an implicit measure, we can also begin toassess whether individuals are interested in only “looking” con-sistent to others (e.g., the researchers) or whether there is a moreintrinsic, automatic desire to maintain consistency among one’scognitions, even if just for one’s self. We reason that seeing lowerlevels of religious commitment on implicit measures would pro-vide support for the latter idea.

Research suggests that individuals are quite efficient at inhibit-ing or blocking conflicting thoughts and beliefs (Perloe, 1960;Shah, Friedman, & Kruglanski, 2002) in order to maintain a senseof personal consistency. With respect to identity research in par-ticular, research reveals that when a specific identity is active,identities that are dissociated or highly distant from that activeidentity experience greater inhibition (e.g., Hugenberg & Boden-hausen, 2004). Accordingly, we reason that when the brand-relatedaspects of one’s identity are salient, the religious aspects of one’sidentity will become inhibited and, thus, religious constructs willbe less accessible. We therefore expect that when brands aresalient in our experimental context, people’s tendency to think ofreligious-focused words in a word-completion activity will de-cline. Further, given that this proxy for religious commitmentinvolves a more objective result (number of religious wordsformed) than the prior studies, and depends on individuals’ currentknowledge of religious terminology, we expected to find thegreatest impact of brand salience among people for whom reli-gious words were normally very salient: frequent religious serviceattenders. More specifically, we expected that when brands werenot salient (and thus, religious constructs uninhibited), individualswho attend religious services regularly (and are consequentlyexposed to religious constructs more often) would show greateractivation of religious constructs than those who do not. In otherwords, at baseline, individuals who attend religious servicesshould demonstrate greater accessibility of religious words thanthose who do not attend religious services. We expected, however,that making brands salient would dampen this heightened activa-tion of religious constructs among high-religious service attenders.

(Individuals who do not normally attend religious services shouldhave low accessibility of religious words at baseline, and thus havelittle room for further inhibition of these words when brands aresalient.)

Method

To test our hypothesis, we recruited 120 adults (84 women; ages18–81; one person failed to complete the dependent measures)online via Amazon Mechanical Turk. The experiment’s design wasa 2 (brand salience: high vs. low; manipulated) � 2 (religiousservice attendance: high vs. low; measured). Participants were firstrandomly assigned to the high-brand salience or low-brand sa-lience condition in the product choice task described in Study 1B.After making their choices, participants completed the dependentmeasure—a word-stem completion exercise, which is a standardmeasure of concept activation (e.g., Gilbert & Hixon, 1991; Tulv-ing, Schacter, & Stark, 1982). Participants saw eight strings ofletters that could be completed as religious or nonreligious words(and 10 filler words). Participants then completed demographicmeasures and reported their past religious service attendance be-havior. Specifically, participants indicated whether they attendedreligious services: more than once a week, once a week, once amonth, specific holidays only, once a year, less than once a year,or never. Individuals were labeled high attenders if they attendedreligious services at least once per month; otherwise, they werelabeled low attenders.4

Results and Discussion

Results revealed no significant main effects, but did reveal aninteraction of brand salience and religious service attendance onthe accessibility of religious words, F(1, 115) � 6.23, p � .01 (seeFigure 1). As expected, among high church attenders, the accessibil-ity of religious words was dampened in the brand condition relative tothe nonbrand condition, F(1, 115) � 4.15, p � .04(Mhigh attenders/brand � 1.60, SD � 1.05; Mhigh attenders/nonbrand � 2.48,SD � 1.78). Among low church attenders, the difference between thebrand and nonbrand conditions was not significant, F(1, 115) �2.08, p � .15 (Mlow attenders/brand � 1.90, SD � 1.26;Mlow attenders/nonbrand � 1.43, SD � 1.43). Viewed differently,among those in the nonbrand salience condition (i.e., at baseline),those who attended religious services frequently were more likelyto generate religious words than those who did not(Mhigh attenders � 2.48, SD � 1.78; Mlow attenders � 1.43, SD �1.43), F(1, 115) � 7.09, p � .01; d � .65. However, in the brandcondition, this pattern disappeared, F(1, 115) � .65, p � .42, as thereligious words of high attenders were inhibited. These resultssuggest that the self-reported declines in religious commitmentrevealed in the prior studies (when brands were salient relative towhen they were not) is also manifested more implicitly via theinhibition of religious constructs.

Study 4

Whereas the focus in Studies 1–3 was on establishing the basicimpact of identity expression through brands on individuals’ levels

4 The results discussed next also hold if attendance is treated as acontinuous variable.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2214 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR

of religious commitment, in Study 4, we sought to explore theunderlying drivers of this relationship. Specifically, we investi-gated whether the perceived distance and conflict between brandsand religion helps to explain the relationship. A second objectivewas to use real behavior in our assessment of religious commit-ment. Toward this end, we gave participants a real opportunity todonate to faith-based and nonfaith-based organizations in additionto using our standard dependent measures.

In order to understand whether the perceived incongruity be-tween religion and brands is an important driver of our results, weasked some of our participants to choose among brands under aself-focused, self-enhancement orientation, similar to the priorstudies. Others, however, were given the opportunity to choosebrands in a context that will also benefit others, thus bringingbrands in line with a more communal orientation and renderingthem more consistent with values perceived to be associated witha religious identity. We expect to replicate our standard effectswhen individuals choose among branded t-shirts and the choice ofbrand is designed to only benefit the self. That is, participants inthe brand condition will demonstrate lower levels of religiouscommitment, as indicated by lower donations to faith-based char-ities (but not nonfaith-based charities5) and lower self-reportedcommitment (i.e., our standard variables). We reason that becausebrand-related values are often inconsistent with religion, individ-uals will dampen religious commitment when brands are salient(and benefit only the self) in order to avoid feelings of inconsis-tency. However, when individuals believe that their choice willbenefit someone other than themselves and is thus more in linewith religious values, the effect of brands on religious commitment(e.g., faith-based donations) should disappear.

To provide further insight into the psychological mechanisms,we also include measures that assess individuals’ perceptions ofthe incongruity between their choices and religious identity.

Finally, we sought to understand how specific the effects ofbrands are. Does brand salience decrease commitment to many

different types of identities, or is the impact of brands specific toreligion? We expect that the explicit disconnect between valuesassociated with religion and brands makes the relationship fairlyunique. We would not expect brands to reduce commitment toother expressive behaviors that do not represent the same level ofdisconnect (e.g., supporting a sports team).

Method

To test our hypothesis, we recruited 260 adults (179 women;ages 18–65; one person ended the survey before completing thedependent variable) online via a market research firm. The studywas a fully crossed 2 (brand salience: brands vs. nonbrands) � 3(beneficiary: self only, other only, self and other) experiment. Inthe brand condition, individuals were exposed to a set of 12branded t-shirts. In the nonbrand condition, people were exposedto the same set of 12 t-shirts, except the brand names wereremoved. All individuals were then asked to make a choice, but theinstructions varied as to who would benefit from their choices:one’s self only, one’s self and another individual, or anotherindividual only. Within the “self-only” condition, participantswere told to select the t-shirt that they would most like to wear andwere informed that they would be entered into a lottery for achance to receive the shirt that they chose. In the “self and other”condition, participants were also told to select the t-shirt from thelist that they would most like to wear, but were told that theywould be entered into a lottery for a chance to receive the chosenshirt and to donate one t-shirt to a well-deserving charity on theirbehalf.6 Finally, in the “other-only” condition, participants weretold to select a t-shirt that they would most like to wear, but weretold that after making their selection, they would be entered into alottery for a chance to donate a t-shirt to a well-deserving charity.Thus, in each condition, participants were given the opportunity toindicate the t-shirt that they liked. Accordingly, individuals in thebrand condition were especially able to use their selection to saysomething about themselves (based on the attributes and associa-tions of the brands) relative to those in the nonbrand condition.Importantly, when layering on the “beneficiary” conditions, thosein the other-only and self and other conditions could use theirchoices to benefit others, bringing them more in line with religiousvalues.

Of note, we use the two different comparison conditions—other-only and self and other—as opposed to just one or the other,for several reasons. First, we wanted to ascertain whether thebenefits of brands needed to be completely “other” focused inorder to mitigate the effects on religious commitment or whethersome level of self-benefit would be acceptable and reconcilablewith religious commitment (e.g., the self and other condition).Second, we wanted to ensure that the weaknesses of either controlcondition would be less plausible as drivers of the results giventhat we witness a similar pattern of behavior with both controlconditions. (For example, whereas the self and other conditiondiffers in the sheer number of beneficiaries versus the self-only

5 Nonfaith-based charities can reflect a variety of secular issues and fitwith one’s brand-focused identity.

6 It was not explicitly stated what t-shirt would be given to the charity(i.e., whether the charity would choose its shirt, receive it at random, orreceive the exact t-shirt that the participant chose).

Figure 1. Study 3: Impact of brand salience on religious word-stemcompletions by religious service attendance. Error bars represent �1standard error.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2215FINDING BRANDS AND LOSING YOUR RELIGION?

condition, this is not true for the other-only condition. Similarly,whereas individuals in the self-only condition could become moreinvested in the self-expression exercise than individuals in theother-only condition because they have an opportunity to receivetheir desired shirt, this should not be true within the self and othercondition, in which individuals also have an opportunity to receivetheir desired shirt.)

After selecting their t-shirts, we assessed reactions to thedependent measures. All participants were first given the op-portunity to donate to charity. They were told that the universitysponsoring the research would be supplementing payment forthe survey and giving each participant $1 (in addition to theirnormal payment). They were asked how much (if any) of this $1payment they would like to donate to charity. They were able toallocate this payment between faith-based charities (“200 well-respected charities that span a wide array of religious affilia-tions and address a variety of important issues”) or nonfaith-based charities (“200 well-respected charities that address avariety of important issues, but are not affiliated with a reli-gion”). They could also choose to keep some or all of the moneyfor themselves. Our objective with this charity measure wastherefore to understand the following: When people are giventhe opportunity to actually demonstrate religious commitment(when real costs are involved) as a function of brand salience,will we replicate the differences that we have seen previouslyon attitudinal measures? Moreover, will patterns of religiouscommitment differ on the basis of whether or not brands aresalient in a communal (and religiously consistent) manner ver-sus a self-focused manner?

After completing this new dependent variable, participantsalso completed the dependent variables from our prior studies,including the standard religiosity scale and the church atten-dance index. Next, to explore whether and when individualsperceive incongruity between expressing the self with brandsand religion, we used two fairly subtle measures that askedindividuals to indicate their agreement with the following,which form our congruity index: “Wearing this t-shirt wouldhelp me communicate religious values” and “I would feelcomfortable wearing this t-shirt to a casual, religious-basedevent” where 1 � strongly disagree and 7 � strongly agree.Lower scores indicate greater perceived incongruity. We hy-pothesized that individuals who expressed themselves withbrands would perceive lower congruence between their t-shirtchoice and religion, unless their t-shirt choice facilitated abenefit to someone else (i.e., those in the other-only or self andother condition). Accordingly, we expected that this measure ofincongruity would mediate the relationship between the brandsalience conditions and religious commitment.

Finally, to assess whether brand expression has a unique impacton religion or instead impacts all forms of identity expressionequally, we asked participants to indicate how important differentforms of expression are to them. In addition to assessing theimportance of “Engaging in faith-based activities,” participantsindicated the importance of “Supporting my favorite sports teams,”“Supporting my favorite artists and musicians,” “Making thingsmyself,” and “Discussing my perspectives with others.” All itemswere measured on a 7-point scale where 1 � not at all importantand 7 � extremely important.

Results

Donation to faith-based charities. Investigating the impactof the brand salience conditions (brand vs. nonbrand) and benefi-ciary conditions (self only, other only, self and other) on donationsto faith-based charities, we found no main effects, but a significanttwo-way interaction, F(2, 253), � 3.01, p � .05 (see Figure 2).Upon probing this interaction, we replicate our standard effect inthe self-only condition such that individuals in the branded t-shirtcondition donated less to the faith-based set of organizations thanindividuals in the nonbranded t-shirt condition, F(1, 253) � 8.22,p � .005 (Mbrand � $0.18, Mnonbrand � $0.47, d � .65). Thiscontrast was not significant, however, when individuals were inthe self and other condition (p � .98) or the other-only condition(p � .83). Moreover, the two-way interaction of brand salience andbeneficiary conditions was not significant when evaluatingnonfaith-based donations (p � .17).

Religious Commitment Scale. The results using the standardReligious Commitment Scale were consistent with those foundabove. There were no main effects, but a significant interaction ofthe brand salience conditions (brand vs. nonbrand) and beneficiaryconditions (self only, other only, self and other) emerged, F(2,253), � 4.86, p � .01 (see Figure 3). We replicate our standardeffect in the self-only condition whereby individuals in the brandcondition reported lower religious commitment than individuals inthe nonbrand condition, F(1, 253) � 8.86, p � .003 (Mbrand �2.74, Mnonbrand � 3.47, d � .78). This contrast was not significant,however, when individuals were in the self and other condition(p � .32) or the other condition (p � .52).

Standardized church attendance index. When assessingchurch attendance, we found similar results as with the priormeasures. There were no significant main effects, but a significantinteraction of the brand salience (brand vs. nonbrand) and benefi-ciary conditions (self, other, self and other) emerged, F(2, 253), �8.85, p � .002. Individuals in the brand condition indicated lowerimportance of church attendance than individuals in the nonbrand

Figure 2. Study 4: Impact of brand salience and beneficiary on donationsto faith-based charities. Error bars represent �1 standard error.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2216 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR

condition, F(1, 253) � 15.63, p � .001 (Mbrand � �.42,Mnonbrand � .35, d � .97). This contrast was not significant,however, when individuals were in the self and other condition(p � .23) or the other condition (p � .20).

Incongruity between religion and brands. In order to assessthe idea that one of the reasons that individuals are less likely tocommit to religion after expressing themselves with brands is thedistance and perceived incongruity between brands and religion,we measured the degree to which individuals experience a sense ofincongruity between brands and religion. We first found a maineffect of brand salience condition, F(1, 253) � 4.10, p � .04, onthe congruity index such that individuals in the brand conditionperceived less congruence between brands and religion than indi-viduals in the nonbrand condition (Mbrands � 3.68, Mnonbrand �3.98). This main effect was qualified by an interaction of brandsalience and beneficiary conditions on perceptions of congruence,F(2, 253) � 4.51, p � .01. Within the self-only condition, indi-viduals in the brand condition showed lower feelings of congru-ence than individuals in the nonbrand condition, F(1, 253) �10.90, p � .001 (Mbrand � 3.35, Mnonbrand � 4.22, d � .74). Thiscontrast was not significant, however, when individuals were inthe self and other condition (p � .36) or the other condition (p �.28).

Notably, we found that such perceptions of incongruity mediatethe effect of the “self-focused” conditions on the key dependentvariables relative to the self and other condition and the other-onlycondition. Exhibiting a pattern of moderated mediation, this is trueonly when individuals express themselves with brands (and notgenerics). More specifically, leveraging 5,000 bootstrapped sam-ples in the procedure recommended by Hayes (2013), we foundthat incongruity mediates the reactions of the self-only conditionrelative to the self and other condition (among those who ex-pressed themselves with brands) on each of the dependent vari-ables, with 95% confidence intervals exclusive of zero.7 Similarly,

perceptions of incongruity mediated the reactions of the self-focused condition relative to the other-only condition (amongthose who expressed themselves with brands) on the dependentvariables.8

Uniqueness of the impact of brands on religion versus otherexpressions of identity. Finally, to assess whether consumers’reactions to religion (faith-based activities) is unique from othertypes of expression (sports teams, art and music, making things,talking with others), we analyzed the interaction of brand salience,beneficiary, and type of activity on individuals’ perceptions of theimportance of the specific forms of expression. A significantthree-way interaction emerged, F(8, 1012) � 2.74, p � .01.

Probing this interaction, we found that the two-way interactionof Brand Salience � Beneficiary is significant when the measureof expression relates to faith-based activities, F(2, 253) � 10.47,p � .001. Specifically, within the self-only condition, individualsconsidered religious expression to be less important in the brandcondition than in the nonbrand condition, F(1, 253) � 15.63, p �.001 (Mbrand � 3.02, Mnonbrand � 5.02, d � 1.22). This was nottrue in the self and other condition (p � .99) or the other-onlycondition (p � .18). Importantly, the two-way interaction for thenonfaith-related activities was not significant (ps � .22).

Discussion

Study 4 suggests that expressing aspects of one’s identitythrough brands can lead to lower levels of religious commitment,as demonstrated not only by attitudes but also by actual (donation)behavior. This study also provides insight by introducing an im-portant boundary condition. Results indicate that when the per-ceived identity conflict between brands and religion is minimizedby giving brands a more communal orientation, the effects onreligious commitment are mitigated. Finally, this study suggeststhat the effects of brand expression on religious commitment maybe unique from its effects on many other forms of expression giventhe particularly high level of distance and conflict between valuesrelated to brands versus religion.

Study 5

In Study 5, the primary objective was to further understand howthe need to minimize conflict and perceive consistency within theself contributes to the relationship between brands and religiouscommitment. We took several steps toward this end. First, weprovided some participants with an opportunity to express the selfwith brands; others were given the opportunity to do so with theirfavorite sports team. Using sports teams in lieu of nonbrandedproducts enables us to further assess the notion that brand expres-sion is unique in its impact on religious commitment, presumablybecause of the heightened perception of conflict and distance withrespect to religious values than what appears for other forms ofexpression, such as sports team support. (A pretest confirmed

7 Faith-focused charitable giving (B � �.05 [�.11, �.003]); religiouscommitment scale (B � �.15 [�.32, �.002]); the church attendance index(B � �.09 [�.21, �.007]).

8 Faith-focused charitable giving (B � �.07 [�.14, �.02]); the religiouscommitment scale (B � �.22 [�.43, �.05]); the church attendance index(B � �.12 [�.27, �.03]).

Figure 3. Study 4: Impact of brand salience and beneficiary on religiouscommitment scale. Error bars represent �1 standard error.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2217FINDING BRANDS AND LOSING YOUR RELIGION?

that individuals perceive greater distance and conflict in therelationship between religion and brands than between religionand sports.9) We then assessed how the effect of brands onreligious commitment can be mitigated when individuals aregiven the opportunity to associate brand expression with morecommunal benefits. Finally, we assessed individuals’ chronicneeds for consistency as a further indicator of the underlyingprocess. Specifically, we expect that the impact of brand expres-sion on religious commitment (when individuals cannot justifytheir behaviors with communal benefits) should be strongest forpeople who have the strongest need to feel and appear consistentin their thoughts and actions in everyday life.

Method

To test our hypotheses, we recruited 160 participants (68women, ages 19–74) through Amazon Mechanical Turk. Thestudy was designed as a 2 (brand salience: brands vs. sports) � 2(beneficiary: self vs. others) � (preference for consistency, con-tinuous measure) experiment. Participants were randomly assignedto write about their relationship with a favorite brand or a favoritesports team. To manipulate the “beneficiary” factor in the exper-iment, individuals were assigned to write about how the brand/sports team enables them to either express who they are (“self”condition) or give of themselves for others (“others” condition).Next, we asked participants to complete our set of religious com-mitment dependent measures. Similar to Study 4, we first includea behavioral measure of religious commitment by assessing indi-viduals’ donations to faith-based and nonfaith-based charities (asdescribed in Study 410). Participants were informed that the uni-versity would allocate $5 per participant for charitable donationsand that they must indicate how they would like the donationallocated on their behalf (between the faith-based and nonfaith-based choice). In addition, participants completed the ReligiousCommitment scale and importance of church attendance measure.

Finally, in order to assess individuals’ preferences for consis-tency, the Cialdini et al. (1995) Preference for Consistency scalewas included after a filler exercise. We chose to include the scaleas an ostensibly “separate study” after the independent and depen-dent variables so as to not unfairly contaminate individuals’thoughts with a focus on inconsistency beforehand. The scaleresults were not impacted by either of the independent variables ortheir interaction (ps � .48). The scale includes items such as “Idon’t like to appear as if I am inconsistent,” where individualsindicate their agreement on a 7-point scale (1 � Strongly disagreeand 7 � Strongly agree). Scale reliability was high (� � .90).

Results

Donations to faith-based charities.11 There were no maineffects, but the two-way interaction of brand salience (brands vs.sports) and recipient conditions (self vs. others) on donations tofaith-based charities was significant, F(1, 156) � 10.59, p � .001;see Figure 4, replicating the results of Study 4. This was thenqualified by a three-way interaction when incorporating individu-als’ preferences for consistency ( � .85, t � 2.18, p � .03).Probing this interaction (Aiken & West, 1991), we found that athigh levels of the preference for consistency measure (1 SD abovethe centered mean), the two-way interaction between brand sa-

lience and beneficiary was significant ( � 2.92, t � 4.00, p �.001). Within this interaction, when individuals focused onbenefits to the self, the brand condition donated less to faith-

9 Forty participants from the same population were asked to indicate howcongruent or consistent with their religious beliefs it would be to expressthemselves through support for their favorite sports team and their favoritebrand (where 1 � highly incongruent, 7 � highly congruent). Participantsindicated that expression through sports (M � 3.82) would be more congruentthan expression through brands (M � 3.02), t(39) � 3.48, p � .001.

10 One difference in this study (vs. Study 4) is that participants were forcedto allocate the full $5 across the charities. This was done simply to minimizethe executional challenges that would be involved with giving each participantdifferent amounts of money back through Mechanical Turk.

11 As participants were told to allocate the full $5 between the twocharities, the results for donations to nonfaith-based charities do not pro-vide additional information.

Figure 4. Study 5: Impact of brand salience and preference for consis-tency on donations to faith-based charities when the focus is on the self asthe beneficiary (Panel A) versus when the focus is on others (Panel B).Orgs � Organizations.

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2218 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR

based charities than the sports condition ( � �2.52,t � �4.50, p � .001). When individuals focused on benefits toothers, there was no significant difference in donations to faith-based charities (p � .40). At low levels of the preference forconsistency (1 SD below the centered mean), the two-way inter-action of the brand salience and beneficiary conditions was notsignificant (p � .44).

Religious Commitment Scale. As with the charity measure,there were no main effects, but the two-way interaction of brandsalience and beneficiary conditions on the Religious CommitmentScale was significant, F(1, 156) � 4.34, p � .04 (see Figure 5).This was then qualified by a three-way interaction between thebrand salience conditions, beneficiary conditions, and the contin-uous preference for consistency scale ( � .85, t � 2.18, p � .03).At high levels of the preference for consistency, the two-wayinteraction between brand salience and beneficiary was significant( � 1.63, t � 2.92, p � .004). Within this interaction, whenindividuals focused on benefits to the self, the brand condition

reported lower religious commitment than the sports condition( � �1.60, t � �3.76, p � .001). When individuals focused onbenefits to others, there was no significant difference in theirreported religious commitment based on brand salience (p � .95).At low levels of the preference for consistency, the two-wayinteraction of brand salience and beneficiary was not significant(p � .89).

Importance of church attendance. Unlike the prior mea-sures, the three-way interaction between the brand salience con-ditions, beneficiary conditions, and the continuous preference forconsistency scale on the church attendance index was not signif-icant ( � .33, t � 1.46, p � .14). However, probing thisinteraction, we found that at high levels of the preference forconsistency, the two-way interaction between brand salience andbeneficiary was significant ( �.92, t � 2.17, p � .03). Withinthis interaction, when individuals focused on benefits to the self,the brand condition reported lower religious commitment than thesports condition ( � �.84, t � �2.59, p � .01). When individ-uals focused on benefits to others, there was no significant differ-ence in their reported religious commitment based on brand sa-lience condition (p � .78). At low levels of the preference forconsistency, the two-way interaction of brand salience and bene-ficiary was not significant (p � .98).

Discussion

Together, the results of Study 5 provide several important in-sights. First, by revealing that expression through sports teamsdoes not lead to the same declines in religious commitment asexpression through brand name products, Study 5 suggests thatbrands can be quite unique from other sources of self-expressionon their impact to religious commitment. Second, we can nowmore confidently articulate the role of self-consistency in thisprocess after finding that (a) the effect of brands on religiouscommitment is strongest among those for whom self-consistencyis most important and (b) the effect of brands on religious com-mitment is mitigated when individuals are given the opportunity toreconcile brands and religion by focusing on the communal ben-efits that arise from brand expression.

General Discussion

Taken together, this set of findings implies that religious com-mitment may be less stable than previously assumed, given that itcan be deprioritized by something as seemingly trivial as branding.To summarize, in Studies 1A–1B, we found that expressing theself by choosing brands leads to lower religious commitment thanchoosing nonbranded products. In Study 2, we demonstrated thatthis effect of brands on religious commitment is specific to the useof brands for self-expressive, but not utilitarian, purposes. Study 3provides convergent validity by demonstrating decreased religiouscommitment after brand expression through an implicit assessmentof such. The final studies demonstrate that the impact of brands onreligious commitment is mitigated when brands provide communalbenefits that minimize the distance to religious values (Studies 4and 5). They also provide additional process-oriented evidence tosupport the role of self-consistency in the relationship betweenbrands and religion, relying on state-based measures of perceiveddistance (Study 4) and individual differences in the chronic desirefor consistency (Study 5).

Figure 5. Study 5: Impact of brand salience and preference for consis-tency on the Religious Ccommitment Scale when the focus is on the self asthe beneficiary (Panel A) versus when the focus is on others (Panel B).

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2219FINDING BRANDS AND LOSING YOUR RELIGION?

Such findings provide new insights into our understanding ofreligion, as they illustrate that though religion is often an importantaspect of identity, individuals are willing to forgo this aspect ofidentity when conflicting aspects of identity are salient. This isimportant and surprising because, unlike many aspects of identityand forms of expression, religion is thought to embody sacredproperties among believers and be more resistant to change thanother aspects of identity. This research also provides new insightsinto the psychological power of brands. Although prior workacknowledged the role of brands in expression, the present re-search is novel in investigating how expression with brands influ-ences downstream commitments and behaviors. It begins to pro-vide support for the growing speculation that commercialization insociety (and brands as one manifestation of such) can impactindividuals’ commitments to their values and belief systems.

Of course, many questions are raised by these findings. Forexample, how might different types of brand-name products leadto different results? Brands that are communally oriented, evenreligious in nature (e.g., Newman’s Own, Chick-Fil-A, HobbyLobby, Forever 21, etc.) might generate a different pattern ofresults. It is reasonable to think that when brands have such acommunal focus, expressing one’s identity with these brandswould not feel particularly distant to religion and thus should notlead to decreases in religious commitment.

One might also wonder whether our pattern of results should beattributed to satiation as opposed to a concern for self-consistency.In other words, perhaps individuals reduce their religious commit-ment because brands have completely satiated their need for self-expression, rendering religion useless on this dimension. Althoughwe believe that brands’ ability to satiate self-expressive needs maycertainly play a role in their impact on religious commitment, it isnot likely the full story. If it were the full story, then one mightexpect to see declines not only in religious commitment but also inthe stated importance of other means of expression when asked(e.g., Study 4; sports, music/art, etc.), which we did not find.

There are also important questions regarding the ultimate im-plications for society. Although the present research is unable tospeak to how long the effects of brands on religious commitmentendure, it is plausible that the constant barrage of brands thatindividuals encounter on a daily basis could lead to sustainablechanges in religious commitment over time. If true, one must alsowonder about the impact on society given the self-regulationbenefits that are often associated with God and religion (e.g.,greater prosocial behavior, less depression, decreased substanceabuse, reduced sexual promiscuity, etc.) (Hardy & Carlo, 2005;Hardy & Raffaelli, 2003; McCullough & Willoughby, 2009;Pearce, Little, & Perez, 2003; Shariff & Norenzayan, 2007; Wills,Yaeger, & Sandy, 2003). Still, we are careful to avoid resemblingor adopting secularization theories of the past wherein severalresearchers promised a near end to religion. This has certainly notcome to fruition, nor does it seem likely (e.g., Stark, 1999, for areview). Moreover, although our research points to a decline inreligious commitment when brands are salient, it does not neces-sarily mean that individuals’ actual beliefs in the existence of Godchange.

Finally, the present research leverages a general population (andstudents) in the United States, with largely Christian backgrounds.Future research might explore how differences in cultures andreligions (e.g., denominations, intensity of beliefs) impact the

degree to which brands influence religious commitment. For ex-ample, given our finding in Study 3 regarding the inhibition ofreligious constructs after brand choice, it may be the case thatbrands are most likely to influence religious commitment amongthose who are highly religious (given that they have more room todecline in commitment by definition). It is also worth considering,however, that individuals who are at extreme levels of commit-ment (and not adequately reflected in our general populationsamples) may not view secular brands as a source of expression tobegin with and would therefore be less prone to see brands as apotential threat to their expression of identity. It would also beinteresting to consider how different religious teachings wouldinfluence the impact of brands on religion. Notably, although theuse of brand-name products as expressions of identity may not fitwithin mainstream religious values, there are segments of religiouspopulations that view the ability to afford brand-name productsand other luxuries as a sign of one’s hard work and obedience toGod. Such views gained traction with the rise of the IndustrialRevolution (Weber, 1958) and can be seen in various forms today.For example, “Prosperity Theology” or the “Faith movement”represents one area of religion in which obedience to God isbelieved to enhance financial wealth. It is believed to be “God’swill” that people experience (material and financial) prosperity(Harris, 1981; Hunt, 2010; Jackson, 1987).

In sum, this research should provide a useful starting point forfuture work that explores these and other important questionspertaining to how the basic elements of our environments andeveryday life influence religious commitment—what many con-sider to be a hallmark of human life.

References

Aaker, J. L. (1997). Dimensions of brand personality. Journal of MarketingResearch, 34, 347–356. doi:10.2307/3151897

Aaker, J. L. (1999). The malleable self: The role of self-expression in persua-sion. Journal of Marketing Research, 36, 45–57. doi:10.2307/3151914

Aiken, L. S., & West, S. (1991). Multiple regression: Testing and inter-preting interactions. Thousand Oaks, CA: Sage.

Amaldoss, W., & Jain, S. (2005). Conspicuous consumption and sophisti-cated thinking. Management Science, 51, 1449–1466. doi:10.1287/mnsc.1050.0399

Amiot, C. E., De La Sablonniere, R., Terry, D. J., & Smith, J. R. (2007).Integration of social identities in the self: Toward a cognitive-developmental model. Personality and Social Psychology Review, 11,364–388. doi:10.1177/1088868307304091

Batson, C. D., & Stocks, E. (2004). Religion: Its core psychologicalfunctions. In J. Greenberg, S. L. Koole, & T. Pyszczynski (Eds.),Handbook of experimental existential psychology (pp. 141–155). NewYork, NY: Guilford Press.

Baumeister, R. F. (1991). Meanings of life. New York, NY: Guilford Press.Beit-Hallahmi, B. (1986). Religion as art and identity. Religion, 16, 1–17.

doi:10.1016/0048-721X(86)90002-3Belk, R. W. (1983). Worldy possessions: Issues and criticisms. Advances

in Consumer Research, 10, 514–519.Belk, R. W. (1988). Possessions and the extended self. Journal of Con-

sumer Research, 15, 139–168. doi:10.1086/209154Berger, J., & Heath, C. (2007). Where consumers diverge from others:

Identity signaling and product domains. Journal of Consumer Research,34, 121–134. doi:10.1086/519142

Bodner, R., & Prelec, D. (2003). Self-signaling and diagnostic utility in

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2220 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR

everyday decision making. In I. Brocas & J. D. Carrillo (Eds.), Thepsychology of economic decisions (pp. 89–104). Oxford, England: Ox-ford University Press.

Burroughs, J. E., & Rindfleisch, A. (2002). Materialism and well-being: Aconflicting values perspective. Journal of Consumer Research, 29, 348–370. doi:10.1086/344429

Chernev, A., Hamilton, R., & Gal, D. (2011). Competing for consumeridentity: Limits to self-expression and the perils of lifestyle branding.Journal of Marketing, 75, 66–82. doi:10.1509/jmkg.75.3.66

Cialdini, R. B., Trost, M. R., & Newsom, J. T. (1995). Preference forconsistency: The development of a valid measure and the discovery ofsurprising behavioral implications. Journal of Personality and SocialPsychology, 69, 318–328. doi:10.1037/0022-3514.69.2.318

Cutright, K. M., & Samper, A. (in press). Doing it the hard way: How lowcontrol drives preferences for high effort products and services. Journalof Consumer Research. doi:10.1086/677314

Damon, W. (1983). Social and personality development: Infancy throughadolescence: New York, NY: W. W. Norton.

Escalas, J. E., & Bettman, J. R. (2003). You are what they eat: The influenceof reference groups on consumers’ connections to brands. Journal of Con-sumer Psychology, 13, 339–348. doi:10.1207/S15327663JCP1303_14

Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA:Stanford University Press.

Finney, J. M. (1978). A theory of religious commitment. Sociology ofReligion, 39, 19–35.

Fitch. (2003). Brands are the new religion. Retrieved from http://www.britishdesigninnovation.org/?page�ib/newsitem&news_id�2601

Fournier, S. (1998). Consumers and their brands: Developing relationshiptheory in consumer research. Journal of Consumer Research, 24, 343–353. doi:10.1086/209515

Fournier, S., & Richins, M. L. (1991). Some theoretical and popularnotions concerning materialism. Journal of Social Behavior and Per-sonality, 6, 403–414.

Gaines, A. D. (1985). Faith, fashion and family: Religion, aesthetics,identity and social organization in Strasbourg. Anthropological Quar-terly, 58, 47–62. doi:10.2307/3317843

Gardner, B. B., & Levy, S. J. (1955). The product and the brand. HarvardBusiness Review, 33, 33–39.

Gebauer, J. E., & Maio, G. R. (2012). The need to belong can motivatebelief in God. Journal of Personality, 80, 465–501. doi:10.1111/j.1467-6494.2011.00730.x

Gilani, I., Shahid, R., & Zuettel, I. (2012). Global index of religiosity andatheism. Retrieved from http://www.wingia.com/web/files/richeditor/filemanager/Global_INDEX_of_Religiosity_and_Atheism_PR__6.pdf

Gilbert, D. T., & Hixon, J. G. (1991). The trouble of thinking: Activationand application of stereotypic beliefs. Journal of Personality and SocialPsychology, 60, 509–517. doi:10.1037/0022-3514.60.4.509

Glock, C. Y., & Stark, R. (1968). American piety: The nature of religiouscommitment. Berkeley: University of California.

Granqvist, P., Mikulincer, M., & Shaver, P. R. (2010). Religion as attach-ment: Normative processes and individual differences. Personality andSocial Psychology Review, 14, 49–59. doi:10.1177/1088868309348618

Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D., Nosek,B. A., & Mellott, D. S. (2002). A unified theory of implicit attitudes,stereotypes, self-esteem, and self-concept. Psychological Review, 109,3–25. doi:10.1037/0033-295X.109.1.3

Griskevicius, V., Tybur, J., Sundie, J. M., Cialdini, R. B., Miller, G. F., &Kenrick, D. T. (2007). Blatant benevolence and conspicuous consump-tion: When romantic motives elicit strategic costly signals. Journal ofPersonality and Social Psychology, 93, 85–102. doi:10.1037/0022-3514.93.1.85

Hammond, P. E. (1988). Religion and the persistence of identity. Journalfor the Scientific Study of Religion, 27, 1–11. doi:10.2307/1387398

Hardy, S. A., & Carlo, G. (2005). Religiosity and prosocial behaviours inadolescence: The mediating role of prosocial values. Journal of MoralEducation, 34, 231–249. doi:10.1080/03057240500127210

Hardy, S. A., & Raffaelli, M. (2003). Adolescent religiosity and sexuality:An investigation of reciprocal influences. Journal of Adolescence, 26,731–739. doi:10.1016/j.adolescence.2003.09.003

Harris, M. (1981). America now. New York, NY: Simon & Schuster.Hayes, A. F. (2013). Introduction to mediation, moderation, and condi-

tional process analysis: A regression-based approach. New York, NY:Guilford Press.

Heider, F. (1958). The psychology of interpersonal relations. New York,NY: Wiley. doi:10.1037/10628-000

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect.Psychological Review, 94, 319–340. doi:10.1037/0033-295X.94.3.319

Hugenberg, K., & Bodenhausen, G. V. (2004). Category membershipmoderates the inhibition of social identities. Journal of ExperimentalSocial Psychology, 40, 233–238. doi:10.1016/S0022-1031(03)00096-9

Hunt, S. H. (2010). ‘Winning ways’: Globalisation and the impact of thehealth and wealth gospel. Journal of Contemporary Religion, 14, 331–347.

Inglehart, R. (2000). Codebook for World Values Survey. Ann Arbor, MI:Institute for Social Research.

Jackson, R. (1987). Prosperity theology and the faith movement. Theme-lios, 15, 16–24.

James, W. (1890). The principles of psychology. New York, NY: Holt.doi:10.1037/11059-000

Keenan, W. J. F., & Arweck, E. (2006). Introduction: Material varieties ofreligious expression. In E. Arweck & W. J. Keenan (Eds.), Materialisingreligion: Expression, performance and ritual (pp. 1–20). Hampshire,England: Ashgate.

Kim, H. S., & Drolet, A. (2003). Choice and self-expression: A culturalanalysis of variety-seeking. Journal of Personality and Social Psychol-ogy, 85, 373–383. doi:10.1037/0022-3514.85.2.373

King, P. E. (2003). Religion and identity: The role of ideological, social,and spiritual contexts. Applied Developmental Science, 7, 197–204.doi:10.1207/S1532480XADS0703_11

Krause, N., & Wulff, K. M. (2005). Church-based social ties, a sense ofbelonging in a congregation, and physical health status. The Interna-tional Journal for the Psychology of Religion, 15, 73–93. doi:10.1207/s15327582ijpr1501_6

Lane, V. R., & Scott, S. G.. (2007). The neural network model of organi-zational identification. Organizational Behavior and Human DecisionProcesses, 104, 175–192. doi:10.1016/j.obhdp.2007.04.004

Lichterman, P. (2008). Religion and the construction of civic identity. AmericanSociological Review, 73, 83–104. doi:10.1177/000312240807300105

Loewenstein, G. (1999). Because it is there: The challenge of mountain-eering . . . for utility theory. Kyklos, 52, 315–343. doi:10.1111/j.1467-6435.1999.tb00221.x

Luna, D., Ringberg, T., & Peracchio, L. A. (2008). One individual, twoidentities: Frame switching among biculturals. Journal of ConsumerResearch, 35, 279–293. doi:10.1086/586914

Markus, H., & Nurius, P. (1986). Possible selves. American Psychologist,41, 954–969. doi:10.1037/0003-066X.41.9.954

McConnell, A. R. (2011). The multiple self-aspects framework: Self-concept representation and its implications. Personality and Social Psy-chology Review, 15, 3–27. doi:10.1177/1088868310371101

McCullough, M. E., & Willoughby, B. L. (2009). Religion, self-regulation,and self-control: Associations, explanations, and implications. Psycho-logical Bulletin, 135, 69–93. doi:10.1037/a0014213

Mol, H. (1976). Identity and the sacred: A sketch for a new social scientificstudy of religion: Oxford, England: Basil Blackwell.

Muniz, A. M., Jr., & O’Guinn, T. C. (2001). Brand community. Journal ofConsumer Research, 27, 412–432. doi:10.1086/319618

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2221FINDING BRANDS AND LOSING YOUR RELIGION?

Newport, F. (2013). Most Americans say religion is losing influence in U.S.Retrieved from http://www.gallup.com/poll/162803/americans-say-religion-losing-influence.aspx

Osgood, C. E., & Tannenbaum, P. H. (1955). The principle of congruity inthe prediction of attitude change. Psychological Review, 62, 42–55.doi:10.1037/h0048153

Pargament, K. I. (2001). The psychology of religion and coping: Theory,research, practice. New York, NY: Guilford Press.

Pargament, K. I. (2002). The bitter and the sweet: An evaluation of thecosts and benefits of religiousness. Psychological Inquiry, 13, 168–181.doi:10.1207/S15327965PLI1303_02

Payne, B. P., & Elifson, K. W. (1976). Commitment: A comment on uses ofthe concept. Review of Religious Research, 17, 209–215. doi:10.2307/3510611

Pearce, M. J., Little, T. D., & Perez, J. E. (2003). Religiousness and depressivesymptoms among adolescents. Journal of Clinical Child & AdolescentPsychology, 32, 267–276. doi:10.1207/S15374424JCCP3202_12

Perloe, S. (1960). Inhibition as a determinant of perceptual defense. Per-ceptual and Motor Skills, 11, 59–66. doi:10.2466/pms.1960.11.1.59

Pew Research Center. (2002). Pew Research Global Attitudes Project: Whatthe world thinks in 2002. Retrieved from http://www.pewglobal.org/2002/12/04/what-the-world-thinks-in-2002

Pew Research Center. (2010). Religion among the Millennials. Retrievedfrom http://www.pewforum.org/files/2010/02/millennials-report.pdf

Pruyser, P. W. (1976). Lessons from art theory for the psychology ofreligion. Journal for the Scientific Study of Religion, 15, 1–14. doi:10.2307/1384310

Reed, I. I. A. (2004). Activating the self-importance of consumer selves:Exploring identity salience effects on judgments. Journal of ConsumerResearch, 31, 286–295. doi:10.1086/422108

Richins, M. L. (1994). Special possessions and the expression of materialvalues. Journal of Consumer Research, 21, 522–533. doi:10.1086/209415

Richins, M. L. (2004). The material values scale: Measurement propertiesand development of a short form. Journal of Consumer Research, 31,209–219. doi:10.1086/383436

Roberts, B. W., & Donahue, E. M. (1994). One personality, multipleselves: Integrating personality and social roles. Journal of Personality,62, 199–218. doi:10.1111/j.1467-6494.1994.tb00291.x

Roccas, S., & Brewer, M. B. (2002). Social identity complexity. Personality andSocial Psychology Review, 6, 88–106. doi:10.1207/S15327957PSPR0602_01

Rokeach, M., & Ball-Rokeach, S. J. (1989). Stability and change inAmerican value priorities, 1968–1981. American Psychologist, 44, 775–784. doi:10.1037/0003-066X.44.5.775

Rucker, D. D., & Galinsky, A. D. (2008). Desire to acquire: Powerlessnessand compensatory consumption. Journal of Consumer Research, 35,257–267. doi:10.1086/588569

Sandikci, Ö., & Ger, G. (2010). Veiling in style: How does a stigmatizedpractice become fashionable? Journal of Consumer Research, 37, 15–36. doi:10.1086/649910

Schouten, J. W., & McAlexander, J. H. (1995). Subcultures of consump-tion: An ethnography of the new bikers. Journal of Consumer Research,22, 43–61. doi:10.1086/209434

Schwartz, S. H. (1992). Universals in the content and structure of values:Theoretical advances and empirical tests in 20 countries. In M. P. Zanna(Ed.), Advances in experimental social psychology (Vol. 25, pp. 1–65).Orlando, FL: Academic Press.

Schwartz, S. H., & Huismans, S. (1995). Value priorities and religiosity infour Western religions. Social Psychology Quarterly, 58, 88–107. doi:10.2307/2787148

Sedikides, C., & Gebauer, J. E. (2010). Religiosity as self-enhancement: Ameta-analysis of the relation between socially desirable responding andreligiosity. Personality and Social Psychology Review, 14, 17–36. doi:10.1177/1088868309351002

Seul, J. R. (1999). ‘Ours is the way of god’: Religion, identity, and intergroupconflict. Journal of Peace Research, 36, 553–569. doi:10.1177/0022343399036005004

Shachar, R., Erdem, T., Cutright, K. M., & Fitzsimons, G. J. (2011).Brands: The opiate of the nonreligious masses? Marketing Science, 30,92–110. doi:10.1287/mksc.1100.0591

Shah, J. Y., Friedman, R., & Kruglanski, A. W. (2002). Forgetting all else:On the antecedents and consequences of goal shielding. Journal ofPersonality and Social Psychology, 83, 1261–1280. doi:10.1037/0022-3514.83.6.1261

Shariff, A. F., & Norenzayan, A. (2007). God is watching you: PrimingGod concepts increases prosocial behavior in an anonymous economicgame. Psychological Science, 18, 803–809. doi:10.1111/j.1467-9280.2007.01983.x

Sheldon, K. M., & Kasser, T. (1995). Coherence and congruence: Twoaspects of personality integration. Journal of Personality and SocialPsychology, 68, 531–543. doi:10.1037/0022-3514.68.3.531

Sirgy, M. J. (1982). Self-concept in consumer behavior: A critical review.Journal of Consumer Research, 9, 287–300. doi:10.1086/208924

Solomon, M. R. (1983). The role of products as social stimuli: A symbolicinteractionism perspective. Journal of Consumer Research, 10, 319–329. doi:10.1086/208971

Stark, R. (1999). Secularization, R. I. P. Sociology of Religion, 60, 249–273. doi:10.2307/3711936

Tulving, E., Schacter, D. L., & Stark, H. A. (1982). Priming effects inword-fragment completion are independent of recognition memory.Journal of Experimental Psychology: Learning, Memory, and Cogni-tion, 8, 336–342. doi:10.1037/0278-7393.8.4.336

Twitchell, J. B. (1999). Lead us into temptation: The triumph of Americanmaterialism. New York, NY: Columbia University Press.

Twitchell, J. B. (2002). Living it up: Our love affair with luxury. NewYork, NY: Columbia University Press.

Wallendorf, M., & Arnould, E. J. (1988). “My favorite things”: A cross-cultural inquiry into object attachment, possessiveness, and social link-age. Journal of Consumer Research, 14, 531–547. doi:10.1086/209134

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali-dation of brief measures of positive and negative affect: The PANASscales. Journal of Personality and Social Psychology, 54, 1063–1070.doi:10.1037/0022-3514.54.6.1063

Weber, M. (1958), The Protestant ethic and the spirit of capitalism. NewYork, NY: Charles Scribner’s Sons.

Wicklund, R. A., & Gollwitzer, P. M. (1982), Symbolic self-completion.Hillsdale, NJ: Lawrence Erlbaum Associates.

Wills, T. A., Yaeger, A. M., & Sandy, J. M. (2003). Buffering effect ofreligiosity for adolescent substance use. Psychology of Addictive Behav-iors, 17, 24–31. doi:10.1037/0893-164X.17.1.24

Worthington, E. L., Wade, N. G., Hight, T. L., Ripley, J. S., McCullough,M. E., Berry, J. W., . . . O Connor, L. (2003). The Religious Commit-ment Inventory–10: Development, refinement, and validation of a briefscale for research and counseling. Journal of Counseling Psychology,50, 84–96. doi:10.1037/0022-0167.50.1.84

Wu, E. C., Cutright, K. M., & Fitzsimons, G. J. (2011). How asking “who amI?” affects what consumers buy: The influence of self-discovery on con-sumption. Journal of Marketing Research, 48, 296–307. doi:10.1509/jmkr.48.2.296

Ysseldyk, R., Matheson, K., & Anisman, H. (2010). Religiosity as identity:Toward an understanding of religion from a social identity perspective.Personality and Social Psychology Review, 14, 60–71.

Received September 24, 2013Revision received July 29, 2014

Accepted July 30, 2014 �

Thi

sdo

cum

ent

isco

pyri

ghte

dby

the

Am

eric

anPs

ycho

logi

cal

Ass

ocia

tion

oron

eof

itsal

lied

publ

ishe

rs.

Thi

sar

ticle

isin

tend

edso

lely

for

the

pers

onal

use

ofth

ein

divi

dual

user

and

isno

tto

bedi

ssem

inat

edbr

oadl

y.

2222 CUTRIGHT, ERDEM, FITZSIMONS, AND SHACHAR