ng 23 (2006) 321–331www.elsevier.com/locate/ijresmar
Intern. J. of Research in Marketi
Explicit and implicit catalysts of consumer resistance: The effects ofanimosity, cultural salience and country-of-origin on subsequent choice
Dale W. Russell a, Cristel Antonia Russell b,⁎
a INSEAD, Boulevard de Constance, 77305 Fontainebleau Cedex, Franceb San Diego State University, Department of Marketing, 5500 Campanile Drive, San Diego, CA 92182-8239, USA
Keywords: Consumer behavior; Animosity; Cultural salience; Country-of-origin; Cultural industries
This research draws from the extant consumer research toidentify and experimentally test conditions that activateanimosity feelings and in turn affect consumers' choices. Itinvestigates the impact of several important factors on consumerresistance: country-of-origin (Bilkey & Nes, 1982; Gürhan-Canli & Maheswaran, 2000; Johansson, Douglas, & Nonaka,1985), animosity (Klein, 2002; Klein, Ettenson, & Morris,1998), and cultural salience (Aaker & Lee, 2001; Briley &Wyer, 2002). A series of experiments manipulating these factorssheds light on explicit and implicit drivers of consumerresistance and extends the current state of research in thosethree highly relevant areas of international business scholarship.
The research is conducted in the context of culturalindustries, which are an important component of the interna-tional business environment as cultural products influenceconsumers' understanding of the world (Hesmondhalgh, 2002).Movies in particular have been recognized for their ability todisseminate a country's culture and values (Hofstede, 2001;Lull, 1995; Tambyah & Thompson, 1999) and to conveyconsumption habits and ideology (Holbrook & Grayson, 1986;Nowell-Smith & Ricci, 1998). The movie industry is especiallysignificant to international business practitioners and research-
ers considering that the U.S. alone generated $25.2 billion inrevenues worldwide in 2004, including $15.7 billion frominternational sales (MPAA, 2003). The U.S. is by far the largestexporter of movies and, in an era marked by strong anti-U.S.feelings, the interplay of animosity and consumption of culturalproducts is a relevant and important venue for research.
1. Processes of resistance
Previous research suggests that consumer resistance usuallyrequires strong awareness or deep resentment (Ger & Belk,1996). Yet, at the same time, some evidence suggests thatconsumers may be able to separate their feelings toward a nationfrom their purchasing behavior. This appears especially true inthe global cultural industries. For instance, according a recentPEW Research Center (2005) report, anti-Americanism is nowat an all time high and yet, as global citizens reject the spread ofAmerican ideas and customs, the majority of respondentsreported liking American cultural exports (i.e., movies, musicand television programs). It thus appears that global consumersare able to disconnect their anti-U.S. feelings from theirconsumption of American cultural products. This research isinterested in identifying the conditions under which resistanceagainst foreign cultural products may be heightened. To do so, itbuilds upon existing country-of-origin research as well as two
322 D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
areas of research relevant to resistance: animosity and culturalsalience.
1.1. Country-of-origin research
Existing country-of-origin research has mainly focused onthe effects of country-of-origin information or stereotypes onproduct evaluations and intentions to purchase (Bilkey & Nes,1982; Gürhan-Canli & Maheswaran, 2000; Hong & Wyer,1989; Johansson et al., 1985; Maheswaran, 1994; Peterson &Jolibert, 1995; Samiee, 1994). Country-of-origin is oftenviewed as a piece of information that helps consumers assessthe quality/reliability of products from that country andsubsequently affects purchase intentions (Peterson & Jolibert,1995). A product's country-of-origin activates concepts aboutthe country and the general quality of products originating fromthere (Hong & Wyer, 1989) and the psychological processesunderlying these evaluations can be affected by consumerfactors such as motivation – the personal relevance of thedecision or product to evaluate – (Gürhan-Canli & Mahes-waran, 2000) or product familiarity or experience (Maheswaran,1994).
This research proposes that information about a product'scountry-of-origin (domestic vs. foreign) serves as a cueaffecting consumers' future receptiveness or resistance todomestic versus foreign products. For instance, in the contextof movie consumption, triggering the thought of a foreignmovie in countries such as the U.S., where domestic moviesdominate, may offset consumers' inherent bias against foreignmovies (Schooler, 1971). However, triggering the thought of adomestic movie in countries where foreign movies dominate,may, conversely, remind consumers of local productions. Thislatter effect may be especially visible in markets where importedcultural materials are perceived as a challenge to local culturaldiversity and uniqueness; in those cases, exposing consumers tomovies from a country perceived as culturally invasive mayactivate resistance. This factor will thus be at the core of ourresearch.
1.2. Explicit and implicit catalysts of resistance
The underlying construct of consumer resistance againstproducts from a particular country is animosity, which reflectsthe “remnants of antipathy related to previous or ongoingmilitary, political, or economic events” (Klein et al., 1998,p. 90). Animosity can adversely impact the consumption ofproducts from another country, regardless of judgments ofproduct quality (Klein et al., 1998), a characteristic thatdistinguishes it from ethnocentrism (Shimp & Sharma, 1987),which also affects product judgments (Klein, 2002). Since theentertainment industry is dominated by U.S. productions, theconstruct of animosity is necessary to explain resistancemechanisms linked to high degrees of economic, political,and war animosity toward the U.S. Recent international tensionsover the war in Iraq have increased anti-American sentimentsand consumers' boycotting of U.S. products throughout theworld (CNN, 2003; Economist, 2005). The same pattern was
also visible within the U.S. against countries that expresseddisagreements with American policies.
There are a few gaps in existing animosity research whichthis research attempts to address. For one, effects haveconcentrated on consumer preferences or existing ownershipsor choice likelihoods (Klein, 2002), not actual choice. This is apotentially important limitation given the extensive debateregarding whether consumers' stated preferences coincide withtheir subsequent behavior (see e.g., discussion in Eagly &Chaiken, 1993, pp. 168–192). A related area that remains opento investigation is that feelings of animosity may be latent andresistance due to animosity may occur through implicit asopposed to explicit means. Further, instead of measuringanimosity, this research suggests that animosity may be subjectto manipulation.
This research proposes and explores two factors that mayserve as catalysts of animosity feelings and, in turn, may affectconsumer behavior. The first is an explicit scenario designed toenhance or reduce animosity. Indeed, the possibility that mediacoverage of situations between two nations affect animosity –for example via exposure to different press coverage – has notbeen previously explored. The second is an implicit catalyst,cultural identity salience, which may also impact consumers'animosity (Aaker & Lee, 2001; Briley & Wyer, 2002). Theconsumer psychology literature suggests that calling people'sattention to cultural identity is likely to affect how consumersrespond to domestic or foreign products (Briley & Wyer, 2002;Druckman, 1994). Research has shown that calling people'sattention to their own cultural identity induces feelings ofallegiance to one's country and increases the tendencies toespouse values common in that culture (Briley & Wyer, 2002;Druckman, 1994). Thus, making one's own cultural identitysalient could increase one's preference for products from theirown country. Similarly, prompting another culture could impactresistance to foreign products but, unlike an explicit scenariodesigned to openly activate animosity, it might do so in a moreimplicit fashion. In particular, making salient a cultureperceived as invasive might increase the threat of culturalincursions and render consumers more defensive and resistantto foreign products. Thus, cultural salience is likely to affectresistance processes by increasing the perceived threat ofcultural invasion and activating a prevention-focus in the formof resistance to products from powerful competitors.
2. Research objectives and context
To test the effects of animosity and cultural salience onconsumers' reactions to domestic versus foreign products, thisresearch focused on consumer movie choices. This extendscountry-of-origin and animosity research, which to date hasbeen mainly confined to functional products, such as cars orappliances, in the realm of experiential products. The context isespecially fitting given that the threat of worldwide dominationby Hollywood and the presumed cultural homogenization bythe U.S. through film and television are common in many moviemarkets (Ger & Belk, 1996; Mathy, 2000). Non-Americanaudiences may reject overpowering messages from U.S.
323D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
movies, especially when reminded of the dominance of U.S.imports in their local movie scene. Such expressions ofresistance may be particularly visible amongst Europeans whooften protest the threat of worldwide domination by Hollywoodand the presumed cultural colonization of the U.S. through filmand television (Mathy, 2000).
This research proposes that the decision to select a domesticversus a foreign movie is affected by initial exposure to country-of-origin information about a movie and by a respondent's levelof animosity toward the country. Based on previous research, itis further proposed that high animosity conditions shouldmoderate the effect of foreign movie exposure; low animosityconditions should increase the effect of foreign movie exposureon subsequent choice and high animosity conditions shoulddecrease that effect. This moderating effect is tested in twoways: first by creating an explicit animosity situation and thenby activating it implicitly. In the first studies, an explicitscenario triggers high versus low animosity whereas in the finalstudy an implicit animosity manipulation is used via a culturalidentity salience task. Finally, as in previous research (Klein,2002; Klein et al., 1998), it is predicted that, throughout theexperiments, future consumption choices should be affected bythe different experimental manipulations independently ofconsumers' responses to products from the country.
Recognizing that these effects should articulate differentlydepending on the conditions in which movie consumption occurs,these propositions are tested in two markets: one dominated bydomestic movies, the U.S., where it is anticipated that the inherentbias toward domestic choices can be reduced with exposure to aforeign movie in a non-animosity context; and another, France, amarket dominated by U.S. imports but which maintains a healthynumber of local productions, especially when compared to othernational markets. The intense two-sided competition between theFrench and American film industries dates back to the 1920s whenthe industry was in its infancy (Mathy, 2000). Furthermore, thehigh degree of political and economic animosity between the twocountries since 2003 over the war in Iraq makes this selectionparticularly current and insightful.
Since this research relied on experimental manipulations, itwas important to sample a homogeneous population. In addi-tion, given the previously documented link between animosityand age (Klein, 2002), recruitment of comparable and relativelyuniform populations was essential (Netemeyer, Durvasula, &Lichenstien, 1991). For these reasons, the research relied oncollege student samples, who are primary consumers of cinema.In the U.S., those aged 12–24 represented 29.1% of all movie-goers in 2002 (MPAA, 2003); and, in France, those aged 15–24comprised 25.5% of moviegoers in 2002 (SEP Publicité, 2002).
2.1. Study 1: Explicit resistance catalyst and country-of-origineffects in the U.S.
In the first study, conducted in the U.S. where exposure toforeign movies is limited, it is hypothesized that exposure to aforeign movie synopsis would make consumers more receptiveto future foreign movie experiences (Truitt, 1975). Animosityshould moderate the effect of exposure to a foreign movie
synopsis on subsequent choice such that high (low) animosityshould decrease (increase) the beneficial effect of exposure to aforeign movie synopsis on foreign movie choices. Finally, as inprevious research (Klein, 2002; Klein et al., 1998), it is furtherpredicted that future choices would be affected by animosityand country-of-origin information about a first movie indepen-dently of consumers' product judgments.
2.1.1. Methodology
2.1.1.1. Experimental design. A 2×2 between-subject exper-iment manipulated the country-of-origin of the movie andanimosity, through the use of a scenario. Following the traditionof experimental scenarios (Carlsmith, Ellsworth, & Aronson,1990), animosity (high vs. low) and country-of-origin (Francevs. U.S.) were manipulated by exposing participants to state-ments activating either high or low degrees of animosity beforeintroducing them to either a fictional American Film Institute ora French Film Institute movie survey. The study began with adescription of the sponsoring organization and the rationale forthe research. To reinforce the country-of-origin manipulation,the name of the sponsor organization was featured throughoutthe questionnaire. Participants were fully debriefed after theirparticipation in the study.
2.1.1.2. Animosity manipulation. Animosity was manipulatedwith the use of a seeming press article presented at the onset ofthe study. The experimental scenarios consisted of either a highor low animosity statement regarding the history and currentstatus of France–U.S. trade relations (see Appendix 1). Thescenarios were pre-tested in a similar population of undergrad-uate students to ensure their success (23 male and 27 female U.S.students). After reading the assigned scenario, participants ratedtheir level of agreement with statements on a Likert like scale(1=strongly disagree; 5=strongly agree). Items were adaptedfrom Klein's research (Klein et al., 1998; Klein, 2002) tomeasure animosity and from Shimp and Sharma's (1987)research to measure consumer ethnocentrism (see all measuresin Appendix 2). The animosity and ethnocentrism scales exhi-bited good levels of internal reliability (α=.70 and .80, respect-ively). The pre-test data demonstrated that the scenariossignificantly affected animosity (2.63 vs. 2.06; t(50)=3.07,p<.05) but not ethnocentrism (1.70 vs. 2.05; t(50)=1.75,p>.05).
2.1.1.3. Country-of-origin manipulation. Participants wereasked to react to the synopsis of a new movie from their own oranother country. This manipulation was operationalized byintroducing the study as conducted by either the American FilmInstitute or the French Film Institute. The introduction describedthe commissioning organization and the rationale for pre-testinga new American (French) movie. To reinforce the country-of-origin manipulation, the name of the organization was featuredthroughout the questionnaire, at the beginning of each series ofquestions. The synopsis of the new movie, positioned either asAmerican or French, was created based on a search of existingmovie databases to ensure that it could easily be construed as an
Table 1Means as a function of animosity level and movie country-of-origin
Study 2 (France)% Selecting domestic movie 45.3% 39.7% 39.4% 58.3%% Selecting U.S. movie 42.2% 39.7% 45.1% 31.7%Animosity 3.00 2.82 3.14 3.23Ethnocentrism 2.26 2.26 2.32 2.40Movie attitude 3.09 2.85 3.06 2.84Movie involvement 3.06 2.68 3.02 2.63
324 D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
American or French story. The description, keeping gender andregion neutral, was as follows: “After graduating from theuniversity, the main character leaves friends and family in asmall country town and moves to the city to start a new career. Anew career and life in the city present many unexpected events”.All participants were debriefed at the conclusion of the expe-riments and informed of the true nature of the study.
2.1.1.4. Dependent measures. Several types of responses tothe cultural product were collected including traditionalvariables in consumer research: future choice, attitude towardthe movie, and involvement with the movie (see Appendix 2).
2.1.1.5. Choice. Choice for patronizing movies in the futurewas the main dependent variable. It was determined through theparticipants' selection of the free movie tickets for a domestic ora foreign movie. This actual choice measure was deemed anexternally valid indicator of receptiveness/resistance to foreigncultural products because it requires a great degree of personalcommitment (Peterson & Jolibert, 1995). The use of foreignversus domestic movie categorization was preferred to refe-rences to a specific country-of-origin (e.g., French) to preservethe experimental manipulations' integrity and avoid demandartifacts (Sawyer, 1975).
2.1.1.6. Responses to the movie. A three-item five-point atti-tude toward the movie measure, representing a direct evaluationof the stimulus (Peterson & Jolibert, 1995), was administeredfollowing the first exposure to the movie synopsis. Since pre-vious research identified involvement as an important moder-ator of country-of-origin effects, it was also measured using athree-item five-point scale adapted from Zaichkowsky's (1994)research.
2.1.1.7. Process and control measures. Participants' levels ofanimosity toward the other country (Klein et al., 1998) andethnocentrism (Shimp & Sharma, 1987) were measured with aseries of five-point scale questions as in the pre-test (seeAppendix 2). In all studies, these measures were collected lastand the items were randomly distributed to reduce bias. Addi-tional control questions included movie consumption (# domes-tic and foreign movies watched in the last month at the theater),the factors that affect movie choices and the preferences ofmovie genres. These measures were collected in the final por-tion of the study. Demographic information was collected at thevery end (age, gender, nationality, languages spoken).
2.1.1.8. Sample. Data were collected at two U.S. universities,one from the northeast and one from the northwest. An invitation toparticipate in a “movie study” was circulated electronically. Alottery incentive for free movie tickets to a theater of their choicewas offered as an incentive for participation. The experiment wascomputer-based and conducted individually. Participants wererandomly assigned to one of four experimental conditions. Thefinal sample, after removing non-U.S. natives based on selfreported country-of-origin, was comprised of 251 participants (130females), with a mean age of 23.1.
2.1.2. Results
2.1.2.1. Manipulation checks. To verify the success of theanimosity scenario manipulation in the main experiment, meananimosity scores (α=.89) were compared across conditions in a 2(animosity: high vs. low)×2 (movie origin: U.S. or France)between-subjects ANOVA. The manipulation was successful, asindicated by the significant animositymain effect (F(1, 247)=4.73,p<.05; η2=.02). As desired, the scenarios yielded significantlyhigher levels of animosity in the high than in the low animosityconditions (2.17 vs. 1.96, t(247)=2.19, p<.05). Neither thecountry-of-origin main effect nor its interaction with animositywere significant (both F values (1, 247)<.61, p>.05). Themanipulation did not affect levels of consumer ethnocentrism, asdesired (α=.81; all F values (1, 247)<1.83, p>.05). All means byexperimental condition can be seen in Table 1.
Mean attitudes toward the movie (α=.90) and mean involve-ment with the movie (α=.86) were compared across conditionswith a 2×2 MANOVA. As expected, the multivariate tests werenot significant (all F values (2, 244)<2.74, p>.05). Univariateanalyses reveal that, for involvement, there was a significant maineffect of country-of-origin (F(1, 245)=5.38, p<.05) qualified byan animosity×country-of-origin interaction (F(1, 245)=4.92,p<.05). The animosity main effect was not significant (F(1,245)=1.11, p>.05). While there were no differences in involve-ment levels in the low animosity condition, country-of-origingenerated significantly different levels of involvement in the highanimosity condition. Attitude toward the movie was unaffected bythe manipulations (all F values (1, 245)<1.80, p>.05), corrobo-rating previous research findings that animosity does not affectproduct judgments.
2.1.2.2. Choice. Fig. 1-A shows the proportion of participantschoosing foreign vs. domestic movie tickets per condition.Overall, U.S. participants preferred domestic movie tickets butthis preference was more pronounced in the high animosity U.S.movie condition. The data were analyzed using a 2×2 logisticregression of animosity and movie origin on choice (1=do-mestic, 0=foreign). As expected, in the U.S. movie synopsis
Fig. 1. Effects of animosity and country-of-origin on movie choices. (A) Study 1 (in the U.S.). (B) Study 2 (in France).
325D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
condition, the proportion of participants choosing domestictickets was greater under high than under low animosityconditions (84.4% vs. 69.4%, χ2(1)=4.20, p<.05), whereas inthe French movie synopsis condition, there were no suchdifferences (70.9% vs. 70.3%, χ2(1)= . 01, p>.05). Fig. 1-Aclearly illustrates that, as anticipated, high animosity conditionsand exposure to a U.S. movie synopsis increased domesticmovie choices.
2.1.3. DiscussionThis experiment showed that exposing U.S. consumers to a
U.S. movie after increasing their levels of animosity towardanother country reduced their desire for foreign movies. Thisshift in choice was not caused by any effect of attitude toward orinvolvement with the original movie therefore supporting the
prediction that future choices were not influenced by partici-pants' attitudes toward or involvement with the manipulatedmovie. Exposure to a U.S. movie synopsis in the presence ofanimosity led U.S. participants to resist foreign movies andrevert to domestic choices. The experiment also confirmed thepre-test's findings that animosity can successfully be manipu-lated with the use of press articles. The fact that the highanimosity scenario affected consumer choices only in the U.S.movie condition suggests that exposure to France-relatedinformation (a French movie) can correct the negative effectof animosity on choice.
Overall, this experiment confirmed previous findings that,if prompted to think about international affairs in a positivemanner, consumers can become more receptive to foreigncultural productions (Suh & Kwon, 2002). Consumer choices
326 D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
were affected by the manipulations: low animosity conditionsand/or exposure to a foreign movie synopsis substantiallyreduced American consumers' negative predispositions to-ward foreign films (Schooler, 1971). One limitation, however,is that, across conditions, overall levels of animosity towardFrance were low, suggesting that the effect of animosity onresistance may be better studied in a different context, hencestudies 2 and 3.
2.2. Study 2: Explicit resistance catalyst and country-of-origineffects in France
The second study, an almost exact replica of study 1, wasconducted in France. The local conditions of the cultural industrysuggested a different pattern of cultural resistance: triggeringanimosity toward a country perceived as invasive or detrimental tothe local culture should foster resistance. In particular, it ispredicted that combining a U.S. high animosity scenario withexposure to a U.S. film synopsis would trigger resistance toAmerican productions. This condition, making the cultural threatof Hollywood more salient, was expected to activate a prevention-focus amongst French consumers (Briley & Wyer, 2002), whoshould revert to domestic movies. Increasing French consumers'animosity toward the U.S. culture and exposing them to a U.S.movie was thus expected to increase domestic movie choices andreduce U.S. movie choices. As in study 1, it is predicted that theeffects would operate independently of product judgments.
2.2.1. MethodologyThe experimental design, manipulations, and measures were
similar to those in study 1. French versions of the instrumentswere produced. When available, previously translated researchinstruments were used (Netemeyer et al., 1991). Appropriatetranslation/back-translation methods and no reverse-wordeditems were used to ensure cross-cultural measurement equiv-alence (Wong, Rindfleish, & Burroughs, 2003). The choicemeasure was adapted to the local market conditions: instead ofthe simple dichotomous domestic vs. foreign choice option,three options were provided: French, American, or other(foreign, non-American).
2.2.1.1. Sample. Data were collected at two state universitiesusing the same recruiting and methodological procedures as instudy 1. The final sample, after removing non-French natives basedon their self reported country-of-origin, leaving, was comprised of253 participants (188 females), with a mean age of 21.4.
2.2.2. Results
2.2.2.1. Manipulation checks. To verify the success of theanimosity manipulation, mean animosity scores (α=.75) werecompared across conditions in a 2 (animosity: high vs. low)×2(movie origin: U.S. or France) between-subjects ANOVA. Theresults reveal a main effect of animosity (F(1, 252)=9.26,p<.05; η2 = .04) such that, as desired, the scenarios yieldedsignificantly greater levels of animosity in the high animositythan in the low animosity conditions (3.18 vs. 2.92, t(250)
=2.95, p<.05). Neither the country-of-origin main effect (F(1,252)= .28, p>.05) nor its interaction with animosity (F(1, 252)=2.31, p>.05) were significant. It was also ensured that themanipulation did not affect levels of consumer ethnocentrism(α=.78) and indeed no effects of animosity level, country-of-origin or their interaction were found (all F values (1, 252)< .90, p> .05). The experimental scenarios successfully affectedanimosity and no other construct.
All means by experimental condition can be seen in Table 1.MANOVA analyses reveal that country-of-origin of the movieaffected both attitude toward (α=.88) and involvement with(α=.88) the movie (F(2, 248)=5.15, p<.05). Univariate testsshow that the French movie generally yielded more positiveattitudes (3.08 vs. 2.84; F(1, 249)=6.56, p<.05) and greaterlevels of involvement (3.04 vs. 2.63; F(1, 249)=10.01, p<.05)than the U.S. movie. Neither animosity levels nor theirinteraction with country-of-origin affected these responses (allF values (1, 249)< .13, p>.05). The finding that French consu-mers find domestic movies better and more meaningful thanAmerican movies, ceteris paribus, is a clear indicator that, inFrance, country-of-origin information is an important and rele-vant cue for forming these evaluations.
2.2.2.2. Choice. As Fig. 1-B depicts, French participants'choices were equally split between domestic and U.S. movietickets except when they had been exposed to a high animosityscenario and a U.S. movie. In that condition, the majority (58.3%)selected domestic movies and fewer (31.7%) selected U.S. moviescompared to the other conditionswhere between 39.7% and 45.1%opted for U.S. movies. This preference shift was significant: theproportion of domestic choiceswas significantly greater in the highanimosity–U.S.movie condition than in the low animosity–U.S.movie condition (χ2(1)=4.12, p<.05) and the high animosity –French movie condition (χ2(1)=4.65, p<.05). The 2×2 logisticregression on the choice variable (1=domestic, 0=foreign)confirms this: the movie origin×cultural salience interaction wassignificant (Wald's statistic=3.79, p<.05) but not the main effects(Wald's statistics<.48, p>.05).
2.2.3. DiscussionThe pattern of results is evidence of cultural resistance: the
interaction of high animosity conditions and U.S. moviesynopsis exposure significantly increased French consumers'preference for domestic movies. As expected, preference fordomestic movies increased when French consumers' animositytoward the U.S. was heightened and they were reminded of theU.S.'s presence in their movie industry. The findings alsoconfirm the special significance of domestic movies amongstFrench consumers: the same brief synopsis yielded significantlymore positive attitudes and greater involvement when it wasannounced as a French movie than when it was presented as anAmerican movie. This illustrates the relevance of country-of-origin information in making cultural choices in this market(Gürhan-Canli & Maheswaran, 2000).
One limitation of the first two studies is that exposure tostrongly polarizing media materials, such as the high animositymanipulation, may activate reactance, a state of motivational
Table 2Means as a function of cultural prompt and movie country-of-origin
Study 3 (France) FR cultural prompt U.S. cultural prompt
327D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
arousal experienced when individuals feel pressured to adopt aparticular position (Brehm, 1966). Such defensive mechanismsare especially likely when consumers are faced with obviousattempts to affect them, as may be the case in these firstexperiments (Friestad & Wright, 1994). Arguably, the explicitanimosity scenarios could have triggered conscious reactions tothe manipulations and perhaps the ensuing corrective mechan-isms. To address this issue, the animosity manipulation wasreplaced with a more subtle manipulation of cultural identity inthe final study.
2.3. Study 3: Implicit resistance catalyst and country-of-origineffects in France
Study 3 experimentally manipulated country-of-origin of amovie and cultural salience as an implicit resistance catalyst.Based on the extant literature on cultural salience (Briley &Wyer, 2002; Hong, Morris, Chiu, & Benet-Martinez, 2000),calling consumers' awareness to their own cultural identityshould affect their cultural product choices such that they avoidthe risk of negative outcomes. The cultural salience literature isless clear about the effects of prompting a foreign culture. It isproposed that making another culture salient may operate as acatalyst of resistance, similarly to the high animosity manipula-tion used in the first studies, if that other culture is perceived asinvasive or detrimental to the local culture. Triggering thoughtsabout such a culture should increase animosity toward thatcountry and foster cultural resistance.
This proposition was tested in France, which, as study 2showed, is characterized by latent animosity feelings toward theU.S. It is predicted that combining a U.S. culture prompt withexposure to a U.S. film synopsis would make the culturaldomination of Hollywood more salient, and activate aprevention-focus amongst French consumers, reflected in anincrease in choices of domestic movies. Therefore, Frenchparticipants were expected to favor domestic movies whenprompted about the U.S. culture and movies because ofincreased animosity toward the U.S.
2.3.1. MethodologyThe study relied on a 2 (cultural salience: France vs. U.S.)×2
(movie origin: French vs. U.S.) between-subject experimentaldesign. The procedures for recruitment at two French stateuniversities were similar to those in the first two studies.Participants were randomly assigned to one of the fourexperimental conditions. They were offered the incentive offree movie tickets of their choice. Responses from non-Frenchnative participants were removed from analysis based on theirself reported country-of-origin, leaving a total of 120 partici-pants (91 females; mean age=20.0).
2.3.1.1. Experimental design. Cultural salience and movieorigin were manipulated by exposing participants to a pictureassociation task prompting their own or the other culture beforeintroducing them to either the fictional American Film Instituteor the French Film Institute movie survey. A completedebriefing was provided upon completion of the study.
2.3.1.2. Cultural salience manipulation. Cultural salience wasmanipulated by prompting participants about their own or anotherculture before exposing them to the movie information. Themanipulation was adapted from previous research (Briley&Wyer,2002; Hong et al., 2000) and comprised the first, separate portionof study 3. It was presented as a picture association exercise wheretwo sets of six pictures representative of a culture (France or theU.S.) were provided and participants were asked to associate apicture from one set (marked with letters) to a picture from theother set (marked with numbers). As in Hong et al. (2000), theimages included symbols (the flags), legendary figures from folk-lore or popular cartoons, or landmarks (e.g. the Statue of Libertyvs. the Eiffel Tower) from each culture. This task was presented asa first, separate questionnaire, distinct from the movie question-naire, which began with a description of the American (French)Film Institute, the country-of-origin manipulation.
2.3.1.3. Dependent and control measures. The measures weresimilar to those in the first two studies. To avoid demand bias, adichotomous choice measure was used by asking participantswhether theywantedmovie tickets for a domestic or foreignmovie.
2.3.2. ResultsData were analyzed using a 2 (cultural salience: France or
U.S.)×2 (movie origin: French or U.S.) between-subjectsANOVA. Table 2 presents the means of all dependent variablesin each experimental condition.
2.3.2.1. Manipulation checks. To verify the effect of thecultural salience manipulation, animosity (α=.80) was comparedacross experimental conditions. The ANOVA (η2=.06) revealed astatistically significant interaction of culture prompt×movie origin(F(1, 120)=4.90, p<.05) without main effects of prompt (F(1,120)=.17, p>.05) or movie origin (F(1, 120)=3.70, p>.05).When their own culture is salient, French consumers reportequivalent levels of animosity toward the U.S. whether exposed toa French or a U.S. movie (2.55 vs. 2.51; t(68)=.25, p>.05).However, when the U.S. culture is salient, a U.S. movie increasesanimosity while a French movie lowers it (2.76 vs. 2.18; t(48)=2.44, p<.05). This effect is similar to the high animosity-U.S.movie condition in study 2, and supports our prediction that thecombined effects of theU.S. culture prompt andU.S.movie triggeranimosity toward theU.S. Ethnocentrism (α=.71)was not affectedby the manipulations (all 1, 116)<1.21, p>.05).
The multivariate analyses comparing attitude and involvementmeans across conditions yielded no significant effects (all F values
Fig. 2. Effects of cultural salience and country-of-origin on choices. Study 3 (in France).
328 D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
(2, 115)<2.59, p>.05). Univariate analyses revealed that, as inprevious studies, there were no differences in mean attitude towardthe movie (α=.86, with all F values (1, 120)<2.05, p>.05),indicating that the effects of cultural salience and country-of-originon future movie choices operated independently of participants'attitudes toward the manipulated movie. As in study 2, Frenchmovies triggered significantly higher levels of involvement(α=.81) than did American movies (3.07 vs. 2.69; F(1, 116)=4.52, p<.05), with all other effects not reaching significance (Fvalues (1, 116)<.34, p>.05).
2.3.2.2. Choice. Fig. 2 shows the proportion of participantschoosing domestic vs. foreign tickets in each condition. Thoseexposed to a U.S. cultural prompt and a U.S. movie, a conditiontriggering animosity, exhibited cultural resistance: they were morelikely to favor domestic movies than those exposed to a U.S.culture prompt and a Frenchmovie (82.6%vs. 55.6%;χ2(1)=4.18,p<.05). This is consistent with the pattern of animosity scoresdescribed above and with study 2's findings. A 2×2 logisticregression on the choice variable (1=domestic, 0=foreign)confirms this movie origin×cultural salience interaction (Wald'sstatistic=6.00, p<.05) and no significant main effects (Wald'sstatistics<1.65, p>.05), thereby replicating study 2.
2.3.3. DiscussionAs expected, exposure to a U.S. cultural prompt and a U.S.
movie synopsis served as implicit catalysts of resistance,awaking French consumers' animosity toward the U.S. and inturn increasing their preference for French movies. Thetendency to revert to domestic movies is clearly reflective ofthe prevention focus identified in previous cultural salienceresearch (Briley & Wyer, 2002). However, the pattern of resultssuggests that the U.S. prompt served as a resistance catalystonly when put in the context of U.S. movies. Similar to study2's findings, this shows that making salient a culture perceived
as invasive can activate resistance but only when the negativeoutcomes of consumers' decisions are also made salient.
3. Limitations
Notwithstanding support for the manipulations and theireffects on movie audiences, findings are limited by selectionbias. The recruitment method, as well as incentives, may havebiased the participation toward highly involved movie patrons.Although this bias is offset by random assignment toconditions, a requirement for internal validity in our experi-ments (Calder, Philips, & Tybout, 1981), it warrants that thefindings be replicated in other cultural contexts and across abroader range of demographics. Also, the mere fact that thesample drew from an educated population, unlikely to feelthreatened by foreign cultural competition, yielded expectedlylow levels of ethnocentrism and animosity. Yet, the experi-mental manipulations were successful and their effects, albeitsmall, were nonetheless significant across all three studies, asuccess which reduces the concern. An additional strong pointof this research is that, in line with previous findings showingno link between animosity and product judgments (Klein et al.,1998), attitude toward the movie was unaffected by theanimosity manipulations. Our findings thus validate thisproposition despite the use of different methodology, setting,and operationalization.
Inherent in the selection of French and American films in thecountry-of-origin manipulations is the distinction between highversus low cultural products. Unlike Hollywood films, whichhave been described as following classical narration, Frenchfilms generally rank higher on the art dimension of the filmcontinuum, and follow different modes of narration (Tan,Eggermont, & Joosten, 1989). Thus, it is possible thatmoviegoers' differing levels of sophistication played a role inour study and affected consumer choices, in a manner similar to
329D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
the moderating effect of expertise on country-of-origin effects(Maheswaran, 1994).
4. General discussion and implications
Collectively, these three studies show that resistance processesare articulated differently depending on the local context. Whenexposure to movies from other cultures is limited, as in the U.S.,exposure to foreign movies or low animosity conditions triggerinterest in foreign movies even though one does not necessarilyperceive differences in evaluations of or likely involvement withthe movies. These findings are a clear reminder of the importanceof movies in shaping perceptions of the world and the potential forforeign movies to foster virtual cultural learning. In contrast, inmarkets where foreign competition is rife, as in France, highanimosity conditions activate a prevention-focus amongst con-sumers and increase their preferences for domestic movies. Thefinding that resistance to U.S. movies in France was only activatedin conditions where the U.S. culture was made salient and therespondents were exposed to a U.S. film suggests that anti-American feelings exist latently but do not surface unless triggeredby U.S. prompts. Despite extensive media coverage on thesupposed tensions between France and the U.S., resistance wasonly expressed when animosity was stimulated (explicitly orimplicitly) andwhen theU.S. origin of amoviewas emphasized. Inother words, only when prompted about the negative outcomes oftheir movie choices, did French consumers' animosity toward theU.S. manifest itself and resentment ensue. This echoes the PEW(2005) research findings and might indicate that American moviesare nowpart of the normal acceptedmovie scene, and that non-U.S.audiences may disconnect their feelings toward the U.S. and theirconsumption ofAmericanmovies. To this effect, it has been arguedthat American culture has now become an inseparable part ofEurope's cultural repertoire and self-identity (Monley and Robins,1989). The recent declaration from French President, JacquesChirac, of his intention to initiate a new “counter-offensive againstAmerican cultural domination” with the assistance of otherEuropean states (Agence France Presse, 2005, p. 1) is a reminderof the need for more research on the acceptance and resistanceprocesses at play in the cultural industries.
This suggests how important it is for business leaders andpolicymakers alike to understand historical factors which,although they might appear either dormant or benign, couldadversely affect international trade. As such, this opens newvenues of research for animosity effects and implies thatanimosity may be better thought of as a latent construct that canbe activated by contextual variables.
A more important contribution to international businessscholarship is the finding that animosity can be manipulated andits effects on consumer behavior tested experimentally. In additionto providing insights into the factors affecting the consumption ofcultural products, this program of research provides empiricalevidence of the usefulness of experimental methodologies to studyconsumer responses in the international business environment. Byshowing that animosity can be manipulated experimentally, thisresearch extends a field where, to date, empirical evidence islimited to correlational studies. Cross-cultural psychological
theories and methodologies can allow a better understanding ofthe processes underlying consumer reactions to foreign productsand may yield relevant and far reaching implications forinternational business practice and scholarship. Future researchcould, for instance, draw from social psychological research on theinterplay between social identities – for instance between a sub-identity and a superordinate identity – to provide insights onconsumption of foreign products based on which identity is madesalient (Roccas and Brewer, 2002). For instance, one can envisionpresenting pan-European movies as overlapping multiple cultures,or, conversely, American cultural productions as emphasizing aparticular American ethnic subculture. Finally, research distin-guishing implicit and explicit cultural identity (Devos and Banaji,2005) may be helpful in disentangling conscious and non-conscious effects triggered by animosity or cultural salience.
The finding that a simple, brief statement about trade relations issufficient to activate animosity toward another country issomewhat disquieting. In an age when consumers are constantlybombarded by news reports and sound bites, the fact that theiropinions and, subsequent choices, can be so easily swayed shouldconcern both public advocates and governments. It is one thing tofind that, when animosity toward a culturally threatening country isheightened, resistance to cultural products from that countryincreases and preventive mechanisms ensue. It is another toobserve that a short press release can create animosity, andresistance, even when there are relatively few “remnants ofantipathy related to previous or ongoing military, political, oreconomic events” (Klein et al., 1998, p. 90). The magnitude of theeffects would likely be greater were the research conducted incountries with even stronger levels of animosity.
Acknowledgements
Support for this research was appreciatively received from theAssociation for Consumer Research/Sheth Foundation, the U.S.Fulbright Program and the Center for International BusinessEducation and Research (CIBER) at San Diego State University.
Appendix A. Animosity manipulations
A.1. High animosity statement
As you might know from watching or reading the news, traderelations between the U.S. and France have always beenturbulent. France continues to impose various restrictions andquotas on many U.S. products, including entertainmentproducts, in an attempt to protect its local markets. Manypeople feel these policies are calculated measures to counter freetrade. Such policies greatly restrict the U.S.'s ability to exportproducts to France. As a result, the trade deficit between Franceand the U.S. has increased to more inequitable levels.
A.2. Low animosity statement
As you might know from watching or reading the news, traderelations between the U.S. and France are better now than anytime before. France continues to openly support trade with the
330 D.W. Russell, C.A. Russell / Intern. J. of Research in Marketing 23 (2006) 321–331
U.S. and trades a great number of goods, including entertain-ment products. Many people feel these policies serve to promotefree trade. Such policies greatly enhance the U.S.'s ability toexport products to France. As a result, the trade balance betweenFrance and the U.S. is more equitable than compared to mostother trading partners.
Appendix B. Measures
B.1. Animosity
1. France is not a reliable trade partner.2. France is taking advantage of the U.S.3. France has too much economic influence in the U.S.4. France is violating free trade at the expense of the U.S.5. I will never forgive France for not respecting the U.S.'s
positions.6. France conducts business unfairly with the U.S.
B.2. Consumer ethnocentrism
1. We should purchase American products instead of lettingother countries get rich off of us.
2. We should buy from foreign countries only those productsthat we cannot obtain within our own country.
3. Purchasing foreign-made products is un-American.4. It is always best to purchase American products.5. American consumers who purchase products made in other
countries are responsible for putting their fellowAmericans outof work.
B.3. Attitude toward the movie
1. The movie is likely to be: very bad – very good.2. I think I will like the movie: not at all – a lot.3. Watching this movie will likely be: very unpleasant – very
pleasant.
B.4. Involvement with the movie
1. The story in this movie will likely be: irrelevant to me –relevant to me.
2. The story in this movie will likely be: unimportant to me –important to me.
3. The story in this movie will likely: mean nothing to me –mean a lot to me.
References
Aaker, J. L., & Lee, A. Y. (2001). ‘I’ seek pleasures, ‘we’ avoid pains: The roleof self regulatory goals in information processing and persuasion. Journal ofConsumer Research, 28(1), 33−49.
Agence France Presse. (2005, March 22). Chirac vows to fight U.S. culturaldomination. The Strait Times, 1.
Bilkey, W. J., & Nes, E. (1982). Country-of-origin effects on productevaluations. Journal of International Business Studies, 13(1), 89−100.
Brehm, J. W. (1966). A theory of psychological reactance. San Diego, CA:Academic Press.
Briley, D., & Wyer, R. (2002). The effect of group membership salience on theavoidance of negative outcomes: Implications for social and consumerdecisions. Journal of Consumer Research, 29(3), 400−415.
Calder, B. J., Philips, L. W., & Tybout, A. M. (1981, September). Designingresearch for application. Journal of Consumer Research, 8, 197−207.
Carlsmith, J. M., Ellsworth, P. C., & Aronson, E. (1990).Methods of research insocial psychology, 2nd ed. New York, NY: McGraw-Hill.
CNN. (2003). War sparks trans-Atlantic boycotts.: Central News NetworkAvailable from http://www.cnn.com/2003/WORLD/europe/04/01/sprj.irq.france.germany/index.html
Devos, T., & Banaji, M. R. (2005). American=White? Journal of Personalityand Social Psychology, 88(3), 447−466.
Druckman, D. (1994, April). Nationalism, patriotism and group loyalty: Asocial psychological perspective. International Studies Quarterly, 38,43−68.
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth,TX: Harcourt Brace Jovanovich College Publishers.
Friestad, M., &Wright, P. (1994, June). The persuasion knowledge model: Howpeople cope with persuasion attempts. Journal of Consumer Research, 21,1−31.
Ger, G., & Belk, R. W. (1996). I'd like to buy the world a coke:Consumptionscapes of the less affluent world. Journal of Consumer Policy,19(3), 271−304.
Gürhan-Canli, Z., & Maheswaran, D. (2000, June). Determinants of country-of-origin evaluations. Journal of Consumer Research, 27, 98−108.
Hesmondhalgh, D. (2002). The cultural industries. Thousand Oaks, CA: SagePublications.
Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors,institutions and organizations across nations. Thousand Oaks, CA: SagePublications.
Holbrook, M. B., & Grayson, M. W. (1986, December). The semiology ofcinematic consumption: Symbolic consumer behavior in Out of Africa.Journal of Consumer Research, 13, 374−381.
Hong, Y., Morris, M. W., Chiu, C., & Benet-Martinez, V. (2000). Multiculturalminds: A dynamic constructivist approach to culture and cognition. TheAmerican Psychologist, 55(7), 709−720.
Hong, S., & Wyer Jr., R. (1989, September). Effects of country-of-origin andproduct-attribute information on product evaluation: An informationprocessing perspective. Journal of Consumer Research, 16, 175−187.
Johansson, J. K., Douglas, S. P., & Nonaka, I. (1985). Assessing the impact ofcountry of origin on product evaluations: A new methodologicalperspective. Journal of Marketing Research, 22(4), 388−396.
Klein, J. G. (2002). Us versus them, or us versus everyone? Delineatingconsumer aversion to foreign goods. Journal of International BusinessStudies, 32(2), 345−363.
Klein, J. G., Ettenson, R., & Morris, M. D. (1998, January). The animositymodel of foreign product purchase: An empirical test in the People'sRepublic of China. Journal of Marketing, 62, 89−100.
Lull, J. (1995). Media, communication, culture: A global approach. New York,NY: Columbia University Press.
Maheswaran, D. (1994, September). Country of origin as a stereotype: Effects ofconsumer expertise and attribute strength on product evaluations. Journal ofConsumer Research, 21, 354−364.
Mathy, J. P. (2000). French resistance: The French–American culture wars.Minneapolis, MN: The University of Minnesota Press.
Monley, D., & Robins, K. (1989). Over the borderlines: Questioning nationalidentities. Spaces of Identity, 30(4), 21−22.
Motion Picture Association of America Worldwide Market Research. (2003). U.S.movie attendance study. Available from http://www.mpaa.org
Netemeyer, R. G., Durvasula, S., & Lichenstien, D. R. (1991, August). A cross-national assessment of the reliability and validity of the CETSCALE.Journal of Marketing Research, 28, 320−327.
Nowell-Smith, G., & Ricci, S. (1998). Hollywood and Europe, economics,culture, national identity 1945–95. London, England: British Film Institute.
Peterson, R. A., & Jolibert, A. P. (1995). A meta-analysis of country-of-origineffects. Journal of International Business Studies, 26(4), 883−900.
Roccas, S., & Brewer, M. B. (2002). Social identity complexity. Personality andSocial Psychology Review, 6(2), 88−106.
Samiee, S. (1994). Customer evaluation of products in a global market. Journalof International Business Studies, 25(3), 579−604.
Sawyer, A. (1975, March). Demand artifacts in laboratory experiments inconsumer research. Journal of Consumer Research, 1, 20−30.
Schooler, R. (1971). Bias phenomena attendant to the marketing of foreigngoods in the U.S. Journal of International Business Studies, 2(2),71−80.
SEP Publicité. (2002). Annual French movie report. Available from http://www.Sep-Publicite.Com/Cinema/Audience.Asp
Shimp, T. A., & Sharma, S. (1987, August). Consumer ethnocentrism:Construction and validation of the CETSCALE. Journal of MarketingResearch, 24, 280−289.
Suh, T., & Kwon, I. W. G. (2002). Globalization and reluctant buyers. Inter-national Marketing Review, 19(6), 663−680.
Tambyah, S. K., & Thompson, C. J. (1999, December). Trying to becosmopolitan. Journal of Consumer Research, 26, 214−241.
Tan, E., Eggermont, L. J. M., & Joosten, P. D. M. (1989, August). On becominga cinephile: A theoretical and empirical analysis of film taste. TijdschriftVoor Theaterwetenschap, 27/28, 207−233.
Truitt, F. J. (1975). A multimedia approach to teaching the total environment ofinternational business: “What's it like, what's it really like?” Journal ofInternational Business Studies, 6(1), 107−120.
Wong, N., Rindfleish, A., & Burroughs, J. E. (2003, June). Do reverse-cordeditems confound measures in cross-cultural consumer research? The case ofthe material values scale. Journal of Consumer Research, 30, 72−91.
Zaichkowsky, J. L. (1994). The personal involvement inventory: Reduction,revision, and application to advertising. Journal of Advertising, 23(4),59−70.
