TheEffects of Consumer Cosmopolitanism onPurchaseBehaviorofForeign vs.DomesticProducts OliverParts IrenaVida The purpose of this empirical study is to investigate the effects of con- sumercosmopolitanismonforeignproductpurchasebehaviorinthree majorcategoriesofconsumerproducts(alcoholproducts,clothes,fur- niture). Basedon theexisting theoretical and empiricalknowledge,we develop a conceptual model and identify two additional constructs as antecedents of foreign purchase behavior, i.e., consumer ethnocen- trism and consumer knowledge of brand origins. The measurement model is examined using a data set of261 adult consumers and tested viastructuralequationmodeling.The studyresults confirmthestrong total effect of consumer cosmopolitanism in purchase behavior and indicate a strong direct effect of this phenomenon on the behavioral outcome.Themorecosmopolitanconsumershaveastrongertendency to buy foreign rather than local products. On the other hand, the di- rect relationship between cosmopolitanism and consumer knowledge ofbrandoriginwasnotsupportedinthestudy. Key Words: cosmopolitanism, consumerethnocentrism, knowledge ofbrandorigins,foreignproductpurchasebehavior,Slovenia jel Classification:m3 ,p2 Introduction In the marketing field, the five decades of country-of-origin research provide evidence that consumers carry diverse perceptions about prod- uctsbasedonthe(stereotyped)nationalimagesofthecountrywherethe product/brand is believed to be created/produced, and that these per- ceptions affect consumer attitudes, purchase intentions and behaviors (Larocheetal.2005 ;Pharr2005 ).Whilethereisastreamofresearchthat focuses on consumers’ choices regarding products from specific foreign OliverPartsisaLecturerattheTallinnSchoolofEconomicsand Business Administration, Tallinn University ofTechnology, Estonia. DrIrenaVidaisaProfessorattheFacultyofEconomics,University ofLjubljana, Slovenia. Managing GlobalTransitions9 (4 ):355 –370 356 OliverPartsandIrenaVida countries (i.e., country-image studies; for recent reviews, see Dmitro- vi´ c and Vida2010 ; Roth and Diamantopoulos2009 ), another stream of research broadly delves into factors that lead consumers to prefer ei- ther local (domestic) or foreign products/brands (e.g., Crawford and Lamb1982 ; Sharma, Shimp and Shin1995 ; Vida, Dmitrovic and Obadia 2008 ). This research focuses on the latter stream of consumer behavior research by examining consumer cosmopolitanism as a major socio- psychological construct underlying consumer preference for foreign vs. domestic(local)products/brands.Cosmopolitanism,asoriginallyintro- ducedbyMerton(1957 ), refers to individuals who are oriented towards the outside world (rather than their local community). While differ- ent terminology has been used in examining essentially the same phe- nomenon, the construct has been widely applied in the international business and marketing research (Levy et al.2007 ; Riefler and Diaman- topoulos 2009 ), including preference for foreign products (Balabanis and Diamantopoulos2008 ; Crawford and Lamb1982 ; Suh and Kwon 2002 ). However, as the recent comprehensive review on the subject reveals (Riefler and Diamantopoulos 2009 ), many questions regarding cos- mopolitanism effects and measurement issues remain unanswered. In particular, with a few exceptions (Balabanis et al.2001 ; Rawwas, Rajen- dran,andWuehrer1996 )thedirecteffectsofcosmopolitanismonbehav- ior in favor of foreign products brands have been rarely examined, and its role as a driver of consumer ethnocentrism is largely left unresolved (SuhandKwon2002 ,Vida,DmitrovicandObadia2008 ). Despite the voluminous body of research on the effects of product national origin on consumer evaluative processes and behavioral out- comes,thesalienceofproductoriginandconsumeractualknowledgeof the brands’nationalorigins hasbeenquestioned inrecent years(Liefeld 2004 ;Pharr2005 ;Samiee,Shimp, andSharma2005 ).Forinstance,Bala- banisandDiamantopoulos(2008 )recentlyexaminedtheextenttowhich consumersattachanationalorigintoabrand,andconcludedthatfuture researchersshouldadjusttheirresearchdesignstoaccountforthepossi- blyinaccurateknowledgeofastimulusbrand’snationalorigin. We designed this empirical study based on the gaps identified in the literature on consumer foreign vs. local purchase behavior andthe con- fusion regarding existing conceptualizations of cosmopolitanism and its role in consumption behavior. Hence, the aims of this research are ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 357 to examine the direct effects of consumer cosmopolitanism on foreign vs. local product purchase behavior, and to explore its indirect effects (through consumer ethnocentrism and consumer knowledge of brand origins)inthreemajorcategoriesofconsumerproducts. LiteratureReview: ConsumptionofForeignvs.Domestic Products In an attempt to understand consumer preference formation for either foreign or local product alternatives available in the marketplace, re- searchers have resorted to various socio-psychological constructs that help disentangle consumption motivations. The two most commonly appliedsocio-psychologicalconstructsintheexistingempiricalworkex- amine how individuals relate to their social in-group (e.g., family, local community, nation and its artifacts) and how they relate to what they consider their out-group (e.g., other cultures, ethnic groups, nations). Theconceptofconsumercosmopolitanismisamanifestationofpositive orientation towards the out-groups (people, artifacts, etc.), and ethno- centrism captures individuals’in-group vs. out-group orientation. Both constructs have been introduced to marketing from the field of sociol- ogy. As originally coined by Merton (1957 ), the concept of cosmopoli- tanism relates to a ‘world citizen,’ i.e., to an individual whose orien- tationtranscendsanyparticularcultureorsetting. Heposited that there are people who view themselves as citizens of the nation rather than the locality; the world rather than the nation; the broader, more het- erogeneous rather than the narrower, more homogeneous geographic or cultural group (Cannon and Yaprak2002 ;Merton1957 ). In the mar- keting literature, the concept has been advanced by many prominent scholars (CannonandYaprak2002 ;Thomson andTambyah1999 ; Yoon, Cannon,andYaprak1996 )whoarguethatcosmopolitanismisconsumer orientationwithsubstantialimplicationsformarketingpractice.Diverse terminology has been used in the literature to describe the individuals’ positive orientation towards the out-group, including openness to for- eign cultures, internationalism, worldmindedness, worldliness or global openness,etc. While cosmopolitanism has been defined differently across studies, sufficient evidence exists that it can lead to better perceptions of for- eign products, includingtheir quality (Rawwas,Rajndran,andWuehrer 1996 ), and induce a greater desire in individuals to travel as they at- Volume 9 · Number 4 · Winter 2011 358 OliverPartsandIrenaVida tempttoseeknewinsightsintoothercultures(CannonandYaprak2002 ; ThompsonandTambyah1999 ). The other socio-psychological construct commonly used to explain consumer choice behavior for foreign vs. domestic products/brands is the construct of ethnocentrism. This phenomenon was originally conceived as a purely sociological concept that distinguished between in-groups (those groups with which an individual identifies) and out- groups (those regarded as antithetical to the in-groups) (Sumner1906 ). Consumerethnocentrism wasintroducedintomarketing byShimp and Sharma (1987 ) when they stated: ‘Ethnocentric consumers believe it is wrong to purchase foreign-made products because it will hurt the do- mestic economy,causethelossofjobs, andit isplainlyunpatriotic.’The tendency of ethnocentric consumers to exhibit preferences for domes- tic rather than imported products has been confirmedin severalstudies (Cleveland,Laroche,andPapadopoulus2009 ;Dmitrovi´ c,VidaandRear- don2009 ; Rawwas, Rajendran, and Wuehrer1996 ; Sharma, Shimp, and Shin1995 ;Vida,Dmitrovic,andObadia2008 ). Inaddition to thesocio-psychological constructsof cosmopolitanism andethnocentrism,weexaminetheissueofconsumeractualknowledge of the brands’ nationalorigins as a factor underlying consumption mo- tivation forforeign vs.domestic products. Contraryto the conventional wisdom, categorization literature supports the view that most of con- sumers’ learning is unstructured and incidental, resulting in imperfect and biased knowledge (Aboulnasr2006 ). It is this notion that in re- cent years has led to a major criticism of the country-of-origin research stream, i.e., that consumers in reality pay less attention to the product nationalorigininformationcuethanisgenerallyassumedbyresearchers. Moreover, the critics claim that consumer knowledge of the actual na- tional origin of products and brands tends to be inaccurate (Balaba- nis and Diamantopoulos2008 ; Liefeld2004 ;Pharr2005 ). For instance, Samiee, Shimp, and Sharma (2005 ) examined the saliency of the prod- uctorigininformationcueintheus andconcludedthatconsumershold merelyasuperficialknowledgeofproductorigins.Theypositedthatthis knowledgeisbyandlargederivedfromconsumers’associationofbrand names with various languages rather than their actual knowledge of the brands’nationalorigins.SimilarconclusionshavebeenreachedbyBala- banisandDiamantopoulos (2008 ).For this reason,we acknowledgethe importanceofconsumerknowledgeofbrandorigininourinvestigation oftheroleofcosmopolitanisminconsumerpurchasebehavior. ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 359 Consumer Ethnocentrismce Knowledge of Brand Originskbo Foreign Product PurchasingBehaviorfppb Cosmopolitanismcp h2 – h3 + h1 + h4 – h5 + figure1 Conceptual modelforthestudy ConceptualModelandHypotheses Againstthistheoretical backgroundandthegapsidentifiedinthelitera- ture,wedevelopaconceptualmodeloftheroleofconsumercosmopoli- tanisminconsumptionofforeignvs.domesticproducts(figure1 ). In line with the conceptual framework for the study, we propose five research hypotheses. The first two hypotheses are related to the two psycho-sociological constructs (i.e., consumer cosmopolitanism andethnocentrism)directlyand/orindirectlyaffectingbehaviouralout- comes. With a few exceptions, the direct effect of cosmopolitanism or related constructs on behavioral outcomes has been rarely investigated inexistingresearch(CannonandYaprak2002 ;LeeandChen2008 ;Craw- fordandLamb1982 ).Forexample,thedirectimpactofwhatwastermed worldmindedness on Taiwanese consumers’ willingness to buy prod- ucts from neighboring countries was demonstrated by Lee and Chen (2008 ).CrawfordandLamb(1982 )examinedtheeffectofworldminded- ness on willingness to buy foreign products among professional buyers, and found that an individual’s attitude towards foreign countries is in fact related to a person’s willingness to buy products from these coun- tries. Onthe other hand, Cannonand Yaprak (2002 ) concludedin their studythat,whileconsumersarebecomingmorecosmopolitan,thisdoes not necessarily result in their behavior transcending their local culture. Hence,weposit: h1 Cosmopolitanism (cp ) has a direct and positive effect on foreign productpurchasebehavior(fppb ). Contrary to the above, the role of cosmopolitanism or related con- Volume 9 · Number 4 · Winter 2011 360 OliverPartsandIrenaVida structs (e.g., cultural openness, internationalism, global mindedness, worldmindedness)asdriversofconsumerethnocentrismhasbeenwidely examined in the literature (Shankarmahesh2006 ). However, empirical examinations of cosmopolitanism’s antecedent nature have produced onlyequivocalresults.Whileatheoreticallypositednegativerelationship betweencosmopolitanismandethnocentrismhasbeendemonstratedin severalstudies(CannonandYaprak2002 ;Dmitrovic,Vida,andReardon 2009 ; Sharma, Shimp, and Shin1995 ; Vida and Reardon2008 ), there is evidence to the contrary as well. Non significant relationships between these two constructs were identified when examining cultural openness (Vida, Dmitrovic, and Obadia2008 ), and internationalism (Balabanis and Diamantopoulus2004 ). For instance, Suh and Kwon (2002 )found that global openness had a significant negative effect on ethnocentrism in theus sample, but this relationship was insignificant in the Korean sample. Similarly, Strizhakova, Coulter and Price (2008 ) examined this relationship acrossdevelopedandemergingmarkets,andfoundamod- erate negative relationship intheus sample, butnosignificantrelation- ship intheemergingmarketsamples.Sincetheroleofcosmopolitanism inshapingconsumers’beliefsaboutthelegitimacyofpurchasingforeign made goods has yielded contradictory results in the literature, the test- ingofthefollowinghypothesisprovidesanopportunityforresolvingthe existingcontroversy: h2 Cosmopolitanism (cp )hasadirectandnegativee ffect on consumer ethnocentrism(ce ). The set of the remaining hypotheses in this study is related to a rel- atively new concept – consumer knowledge of brand origins – which has been introduced into the conceptual model in response to the crit- icisms of country-of-origin research about the relative absence of con- sumerabilitytorecognizetheactualnationaloriginofproducts(Balaba- nisandDiamantopoulos2008 ;Liefeld2004 ;Samiee,Shimp,andSharma 2005 ). Brand origin is defined by the place, region or country to which the brand is perceived to belong by its target consumers. While there is a dearth of research investigating the role of consumer brand origin knowledge in relation to the constructs identified in our study, we pos- tulate that consumer knowledge of brand origins will be fuelled by cos- mopolitanism(Samiee,Shimp,andSharma2005 )andconsumerethno- centrism (Alden,Steenkamp,andBatra2006 ;Shimp andSharma1987 ), and that a greater overall consumer cognizance of brand national ori- ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 361 gins will result in a greater tendency to purchase foreign rather than lo- cal products (Riefler and Diamantopoulos2009 ). For instance, having examined the relations between consumers’ overall origin classification performance and the degree of ethnocentrism, Balabanis and Diaman- topoulos (2008 ) found the classification performance for domestic as wellasforeignbrandswasthelowestforethnocentricconsumers.Hence, weproposethefollowinghypotheses: h3 Cosmopolitanism (cp ) has a direct and positive effect on consumer knowledgeofbrandorigins(kbo ). h4 Consumer ethnocentrism (ce ) is negatively related to consumer knowledgeofbrandorigins(kbo ). h5 Knowledge of brand origins (kbo ) is significantly and positively re- latedtoforeignproductpurchasebehavior(fppb ). Research Methods datacollectionandsamplecharacteristics The model for the study was tested via the store and outdoor intercept survey method using a sample of adult consumers in Slovenia. A quota sampling method based on gender, age and income was applied. The finalsampleconsistedof261 adultrespondentsin Slovenia.Womenand menwerealmostequallypresentedinthesample.Theaverageageofthe samplewasslightlyover45 years(sd of17 .29 ).Respondentswhoclaimed tohaveabove-averageorbelow-averagehousehold incomeswerealmost equallypresentedinthesample(18 .0 %and15 .9 %,respectively). instrumentdevelopmentandmeasures The measures were derived from the existing literature and adapted to the cultural context of the focal country following the guidelines estab- lished by Craig and Douglas (2000 ). The questionnaire was pretested on a conveniencesample of consumers, after which onlyminor amend- mentswere necessary. Cosmopolitanism was measured with Likert-type items selected from the worldmindedness scale used by Rawwas, Rajendran and Wuehrer (1996 ), who adapted the scale originally developed by Sampson and Smith (1957 ). The three specific items selected for this study are con- sistent with the recent specification of the conceptual domain of cos- mopolitanism (Riefler and Diamantopoulos2009 ) related to (a) gen- eral open-mindedness, (b) diversity appreciation and (c) consumption Volume 9 · Number 4 · Winter 2011 362 OliverPartsandIrenaVida transcending borders. Similar items have been recently used in Lee and Cheng’s (2008 ) study. To measure consumer ethnocentrism, the reduced fiveitemversionofcetscale (ShimpandSharma1987 )wasused,con- sistent with recent studies investigating this concept (Evanschitzky et al.2008 ; Balabanis and Diamantopoulos2004 ). We used a seven-point Likert-typescale,rangingfrom1 absolutelydisagreeto7 absolutelyagree, formeasuringbothpsycho-sociological variables. The measure of consumer knowledge of brand origins (kbo ) was de- velopedbasedonSamiee,ShimpandSharma’s(2005 )researchonBrand Origin Recognition Accuracy. Respondents were asked to identify the national origin of domestic and foreign brands in three different prod- uctcategories:alcoholproducts,clothesandfurniture.Participantswere presentedwithtwoforeignandtwodomesticbrandsineachoftheprod- uct categories; they had to correctly match each brandwith the country of origin from the list of six countries identified in our research instru- ment. If the respondents were unsure about the brandorigin, then they were instructed to make an educated guess, and only leave the question blank if they had no idea of the brand or its origin.kbo was evalu- atedinthealcoholproductgroupwithbrandslikeHeineken,Jägermeis- ter, Quercus, and Zlatorog with the following alternative national ori- gins:Italy,Germany,Netherlands, Russia,Slovenia,andScotland.Inthe clothesproductgroup,kbo wasidentifiedforthebrandsElkroj,Kappa, Lisca, and Zara with possible brand origins from among Croatia, Italy, Germany, Slovenia,Spain, andusa .kbo wasidentified in the furniture productgroupforIkea,Klun,LipBled,andScavolinibrandswithpossi- bleoriginsbeingFrance,Italy,Germany,Poland,Slovenia,andSweden. Comparing our kbo measure to the similar measure in Samiee, Shimpa and Sharma’s study (2005 ), the latter was clearly much more comprehensive in terms of the types of products and their nationalori- gins. Given the limited availability of both domestic and foreign brands inmanyproductcategories, thiswasnotattainableinasmallopenmar- ket economy like that of Slovenia. Moreover, similarly to the recent ori- gin classification performance study by Balabanis and Diamantopoulos (2008 ), knowledge of brand origins was measured collectively for do- mesticaswellasforeignbrands. Theforeign(vs.local)productpurchasingbehavior (fppb )constructin themodelwasmeasuredforalcoholproducts,clothes,andfurnitureus- ing a5 -point semanticdifferential scale, whereby one extreme indicated ‘Ibuyonlydomesticproductsinthisproductcategory.’andtheotherex- treme‘Ibuyonlyforeignproductsinthisproductcategory.’(eier2009 ). ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 363 table1 Scaleproperties, itemsandreliabilities (1)( 2)( 3 ) Cosmopolitanism–cp (Rawwasetal.1996 ) Likert-scalerangingfrom7 -absolutelyagree to1 absolutelydisagree ρvc=0 .56 ;ρr=0 .73 m =3 .67 ;sd =2 .09 Iprefertobe acitizen oftheworldrather thanofanyparticularcountry. 0 .666 Mygovernmentshouldallowforeignersto immigratehere. 0 .680 Productionlocationofaproduct doesnot affectmypurchasingdecisions. 0 .712 ConsumerEthnocentrism –ce (ShimpandSharma1987 ) Likert-scalerangingfrom7 -absolutelyagree to1 absolutelydisagree ρvc=0 .77 ;ρr=0 .94 m =3 .01 ;sd =2 .00 Sloveniansshouldnotbuyforeignproducts because thishurtsSlovenianbusinessand causesunemployment. 0 .876 Slovenianconsumerswhopurchase prod- uctsmadeinothercountriesareresponsible forputtingtheirfellowSloveniansoutof work. 0 .875 ArealSlovenianshouldalwaysbuy Slovenian-madeproducts. 0 .864 Itisnotrighttopurchase foreignproducts because itputsSloveniansoutofjobs. 0 .862 Weshouldbuyfromforeigncountries only thoseproducts thatwecannotobtainwithin ourowncountry. 0 .840 Continuedonthenextpage DataAnalyses andResults Data were analyzed via a structural equation modeling (sem )method using Lisrel8 .8 software. Following Gerbing and Anderson’s (1988 )rec- ommendations, a measurement model was analyzed first, followed by the evaluation of a structural model in order to assess the hypothesized relationshipsbetweenconstructs.Finalmodelitems,scalereliability,av- eragevarianceextractedandfactorloadingsarepresentedintable1 . Reliability of the scales was established using composite reliability (rho) which ranged from0 .73 to0 .94 –wellabovethe0 .7 recommen- dation by DeVellis (2003 ). The validity of each of the scales was tested with confirmatory factor analysis (cfa ). The final measurement model includedfour latent constructs and13 indicators used to measure them. The fit statistics of the model indicate a very good fit to the data with Volume 9 · Number 4 · Winter 2011 364 OliverPartsandIrenaVida table1 Continuedfromthepreviouspage (1)( 2)( 3 ) Foreign vs Domestic Purchase Behavior–fppb (adapted fromeier2009 ) Semanticdifferential scalefortypicalpur- chaseinspecificproduct category(anchored 5 –onlyforeignto1 –onlydomestic) ρvc =0 .66 ;ρr =0 .81 m =2 .80 ;sd =0 .90 Clothes 0 .823 Furniture 0 .776 Alcoholproducts 0 .711 KnowledgeofBrandOrigins– kbo (adaptedfromSamieeet al.2005 ) Three product categorieswithtwodomestic andtwoforeignbrandsandsixcountriesof originforeachbrand ρvc=0 .81 ;ρr =0 .92 Domesticbrandsorigins 0 .898 m =0 .69 ;sd =0 .21 Foreignbrandsorigins 0 .898 notes Columnheadings are as follows:(1)constructsandcoefficients –ave (ρvc)in cr (ρr),(2 )items,(3 )factorloading.m –meanvalue,sd –standarddeviation. rmsea of0 .046 andsrmr of0 .043 andotherindiceswellover0 .90 (gfi =0 .950 ,nfi =0 .950 ,nnfi =0 .980 ,cfi =0 .980 ,rfi =0 .940 ).The con- vergentvalidity ofscaleswastested through examinationofthe t-values of the Lambda-X matrix ranging from3 .45 to15 .88 ; all values were well above the2 .00 level specified by Kumar, Stern and Achrol (1992 ). The averagevarianceextracted (ave ) ranged between0 .56 to0 .81 ,exceeding 0 .50 for all constructs (Fornell and Larcker1981 ). Discriminant valid- ity was assessed by setting the individual paths of the Phi matrix to1 andtestingtheresultantmodelagainsttheoriginal(GerbingandAnder- son1988 ) using the D statistics (Joreskog and Sorbom1993 ). The high Dsquaredstatisticsindicatedthattheconfirmatoryfactormodelforthe scales fits significantly better than the constrained models for each con- struct,thus showingdiscriminantvalidity. Once the construct reliability, convergent validity and discriminant validitywereestablished,thestructuralmodelwasruninordertotestthe hypothesized relationships between constructs. The Chi-Squared statis- tic was significant, but the rest of the structural model fit measures in- dicate that the data conformed well to the model (i.e.,rmsea of0 .059 ; standardizedrmr of0 .052 –slightlyhigherthantherecommendedvalue of0 .05 ;gfi =0 .936 ,nfi =0 .939 ,nnfi =0 .961 ,cfi =0 .970 ,rfi =0 .921 ). Hypotheses were tested using t-statistics from the structural model. As seenintable2 ,theresultsofouranalysesconfirmedfourhypothesesout ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 365 table2 Hypothesestesting andresults (1)( 2)( 3)( 4)( 5)( 6 ) h1 Cosmopolitanism fppb 0 .13 3 .35 Supported h2 Cosmopolitanism ce –0 .38 –3 .19 Supported h3 Cosmopolitanism kbo 0 .01 0 .46 NotSupported h4 ce kbo –0 .03 –3 .95 Supported h5 kbo fppb 1 .29 3 .65 Supported notes Column headings are as follows: (1 ) hypothesis, (2 ) antecedent, (3 ) criterion variable,(4 )estimate,(5 )t-value, (6 )result. of five. We found a direct positive effect of cosmopolitanism onfppb (h1 ), a strong negative and significant relationship between cosmopoli- tanism and consumer ethnocentrism (h2 ), an inverse relation between consumer ethnocentrism and knowledge of brand origin (h4 ), and a positive and significant relationship between knowledge of brand ori- gin andforeign productpurchase behavior(h5 ).On theother hand,no support was found for the relationship between cosmopolitanism and consumerknowledgeofforeignbrands(h3 ). DiscussionandConclusions While cosmopolitanism has been widely studied in the management and marketing literatures, previous research has rarely explored the di- rect effects of cosmopolitanism on behavioral outcomes, as in the case offppb in our model (Cleveland, Laroche, and Papadopoulus2009 ; Sharma, Shimp, and Shin1995 ). Moreover, in examining consumer for- eign and domestic purchase behavior, consumer actual knowledge of brands’nationalorigin has seldom been accounted forin existing mod- els, even despite the growing concern that consumer knowledge of the product/brandnationalorigins tends to be inaccurateand superficial at best(BalabanisandDiamantopoulos2008 ;Liefeld2004 ;Samiee,Shimp, andSharma2005 ).Ourresults confirmthatcosmopolitanismexhibits a direct and positively significant effect onfppb , suggesting that the seg- mentofconsumerscharacterizedas‘worldcitizen’hasagreatertendency topurchaseforeignratherthandomesticbrandsinthethreeproductcat- egoriesinvestigated,i.e.,alcohol,clothesandfurniture. Our empirical study found no support for the direct relationship be- tweencosmopolitanismandconsumerknowledgeofbrandorigins(h3 ), suggesting that the worldly individuals who are open to foreigners do Volume 9 · Number 4 · Winter 2011 366 OliverPartsandIrenaVida notnecessarilymoreaccuratelyassessthenationaloriginofbrandsthan less cosmopolitan consumers. This hypothesis was largely exploratory in nature as we were able to identify only one study examining the impact of international experience on brand origin recognition accu- racy –bora (Samiee, Shimp and Sharma2005 ). In Samiee, Shimp and Sharma’s study,bora was measured separately for foreign brands and domesticbrands. Despite the conceptual confusion about the nature of cosmopoli- tanism as an antecedent of consumer ethnocentrism in some previous studies (e.g., Balabanis et al.2001 ; Shankarmahesh2006 ), our findings are in line with Sharma, Shimp, and Shin’s (1995 )original modelin that consumer positive orientation towards the out-groups directly affects an individual’sethnocentric tendencies, i.e., it reduces consumerpreju- dice towards imports, and ultimately (through consumer knowledge of brandorigins)affectspurchasebehavior.Moreover,weconfirmedthatin general, moreethnocentric consumersare less knowledgeableabout the overall brand origins. This is consistent with the findings of Balabanis andDiamantopoulos(2008 )whoconcludedthatconsumers’countryof origin classification performance is negatively related to the degree of ethnocentrism.Ontheotherhand,thisresult isonlypartiallyconsistent withSamiee,Shimp,andSharma(2005 ).Theseauthorsfoundthatce is positively related tobora for domestic brands but negatively tobora for foreign brands. Lastly, our empirical results suggest that consumer ability to correctly identify brands’ national origin is positively related to their purchase behaviors in favor of foreign products. While largely exploratory,weproposedandfoundthatconsumerknowledgeofbrand origins is a mediating variable between consumer ethnocentrism and purchase behavior in favor of foreign products. This finding suggests that more ethnocentric individuals possess poorer overall knowledge of brandoriginsthantheirlessethnocentriccounterparts,whichultimately leads to purchase preferences for domestic rather than foreign products inthe productcategoriesinvestigated inthisstudy. Understandingthedirectandindirecteffectsofconsumercosmopoli- tanismclearlyoffersvariousimplicationsforactionablemarketingprac- ticeinlocalaswellasgeographically andculturallydistantinternational markets. Using cosmopolitanism as a market segmentation variable, marketers can better understand the intensity of cosmopolitan values in their target segment and can ultimately effectively adapt the mar- keting mix to the local consumer preferences. This is particularly rel- ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 367 evant in branding activities and in the ability to develop prudent pro- motional campaigns. Our findings suggest that cosmopolitanism is a strong predictor of consumer behavioral preferences for foreign rather than local goods, and an equally effective predictor of consumer ethno- centrism. studylimitationsandfutureresearch In this research, deliberate efforts have been undertaken to utilize an externally valid consumer sample, solid measures and relevant analyti- cal methods to test the model. However, several limitations still apply, which, in turn, open questions for future research venues. In this study we examined the direct and indirect effects of consumer cosmopoli- tanismonconsumerpurchasebehavior in favorofforeignrelative todo- mesticpurchasebehavior collectively forthree categories of consumables (alcohol, clothes and furniture). Previous studies focusing on the role of socio-psychological constructs have shown that the impact of cos- mopolitanism and ethnocentrism varies according to whether the out- come measure is conceptualized as domestic or foreign consumption (Balabanis et al.2001 ; Suh and Kwon2002 ). Moreover, while some re- searchers demonstrated that product national origin affects consumer attitudes regardless of the product category (e.g., Ahmed et al.2004 ), others asserted that the effects tend to vary by product category (e.g., Balabanis et al.2001 ). Hence, future examinations of consumer foreign vs. domestic choice alternatives should attempt to overcome these limi- tations.Specifically,futurestudiesshouldincludeotherrelevantproduct categories, examine the cosmopolitanism effects independentlyforeach product category, and use independent measures of purchase behavior forforeignandfordomesticproducts. Our measure of consumer knowledge of brand origin was delimited to three product categories, with two domestic and two foreign brands and six national origins for each brand. Considering that respondents only matched a limited number of brands to the six countries of origin from our list, future studies will therefore need to improve the measure ofkbo and retest the direct relationship between cosmopolitanism and consumer knowledge of brand origins. An examination of the role of kbo inthemodel,separatelyfordomesticandforforeignbrands,would providevaluableinsightsaswell.Andlastly,acomparativestudyofother cultures and countries is recommended so as to ensure the model’s ex- ternal validity. In particular, a comparison between the mature and the Volume 9 · Number 4 · Winter 2011 368 OliverPartsandIrenaVida emerging markets would enable a deeper understanding of differences in the cosmopolitanism effects across markets based on their economic development,assuggestedinpreviouswork(Dmitrovi´ c,Vida,andRear- don2009 ;Dmitrovi´ candVida2010 ;Shankarmahesh2006 ;Strizhakova, Coulter,andPrice2008 ). References Aboulnasr, K.2006 .‘CountryofOriginEffects: The Role of Informa- tion Diagnosticity, Information Typicality and Involvement.’ Market- ingManagementJournal16 (1 ):1 –18 . Ahmed, Z. U., H. P. Johnson, X. Yang, C. F. Fatt, H. S. Teng, and L. C Boon.2004 . ‘Does Country of Origin Matter for Low Involvements Products?’International MarketingReview21 (1 ):102 –20 . Alden,D.L.,J.B.E.M.Steenkamp,andR.Batra.2006 . ‘Consumer Atti- tudes Toward Marketplace Globalization: Structure, Antecedents and Consequences.’ International Journal of Research in Marketing23 (3 ): 227 –39 . Balabanis,G.,andA.Diamantopoulos.2008 .‘BrandOriginIdentification by Consumers: A Classification Perspective.’ Journal of International Marketing16 (1 ):39 –71 . ———.2004 . ‘Domestic Country Bias, Country of Origin Effects, and ConsumerEthnocentrism: AMultidimensionalUnfoldingApproach.’ JournaloftheAcademyofMarketingScience32 (1 ):80 –95 . Balabanis, G., A. Diamantopoulos, R. D. Mueller, and T. C. Melewar. 2001 .‘TheImpactofNationalism,PatriotismandInternationalismon Consumer Ethnocentric Tendencies.’ Journal of International Business Studies32 (1 ):157 –75 . Cannon, H. M., and A. Yaprak.2002 . ‘Will the Real-World Citizen Please Stand Up! The Many Faces of Cosmopolitan Consumer Behavior.’ JournalofInternational Marketing10 (4 ):30 –52 . Cleveland, M., M. Laroche, and N. Papadopoulos. 2009 . ‘Cosmopoli- tanism, Consumer Ethnocentrism, and Materialism: An Eight-Coun- tryStudyofAntecedentsandOutcomes.’JournalofInternationalMar- keting17 (1 ):116 –46 . Craig, C. S., and S. P. Douglas.2000 . International Marketing Research. NewYork:Wiley. CrawfordJ.C.,andC.W.Lamb.1982 .‘EffectofWorldmindednessamong Professional Buyers Upontheir Willingness to BuyForeignProducts.’ Psychological Reports50 (3 ):859 –62 . Devellis,R.F.2003 .ScaleDevelopment:TheoryandApplications.Thousand Oaks,ca :Sage. ManagingGlobalTransitions TheEffectsofConsumerCosmopolitanismonPurchaseBehavior 369 Dmitrovic, T., I. Vida, and J. Reardon.2009 . ‘Purchase Behavior in Fa- vor of Domestic Products in the West Balkans.’ International Business Review18 (5 ):523 –35 . Dmitrovic, T.,and I. Vida.2010 . ‘Consumer Behaviour Inducedby Prod- uct Nationality: The Evolution of the Field and its Theoretical An- tecedents.’TransformationinBusinessandEconomics9 (1 ):145 –65 . eier .2009 . Elanike Toitumisharjumused Ja Toidukaupade Ostueelistused 1996 –2009 . Tallinn: EstonianInstituteofEconomicResearch. Evanschitzky,H.,V.F.Wangenheim,D.Woisetschläger,andM.Blut.2008 . ‘ConsumerEthnocentrismintheGermanMarket.’InternationalMar- keting Review25 (1 ):7 –32 . Fornell,C.,andD.F.Larcker.1981 .‘EvaluatingStructuralEquationModels with Unobserved Variables and Measurement Error.’ Journal of Mar- keting Research28 (1 ):39 –50 . Gerbing,D.,andJ.Anderson.1988 .‘AnUpdatedParadigmforScaleDevel- opmentIncorporatingUnidimensionalityanditsAssessment.’Journal ofMarketingResearch25 (2 ):186 –92 . Joreskog, K., and D. Sorbom.1993 .lisrel8 : Structural Equation Mod- eling with thesimplis Command Language. Chicago: Scientific Soft- ware. Kumar,N.,L.Stern,andR.Achrol.1992 . ‘Assessing Reseller Performance fromtheperspectiveoftheSupplier.’JournalofMarketing Research29 (2 ):238 –53 . Laroche,M.,N.Papadopoulos,L.A.Heslop,andM.Mourali.2005 .‘The Influence of Country Image Structure on Consumer Evaluations of ForeignProducts.’ International MarketingReview22 (1 ):96 –115 . Lee, S. T., and F. F. Chen. 2008 .‘CountryImageE ffect on Taiwanese Consumers’ Willingness to Buy from Neighboring Countries.’ Inter- nationalJournalofCommerceandRetailManagement18 (2 ):166 –83 . Levy,O.,S.Beechler,S.Taylor,andN.A.Boyacigiller.2007 .‘WhatWeTalk AboutWhenWeTalkAbout“GlobalMindset:”ManagerialCognition inMultinationalCorporation.’JournalofInternationalBusinessStudies 38 (2 ):231 –58 . Liefeld,J.P .2004 . ‘Consumer Knowledge and Use of Country-of-Origin Information at the Point of Purchase.’ JournalofConsumer Behaviour 4 (2 ):85 –96 . Merton, R. K.1957 . ‘Patterns of Influence:Localand Cosmopolitan Influ- ential.’ In Social Theory and Social Structure, edited by R. K. Merton, 387 –420 .NewYork:TheFreePress Pharr,J.M.2005 .‘SynthesizingCountry-of-OriginResearchfromtheLast Decade:Is theConceptStillSalient in an Eraof GlobalBrands?’ Jour- nalofMarketingTheoryandPractice13 (4 ):34 –45 . Volume 9 · Number 4 · Winter 2011 370 OliverPartsandIrenaVida Rawwas,M.Y.A.,K.N.Rajendran,andG.A.Wuehrer.1996 .‘TheInflu- ence of Worldmindedness and Nationalism on Consumer Evaluation ofDomesticandForeignProducts.’International MarketingReview13 (2 ):20 –38 . Riefler, P., and A. Diamantopoulos.2009 . ‘Consumer Cosmopolitanism: Review and Replication of thecymyc Scale.’ Journal of Business Re- search62 (4 ):407 –19 . Roth, K, and A. Diamantopoulos.2009 .‘ AdvancingtheCountryImage Construct.’JournalofBusiness Research62 (7 ):726 –40 . Samiee, S., T. A. Shimp., and S. Sharma. 2005 .‘BrandOriginRecog- nition Accuracy: Its Antecedents and Consumers’ Cognitive Limita- tions.’JournalofInternational Business Studies36 (4 ):379 –97 . Sampson, D. L., and H. P. Smith.1957 . ‘A Scale to Measure Worldminded Attitudes.’TheJournalofSocialPsychology45 :99 –106 . Shankarmahesh, M. N.2006 . ‘Consumer Ethnocentrism: An Integrative ReviewofItsAntecedentsandConsequences.’InternationalMarketing Review23 (2 ):146 –72 . Sharma, S., T. A. Shimp., and J. Shin.1995 . ‘Consumer Ethnocentrism: A Test of Antecedents and Moderators.’JournaloftheAcademyofMar- keting Science23 (1 ):26 –37 . Shimp, T. A., and S. Sharma.1987 . ‘Consumer Ethnocentrism: Construc- tion and Validation of thecetscale .’ Journal of Marketing Research 24 (8 ):280 –9 . Strizhakova,Y .,R.A.Coulter,andL.L.Price.2008 . ‘Branded Products as a Passport toGlobal Citizenship: Perspectives from Developedand Developing Countries.’ Journal of International Marketing16 (4 ):57 – 85 . Suh,T.,andI.-W.G.Kwon.2002 . ‘Globalization and Reluctance Buyers.’ International MarketingReview19 (6 ):663 –80 . Sumner,W. G.1906 . Folkways. NewYork:Ginn. Thompson, C. J., and S. K. Tambyah.1999 . ‘Trying to be Cosmopolitan.’ JournalofConsumerResearch26 (3 ):214 –241 . Vida,I.,T.Dmitrovic,andC.Obadia.2008 .‘TheRoleofEthnicAffiliation inConsumerEthnocentrism.’EuropeanJournalofMarketing42 (3 –4 ): 327 –43 . Vida, I., and J. Reardon.2008 . ‘Domestic Consumption: Rational, Affec- tive,orNormativeChoice?’JournalofConsumerMarketing25 (1 ):34 – 44 . Yoon, S. J., M. H. Cannon, and A. Yaprak.1996 . ‘Evaluating thecymic Cosmopolitanism Scale on Korean Consumers.’ Advance in Interna- tionalMarketing7 :211 –32 . ManagingGlobalTransitions