The impact of consumer ethnocentrism and consumer cosmopolitanism on consumption behaviour

Table of contents

I. Introduction

II. A literature review of consumer ethnocentrism
II.1. The concept of consumer ethnocentrism
II.1.1. Ethnocentrism
II.1.2. Consumer ethnocentrism
II.1.3. Related constructs
II.2. The measurement of consumer ethnocentrism
the CETSCALE
II.2.1. Construction of the CETSCALE
II.2.2. Psychometric properties
II.2.3. Cross-national assessment of the CETSCALE
II.2.4. The impact of consumer ethnocentrism on buying behaviour
II.2.5. Antecedents of consumer ethnocentrism

III. A literature review of consumer cosmopolitanism
III.1. The concept of cosmopolitanism
III.1.1. Definitions of cosmopolitanism
III.1.2. Conceptualizations of cosmopolitanism
III.1.3. Related concepts
III.2. The measurement of cosmopolitanism
III.3. The concept of consumer cosmopolitanism
III.3.1. Cosmopolitanism and consumer behaviour
III.3.2. Definitions of consumer cosmopolitanism
III.3.3. Conceptualization of consumer cosmopolitanism
III.4. The measurement of consumer cosmopolitanism: the CYMYC scale
III.4.1. Empirical studies on consumer cosmopolitanism
III.4.2. The need for a new scale
III.5. Summary of findings

IV. The new consumer cosmopolitanism scale
IV.1. Scale development
IV.2. Construct definition and domain of content
IV.2.1. Construct of consumer cosmopolitanism
IV.2.2. Conceptual dimensions
IV.2.3. Dimensionality
IV.2.4. Definition of consumer cosmopolitanism
IV.2.5. Conceptualization of cosmopolitanism and localism
IV.2.6. Validation items and the nomological network
IV.2.7. Indicators
IV.3. Item pool generation
IV.3.1. Domain sampling and generation of the item pool
IV.4. Response format
IV.5. Expert screening

V. Empirical study on consumer cosmopolitanism
V.1. Measures
V.2. Prestudy
V.3. Sample description
V.3.1. Sample size
V.3.2. Sample description
V.4. Analysis
V.4.1. Item analysis
V.4.2. Exploratory Factor analysis
V.4.3. Reliability analysis
V.4.4. Validity assessment
V.5. Concluding remarks

VI. Conclusion
VI.1. Summary
VI.2. Conclusion
VI.3. Future Research
VII. References

VIII. Appendix

Index of tables and figures

Table 1: Selection of relevant studies on consumer ethnocentrism

Table 2: Summary of the definitions of cosmopolitanism in sociological literature

Table 3: Construct descriptions of cosmopolitanism in the socio-cultural literature

Table 4: Summary of definitions of the constructs related to cosmopolitanism

Table 5: Summary of empirical studies on cosmopolitanism

Table 6: A Contingency Theory for Selecting Cross-National Segmentation Strategies

Table 7: Summary of the descriptions of consumer cosmopolitanism in marketing literature

Table 8: Classification of consumers’ cosmopolitan and local orientation.

Table 9: characteristics of the local and global cosmopolitan consumer

Table 10: Framework to classify cosmopolitans by the need for variety and the quality of experience

Table 11: Focus and methodology of the CYMYC cosmopolitanism studies

Table 12: Summary of settings and samples of the CYMYC studies

Table 13: Summary of the identified dimensions of consumer cosmopolitanism

Table 14: Summary of the coefficient alpha reported in the studies

Table 15: Summary of dependent variables and external validity of the CYMYC studies

Table 16: Meta dimensions and dimensions of consumer cosmopolitanism

Table 17: Comparison of the meta dimensions of cosmopolitanism and consumer cosmopolitanism

Table 18: Summary of the items of the consumer cosmopolitanism scale

Table 19: The consumer cosmopolitanism scale

Table 20: Overview of the correlations with the mean values of the validation constructs

Table 21: Coefficients table of the multiple regression analysis on willingness to buy

Table 22: Coefficients of the multiple regression analysis on product possession 124

Figure 1: Reduced Nomological frame 92 Figure 2: Output of the 3 rd EFA: Cattell’s scree plot

Index of appendix tables

Appendix table 1: CETSCALE of consumer ethnocentrism

Appendix table 2: CYMYC consumer cosmopolitanism scale

Appendix table 3: Correlation matrix

Appendix table 4: Factor loadings CYMYC prestudy

Appendix table 5: Dimensions sorting

Appendix table 6: Coder total agreement matrix

Appendix table 7: Prestudy questionnaire

Appendix table 8: Example of the online questionnaire at surveyz.com

Appendix table 9: Example of the participation appeal posted at the UWC Yahoo group

Appendix table 10: Summary of the sample’s nationalities and places of birth

Appendix table 11: Factor structure changes across the 6 EFAs

Appendix table 12: Overview of the eliminated items

Appendix table 13: Output of the 6 th EFA: total variance explained

Appendix table 14: Overview of the validation items

I. Introduction

“Through constellation of taken-for-granted meanings, beliefs, normative values, and implicit cultural connotations, ideologies become naturalized as part of a cultural way of life and situate individuals’ self-directed actions within a socio- cultural matrix of power relations, such as dominant/subordinate, central/marginal, and included/excluded”(Thompson & Tambyah 1999, p. 215). In view of that, and thus depending on the values promoted by a culture, people tend to acquire diverse cultural identities such as nationalism, patriotism, internationalism, world- mindedness, ethnocentrism and cosmopolitanism

Culture, commonly defined as being shared groups of people with distinctive beliefs, norms and customs (Hofstede 1984), is not a static phenomenon but evolves over time and thus in our day we are being confronted with a particular cultural tendency, namely the so-called worldculture. Today we live in a worldculture (Hannerz 1990), marked by an organization of diversity rather than by a replication of uniformity. The world has become one network of social relationships and between its different regions there is a flow of meanings as well as of people and of goods (Brennan 1997; Albrow 1996; Tomlinson 1999). It has been created through an increasing interconnectedness of varied local cultures and the development of cultures without a clear anchorage to a specific territory due to migration. We are testimonies of a growth of transnational social spaces (Roudomethof 2005), social fields and networks which are born out of an increasing internal globalization or glocalization¹. Thus transnational cultural networks have been created, in which working-class-cosmopolitans (Werbner 1999) move along prescribed pathways, rather than the one-way westernised flow of goods and images, the so-called McDonaldization of the world. It can be observed that processes of globalisation, transnationalisation and localisation are historically and culturally specific and that the economic pathways carrying goods and people criss-cross each other to create complex cross-national topographies. This phenomenon of worldculture or globalization is omnipresent. People can relate in different ways to this global interconnected diversity. The two main directions stated by the ethnic globalization paradox are: ethnocentric / parochial / nationalistic and cosmopolitan / glocal / transnational. The phenomenon of the ethnic globalization paradox (Turner 2002) aims to explain why in a time when nation state paradigms are breaking up from the inside, when the world is literally growing closer together, when barriers between nations and ethnic groups are being lifted, individuals can either adopt an open, encompassing attitude or on the contrary reinforce their ethnic identities leading to diverse manifestations of nationalism. Considering the paradigm that all knowledge claims, cultural representations and identity positions are constituted within ideological systems and therefore ideologies organize the practices of every day life, one can deduce that these general cultural tendencies can similarly be applied to the sphere of consumption (Beck 2002, 2004). In marketing though they are referred to as consumer cosmopolitanism (Cannon, Yoon, McGowan & Yaprak 1994) and consumer ethnocentrism (Shimp & Sharma 1987) and their impact on value driven consumption patterns is explored

The various impacts of these cultural tendencies on consumption behaviour are of central interest to the field of marketing. A crucial question, marketers are currently being confronted with, regards factors upon which consumers base their purchase decision when being confronted with a considerably broad range of products that are perceived similar or even alike. Certainly the decision largely depends on product judgement and overall perception of a product (Kroeber-Riel & Weinberg 2003), which can be subdivided into intrinsic information cues, such as product features and design, and extrinsic information cues, such as price and brand. As the choice of products are commonly not all made domestically but stem from a variety of different countries, Papadopoulos and Heslop (2003) infer that the country of origin effect ought to be considered as an additional extrinsic information cue explaining buying behaviour and as an important determinant of consumers’ bias against certain products. Consumers’ favourableness or unfavourableness towards products has been found to vary according to product specific origins (Peterson & Jolibert 1995), interacting as well with product category (Balabanis, Diamantopoulos, Mueller & Melewar 2001)

In the country-of-origin literature, “made in” influence is assumed to impact on consumers’ willingness to buy foreign products and on buying decisions indirectly via product judgments (e.g. Bilkey & Nes 1982; Wall, Liefeld & Heslop 1991; Papadopoulos & Heslop 1993; Peterson & Jolibert 1995; Verlegh & Steenkamp 1999). Therefore consumers’ attitude toward foreign and domestic products defined in terms of one’s cultural identity, is fundamental to the process of consumption decision making

Consumer ethnocentrism and consumer cosmopolitanism seem of particular interest in this context, as they constitute cultural identities which behave according to the assumptions of the globalisation paradox (Turner 2002), but are applied to the consumption sphere. While the former has been subject to many studies previously carried out (e.g.: Shimp & Sharma 1987; Netemeyer et al. 1991; Herche 1992; Balabanis et al. 2001, 2004; Douglas et al. 2003), the latter has only recently gained increasing attention in the marketing literature

Consumer ethnocentrism as an application of ethnocentrism to the economic sphere has inherited the main premises and properties of ethnocentrism. It is defined as a trait-like property of an individual's personality and encompasses the belief held by the consumers about the appropriateness and morality of purchasing foreign-made products, and thus an anxiety about harming one's beloved country when purchasing imports and a general prejudice against imports (Shimp & Sharma 1987). Shimp and Sharma (1987) have developed a multi-item scale, namely the CETSCALE, to capture consumer ethnocentric tendencies. The CETSCALE, a 17 items, 7 point Likert scale has been proved to be valid and reliable in a number of studies (e.g. Netemeyer et al. 1991; Herche 1992; Balabanis et al. 2001, 2004; Douglas et al. 2003). In these studies consumer ethnocentrism has been established as an independent construct explaining consumers’ preference of domestic over imported goods. Furthermore it has been demonstrated that consumer ethnocentrism serves as a better variable to predict import purchase than demographics

Consumer cosmopolitanism, on the other hand, the perspective of a cosmopolitan individual entails relationships to a plurality of cultures understood as permissive entities (Beck 2004). The notion of cosmopolitanism furthermore includes a stance towards diversity itself (Hannerz 1990) and towards the coexistence of cultures in the individual experience (Thompson & Tambyah 1999). It is first of all an orientation, a willingness to engage with others (Werbner 1999). It represents an intellectual and aesthetic stance of openness towards divergent cultural practices (Featherstone 2002). Hannerz (1990) retains cosmopolitanism to be as well a matter of competence regarding two aspects: readiness and a built-up skill of manœuvring, relating to considerations of the self as well as the other. Certainly it extends the etymological sense of world-citizenship

Similarly consumer cosmopolitanism is meant to describe a consumer whose orientations transcend any particular culture or setting, who seeks the best offer, simply accepting the cultural diversity of life and whose cosmopolitan perspectives are conditioned by experience of life (Cannon & Yaprak 2002; Thompson & Tambyah 1999). The factors driving economic cosmopolitanism are generally acknowledged as being competition, technological change, global communication, consumer experience, saturation of low level needs, the desire for quality, the desire for authenticity, the construction of a self-concept and value orientation in terms of cultural diversity (Cannon et al. 2002; Thompson et al. 1999; Cannon et al. 1998). Furthermore several authors propose conceptual dimensions of innovativeness (Gatignon, Eliashberg, Robertson 1989), communal, social and organizational orientation (Merton 1957; Gouldner 1957; Glaser 1963; Cannon & Yaprak 2002) and similarly cultural open-mindedness (Thompson & Tambyah 1999). One measurement instrument, namely the CYMYC Scale, has been developed to empirically capture consumers’ cosmopolitan tendencies. The CYMYC scale (Canon, Yoon, McGowan & Yaprak 1994) incorporates a value system model and globalization theories, and is based on the presumption that consumption behaviour is value driven. The scale however has only been used in a few studies in which its psychometric properties revealed to be on the whole rather disappointing and thus the scale still lacks scientific establishment

This diploma thesis seeks to empirically investigate these two potential key antecedents of consumption behaviour, namely consumer ethnocentrism and consumer cosmopolitanism, which assumingly have a negative or, respectively positive impact on foreign product evaluation and willingness or reluctance to buy foreign products. The fundamental questions to be discussed in this study are whether an individual’s ideological frame in terms of cultural identities influences the consumption decision regarding domestic versus foreign product choice and how the two phenomena relate to one another

To this aim, a conceptual model has been established, which illustrates the presumed relations between consumer cosmopolitanism and 1) the cultural antecedents such as cosmopolitanism and ethnocentrism 2) the related beliefs such as variety seeking and consumer ethnocentrism 3) the behavioural consequences in the sphere of consumption stemming from product perception, via willingness to buy to product possession. By means of the development of a new consumer cosmopolitanism scale, intended as an instrument that can validly measure a cosmopolitan mind-set, the way has been paved for a revision of this conceptual model, though considering only a limited number of factors due to the restricted breadth of a diploma thesis

This thesis is principally divided in seven parts. The first part is dedicated to a literature review on consumer ethnocentrism, drawing the bow from a definition and delineation of consumer ethnocentrism (section II.1), to an assessment of the CETSCALE (section II.2). The second part provides a literature review on consumer cosmopolitanism. Thus after a discussion of the conceptualization of cosmopolitanism in marketing and social-cultural research (section III.1), cosmopolitanism is demarcated from related cultural identities (section III.1.3) and measurement attempts of cosmopolitanism in the field of sociology and psychology are presented (section III.2). In the next step the ideation of cosmopolitanism is applied to the consumption sphere whilst the reconceptualizing of consumer cosmopolitanism and its importance and potential usage in marketing is underlined (section III.3). Consequently the various definitions are discussed and the dimensions of consumer cosmopolitanism are extracted. Next the CYMYC consumer cosmopolitanism scale is discussed (section III.4). Furthermore an in depth analysis of the CYMYC scale is undertaken, seeking to identify specific problems of the scale in terms of its content, dimensionality, reliability and validity. By means of a comprehensive examination of the CYMYC scale the need for a new scale is highlighted

The development process of the new consumer cosmopolitanism scale is illustrated from the construct definition (section IV.2) over the item pool generation (section IV.3) the choice of the response format (section IV.4), to the expert screening (section IV.5) and last of all the inclusion of the validation items

On the completion of these preliminary preparations, an exploratory investigation of the properties of the newly developed consumer cosmopolitanism scale was carried out by means of empirical study. The operational qualities of the study are presented (section V.1, V.2 and V.3) followed by an in depth analysis of the scale items’ characteristics which served to trim down the item pool (section V.4.1). By means of an exploratory factor analysis (section V.4.2) the dimensional structure of the concept was determined before proceeding with a detailed enquiry regarding the scale’s reliability, internal and external validity (section V.4.3 and V.4.4)

Lastly the present diploma thesis offers a discussion and conclusion and finishes with an outlook on possible future research efforts (section VI)

II. A literature review of consumer ethnocentrism

II.1. The concept of consumer ethnocentrism

II.1.1. Ethnocentrism

The general concept of ethnocentrism was introduced by Sumner (1906) in his publication about Folkways. Originally, it is a purely sociological concept to distinguish between “ingroups” and “outgroups” and to further explore their respective characteristics. The term “ingroups” refers to groupings with which an individual identifies whereas “outgroups” are regarded as antithetical to the “ingroup”

In general the conception of ethnocentrism represents the universal proclivity for people to view their own group as the centre of the universe, to interpret other social units from the perspective of their own group, and to reject persons who are culturally dissimilar while blindly accepting those who are culturally like themselves (Booth 1979). Correspondingly the symbols and values of one’s own ethnic or national group become objects of pride and attachment, whereas symbols of other groups may become objects of contempt (Campbell et al. 1972)

Over the years, sociological ethnocentrism was introduced into psychosocial research and the concept similarly accounted for the level of individual personality as well as for the more general cultural- and social analytic frameworks (Levine & Campbell 1972)

II.1.2. Consumer ethnocentrism

Due to limited diffusion of foreign-manufactured goods by means of international trade, domestic products have historically provided the sole frame of reference whereby consumers have evaluated imported goods whereas foreign products were often considered inferior or abstained. At the rise of the era of globalized markets, an ever-increasing tendency manifests itself to consider foreign-made goods as valuable alternatives to national-made products. However, it can be observed that some consumers still persistently refuse to buy imported goods to the extreme reaction of condemning fellow consumers for doing so. Their disapproval is to be explained by their conviction that buying foreign products hurts the economy, causes unemployment among co-nationals, or is considered blatantly unpatriotic (Shimp & Sharma 1987)

The concept of consumer ethnocentrism is adapted from the general concept of ethnocentrism in order to suit consumer behaviour and allow for marketing interventions. Though the general applicability of ethnocentrism to the study of consumer behaviour was already recognized in the 1970s (Markin 1974; Berkman & Gilson 1978), it remained to Shimp and Sharma (1987) to reformulate the concept specifically to suit the field of marketing and consumption motivation or reluctance. Consumer ethnocentrism as introduced by Shimp and Sharma (1987, p.280) is defined as “the beliefs held by American consumers about the appropriateness, indeed morality, of purchasing foreign-made products”

Ethnocentric consumers, thus, hold firm to their conviction that purchasing imported goods is erroneous because it would potentially hurt the domestic economy, cause a loss of jobs, and would be plainly unpatriotic. Put differently, products from other countries (perceived as outgroups) are undesirable highly ethnocentric consumers. In contrast, to non-ethnocentric individuals foreign products are objects to be evaluated on their own merit without consideration of their origin or possibly even appraisal due to their “exotic” origin (Shimp & Sharma 1987)

Consumer ethnocentrism endows the individual with a sense of identity, feelings of belonging, and, most importantly in reference to consumption, an understanding of which purchase behaviour is acceptable or unacceptable to the ingroup. Accordingly, in the evaluation of domestic products ethnocentric consumers tend to demonstrate a positively biased behaviour, overrating the qualities of nationally manufactured goods. Shimp and Sharma (1987) thus proposed consumer ethnocentrism as a construct that may partially explain why consumers evaluate domestic products more favourably than foreign goods. The underlying concept incorporates the emotional dimension of purchasing imported goods and the implication that the choice in some way threatens the domestic economy or even national security

Several studies have been conducted in order to further explore the proclivity of consumers to display ethnocentric tendencies (see Table 1). Table 1 is meant to provide a brief overview of selected publications on consumer ethnocentrism, whereas the detailed discussion of these empirical studies will follow in section II.2

Table 1: Selection of relevant studies on consumer ethnocentrism

Abbildung in dieser Leseprobe nicht enthalten

II.1.3. Related constructs

Nationalism and patriotism are the most relevant cultural concepts which are closely related to ethnocentrism. Despite being closely related, they are distinct concepts. In the following, concept descriptions of nationalism and patriotism are provided demonstrating the distinctiveness

Nationalism together with patriotism are the most commonly cultural identities mentioned in the sphere of economic behaviour, often erroneously perceived as interchangeable synonyms of consumer ethnocentrism and moreover often used interchangeably

However, although being similar, the three concepts of nationalism, patriotism and ethnocentrism have been shown empirically not to overlap, as will be discussed in the section II.3.3, dealing with the impact of antecedents on consumer ethnocentrism

Patriotism

Patriotism refers to “a commitment – a readiness to sacrifice for the nation” and “a strong feeling of attachment and loyalty to one’s country without the corresponding hostility towards other nations” (Balabanis et al. 2001, p.158, p.160)

Already Adorno (1950) explicitly explored the distinction between healthy patriotism (love of country) and ethnocentric patriotism or pseudopatriotism². Kosterman and Feshbach (1989) empirically supported this division and delineated the concepts by introducing three meaningful factors, namely nationalism (equivalent to Adorno’s pseudopatriotism), patriotism (equivalent to Adorno’s healthy patriotism) and internationalism³

Thus, patriotism may be viewed as a synonymous term for “love for one’s country”

Nationalism

Nationalism, on the other hand, is defined as “a commitment plus exclusion of others, a readiness to sacrifice bolstered by hostility towards others” (Druckman 1994), or as Balabanis et al. (2001, p. 160) put it, “n ationalism encompasses views that one’s country is superior and should be dominant and thus implies a denigation of other nation”

Nationalism as a concept has been linked to collective behaviour, governmental policies and to individual behaviour in the context of elections, inter-racial relations, stereotyping and xenophobia (Billig 1995). To sum up, nationalism reflects a commitment to one’s country accompanied by hostility towards others, whereas patriotism lacks this aspect of resentment

Although all levels of nationalism result in implications for international marketing, consumer ethnocentrism represents a distinct phenomenon as it focuses hereby upon the socio-psychological barriers that impose constraints on the consumers regarding the purchase of now more easily available foreign products

These definitions of related cultural constructs were intended to underline the distinctive features of consumer ethnocentrism and thus allow for a clear delineation

II.2. The measurement of consumer ethnocentrism: the CETSCALE

he classic measure of ethnocentrism is the California Ethnocentrism Scale by Adorno et al. (1950). This scale, however, was developed for sociological purposes and, thus, consisted of items not directly relevant for consumption behaviour studies⁴. For this reason, Shimp and Sharma (1987) developed the CETSCALE (consumer ethnocentric tendencies scale), which is a 17-item scale assessing specifically consumers’ ethnocentric tendencies for marketing research (see Appendix table 1). The CETSCALE was intended as a measure of tendency rather than attitude, because the latter term suggests a greater degree of object specificity than the CETSCALE was intended to capture. According to Shimp and Sharma (1987), tendency captures the more general notion of a disposition to act in some consistent fashion towards foreign products in general, whereas attitude is used most appropriately in reference to a consumer’s feelings towards a specific object

II.2.1. Construction of the CETSCALE

The authors based the domain specification and item generation on the results of an exploratory study asking American consumers to describe their thoughts and feelings about buying foreign-made products. Seven facets of consumer orientation emerged, but after scale purification and refinement, only four facets were retained. These were consumer ethnocentric tendencies, price-value perceptions, self-interest and reciprocity norms, and the CETSCALE⁵ was found to be psychometrically uni- dimensional

Shimp and Sharma (1987) conducted four separate studies to assess the reliability and the construct validity of the final 17-item CETSCALE among US consumers and students. The Four-areas study was carried out via mail questionnaires among consumers in metropolitan areas of Detroit (N= 322), Denver (N= 323), Los Angeles (N= 315), North and South Carolina (N= 575). Respondents of the Carolinas study⁶ were surveyed two times for assessing test-retest reliability. The National-consumer-goods study was a national telephone quota survey of consumers. The Crafted-with-pride-study consisted of a test-retest college student survey with 145 participants

II.2.2. Psychometric properties

The CETSCALE proved already in the establishing studies (Shimp & Sharma 1987) to rigorously adhere to the criteria of psychometrically sound scales in terms of content, dimensionality, reliability and validity. However, a detailed discussion of the scale’s psychometric characteristics will be provided in the following sections while including various additional studies (e.g. Netemeyer et al. 1991; Herche 1992; Douglas et al. 2003;) in order to compensate for the limits of Shimp and Sharma’s study. In terms of conceptualization and sample their study was restricted to the contemporary white American society and thus rendering the scale’s application and the implied transfer of its psychometric properties to other cultures problematic (section II.2.3)

Furthermore as in theory the construct of consumer ethnocentrism is not product specific (Shimp & Sharma 1987), their study failed to investigate the predictive power regarding purchase behaviour across product categories, which was in later studies (Herche 1992) found to be inconsistent (section II.2.4)

These critical aspects regarding the CETSCALE were assessed in further studies and will in turn be discussed

Dimensionality

In the study conducted by Netemeyer et al. (1991), the CETSCALE’s unidimensional structure across all four countries⁷ was confirmed by means of confirmatory factor analysis as the correlations among the CETSCALE items are consistent with the hypothesized uni-dimensional structure⁸. The pattern of factor loadings proved invariant across the four countries determined by multiple group analysis and coefficient of congruence

Reliability

Regarding Shimp and Sharma’s (1987) study, the CETSCALE performed brilliantly on the reliability tests with Cronbach’s co-efficient alpha for the four studies ranging from 0.94 to 0.96; and the test-retest reliability in the college sample yielding up to

0.77. These results indicate that the CETSCALE is a reliable measure of consumers’ ethnocentric tendencies

Netemeyer et al. (1991) further assessed the CETSCALE’s composite reliability (analogous to co-efficient alpha) which showed to be high and nearly equal across all countries ranging from 0.91 and 0.95. The variance extracted (i.e. the amount of variance captured by a construct's measure in relation to variance due to random measurement error) exceeded the advocated level of 0.50 (Netemeyer et al. 2003) Similarly, item loadings and item-to-item correlations provided further support for the internal consistency of the CETSCALE across all four countries under examination

Validity

Shimp and Sharma (1987) assessed the external validity and convergent validity in the Carolinas study by relating the CET score to prior information on automobile ownership, purchase intentions, attitudes and beliefs. Discriminant validity was assessed in the crafted-with-pride study in relation to patriotism, polito-economic conservatism and dogmatism

Netemeyer et al.’s (1991) study further delineated the CETSCALE’s discriminiant validity as it investigated the correlation of consumer ethnocentrism with attitude towards the home country which turned out to be non significant

Nomological validity

Shimp and Sharma (1987) established the nomological validation in terms of correlation with measures of attitude toward foreign-made products, general feelings towards foreign-made products, attitude towards buying American-made apparel and intention to purchase American-made apparel, importance of Country-of-origin country bias, cognitive and attitude measure of car ownership. Two general predictions were formulated: H1) CETSCALE scores should negatively correlate with beliefs, attitudes, purchase intentions towards foreign made products and H2) CETSCALE scores should be negatively correlated with foreign product behaviour. All correlations were found to be strong and significant, the hypotheses were supported and therefore more than sufficient evidence of nomological validity was provided

The study carried out by Netemeyer et al. (1991) yielded results confirming nomological validity by demonstrating that the correlation of the CETSCALE score with the importance of buying domestic products was significant whereas the correlation with the importance of buying the specific product category domestically was only significant for two of four countries. However, as most correlations of the CETSCALE with attitudinal, belief and ranking variables were significant and, importantly, the pattern of correlation was consistent across each country’s sample, further evidence of nomological validity is provided

II.2.3. Cross-national assessment of the CETSCALE

Given a rising concern for the development of measures that are useful in assessing consumer attitudes and preferences for both domestic and foreign products, the paramount issue in such an effort is the cross-nationally valid psychometric property of the respective scale. The scale certainly should not reflect artefacts due to unreliability and lack of validity

Thus establishing equivalence in cross-cultural research is viewed as key in making valid cross-cultural comparisons. According to Douglas and Nijssen (2003) equivalence is viewed as specific to a given cross-cultural comparison, and as a function of the characteristics of the research instrument, and the cultural groups or contexts compared. Three different levels of equivalence have been identified: construct equivalence, measurement unit equivalence and scalar equivalence (Douglas et al. 2003). Construct equivalence requires that a construct or concept occurs and has the same underlying meaning or significance across countries or cultures

Up to the study by Netemeyer, Durvasula and Lichtenstein (1991), the degree of cross-national correspondence between the CETSCALE and the construct of consumer ethnocentrism was not investigated, and consequently the applicability of the consumer ethnocentrism construct for understanding consumer behaviour across different cultural settings was not clarified

Netemeyer et al. (1991) carried out their study with a sample of 71 US business students, 73 German, 70 French and 76 Japanese students. They administered to their sample the short ten-item CETSCALE⁹ (Shimp & Sharma 1987) together with scales¹⁰ assessing the general attitude toward the home country, the importance of buying home products, the general attitude toward purchasing foreign products, the attitude toward buying foreign products of a specific product category¹¹ and the general attitude toward buying products from specific foreign countries i.e. USA, France, Germany and Japan

The CETSCALE’s unidimensional structure across all four countries was confirmed by means of confirmatory factor analysis and the pattern of factor loadings proved invariant across the four countries determined by multiple group analysis and coefficients of congruence

Composite reliability (analogous to co-efficient alpha) was high and nearly equal across all countries ranging from 0.91 and 0.95. Item loadings and item-to-item correlations provide further support for the internal consistency of the CETSCALE across all 4 countries

Nomological validity yielded results demonstrating that the correlation with the importance of buying domestic products is significant, the importance of buying the specific product category domestically is only significant for 2 countries, however as most correlations of the CETSCALE with attitudinal, belief and ranking variables are significant and, most importantly, the pattern of correlation is consistent across each country’s sample

In later studies Sinkovics (1999) and Douglas, Nijssen (2003) further investigated the cross-national applicability of the CETSCALE

Douglas and Nijssen (2003) conducted their study in the Netherlands, which provide a markedly different context from previously examined countries, with regard to factors related to ethnocentrism. These include for instance nationalism, national pride in terms of patriotism due to its position in Central Europe, country size, small size of the internal market, the resulting high level of international trade, imports and the individual level constant interactions with other nationalities in form of extensive travel practice. Consequently operationalizing the identical US-concept was expected to pose severe problems in the Netherlands marked by distinctive cultural features. The sample consisted of 1253 Dutch consumers. Douglas and Nijssen (2003) on the contrary, opted for a plurality of CETSCALE in order to allow for comparisons, namely the 10-item short version of the CETSCALE in a literal translation, the original 17-item CETSCALE in a literal translation and a version adapted to the national context. Furthermore, in-depth interviews were carried out. The product categories for the purpose were soft drinks and beer from the Netherlands and Germany, because of considerations of availability of a domestic alternative. Though the factor structure¹² differed from all previous research efforts regarding the CETSCALE, this investigation revealed no severe psychometric shortcomings of the application of the CETSCALE in a cross-national context. However, the authors underline that the use of “borrowed” scales is fraught with danger, in particular where the scale is used in a country or context that differs substantially in certain key respects from that in which the scale was originally developed. The attendant pitfalls may be less evident, associated with the size of an economy and its openness to other cultural influences and ideas

Sinkovics aimed to construct and validate a consumer ethnocentrism scale adapted to the Austrian context. The transfer of the ethnocentrism construct was carried out in a culture sensitive mode in order to integrate Austrian culture specific aspects, but based on the work of Shimp and Sharma (1987). Sinkovics’ (1999) two samples (N

= 1105 and N = 1084) were sought to be representative of the Austrian population. The empirical project consisted of five research stages and in the end the AT- CETSCALE was extracted through qualitative and quantitative methods of data reduction from a large item-pool. The initial scale length of 35 items was found to be highly reliable in the Austrian context but too long for an economic application and therefore further reduction took place. By means of item-to-item correlation, Cronbach’s Alpha, confirmatory factor analysis (sought unidimensionality) and lastly qualitative interpretation by rating experts the size of the scale was shortened. In the end two versions of the scale were developed to allow for parallel forms of testing and multiple forms of usage. The focus remained on the type A - AT Scale with 20 items. Sinkovics’ study revealed an important particularity of the Austrian context or generally of small countries to consider in further research, and that is the problematic aspect that Austrian consumers, despite their possible ethnocentric identities, are often constrained in their preference for domestic products due to problems of availability of a national alternative

The psychometric stability of the CETSCALE has certainly been confirmed (Netemeyer et al. 1991), whereas the findings of both Douglas & Nijssen (2003) and Sinkovics (1999) allude to the substantial influence of moderating factor such as the availability of domestic alternatives

II.2.4. The impact of consumer ethnocentrism on buying behaviour

The primary objective stated in the development process of the CETSCALE (Shimp

& Sharma 1987) was the establishment of consumer ethnocentrism as an explanatory construct of domestic versus foreign product choice and, thus, the creation of an instrument, which displays consistent predictive power regarding consumers’ buying behaviour. Furthermore in the developmental process of the CETSCALE, Shimp and Sharma (1987) proposed that the notion of consumer ethnocentric tendencies is based on a general predisposition toward or against all foreign products, hence the construct is not product specific. Therefore the focus of this section is placed on the discussion of the predictive validity of the CETSCALE and its impact across product categories. In addition, a review of the effects of moderating factors, i.e. antecedent constructs, or demographics, on consumer ethnocentrism is undertaken

Predictive validity of the CETSCALE

Shimp and Sharma (1987) already showed that the scale correlates with consumers’ product evaluations (National consumer good study), intent to purchase and actual possession of domestic products (Four-areas and Carolinas study) and preference for domestic products (crafted-with-pride study). Thus, at first impression, the CETSCALE’s underlying construct of consumer ethnocentrism seems to be a promising predictor variable for consumption behaviour

However, criticism voiced with respect to the construction of the CETSCALE by Shimp and Sharma (1987) regarded its incremental ability to predict import purchase behaviour beyond the use of demographic variables as predictors. Taking on this criticism, many authors have conducted research, from which, for the present discussion Herche (1992) and Balabanis et al. (2004) have been chosen for their focus on the aspect of predictive validity across product categories

Herche (1991) administered the original CETSCALE and an altered version to 1000 personal computer owners and 1000 recent-model automobile owners, by means of US wide mail survey among specialist journal readers. The results of Herche’s study revealed distinctive predictive validities across product categories, however ranging between 10% and 30% in explaining variance. These findings suggest that the CETSCALE's contribution for understanding import buying behaviour is superior to that of demographic variables. However, it seems also to suggest that the scale's predictive validity might be product specific as the amount of variance explained was shown to be higher for the purchase of cars than for computers (as will be discussed in detail in the subsequent subsection)

In a similar vein Balabanis and Diamantopoulos (2004) examined the predictive power of consumer ethnocentrism for domestic versus foreign products. Additionally, the authors sought to determine whether the predictive power arises independently of country characteristics. According to the findings of Balabanis and Diamantopoulos (2004) consumer ethnocentrism facilitates the prediction of product rankings for domestic products, as well as it helps to foresee bias towards home products. However, bias against foreign products was not well explained in their study. Neither economic competitiveness of the country of origin of the imported product nor cultural distance between a consumer’s country and the country of origin of the imported good was found to influence consumption behaviour. Therefore one can deduce that the phenomenon of consumer ethnocentrism is independent of the product’s country of origin. As a consequence it appears to be sufficient to use in further research the generic term of “foreign country origin” of a product as opposed to specifications of the actual country of a product’s origin

The impact of product categories

The impact of product categories on the predictive power of the CETSCALE has been assessed in a number of studies but also in Netemeyer et al.’s (1991) study. Here, respondents were asked not only to evaluate products in general, but also two specific product categories, i.e. cars and TV-sets, from the US, Germany, France, and Japan. The correlation of the CETSCALE with the importance of buying domestic products was significant and positive across all four countries, and with the exception of the German sample, the correlation with the general attitude toward buying foreign products was significant and negative across countries. However, the correlation with the general attitude toward buying a foreign car is significant for only two of the four countries, i.e. USA and Japan. Regarding the ranking variables, the results afford evidence of the validity of the CETSCALE in offering a consistent pattern in the predicted directions

Similarly, one objective of Herche’s (1992) study was to ascertain whether the scale’s ability to predict purchase behaviour differs across two product categories, i.e. personal computers and cars. The findings of Herche (1992) suggested that consumers change their preference maps across product categories, therefore consumer ethnocentrism, domestic-country-bias and country-of-origin ought to be examined whilst accounting for the product category. Distinctive predictive validities were revealed: for automobile owners the reduced model explains 10% of the variance in purchase behaviour, while the full model explains 30%, whereas the CETSCALE's contribution to explained variance is less pronounced among computer owners (4.6% for the reduced model and 10% for the full model)

The focus of the research carried out by Balabanis and Diamantopoulos (2004) was once more on the consistency of consumer preference for domestic products across different product categories. The eight product categories chosen for the study were cars, food products, TV-sets, toiletries, fashion wear, toys, do-it-yourself equipment, and furniture from six economically highly developed countries: Britain, US, France, Germany, Japan and Italy. These categories were selected because a large percentage of those products are imported, they represent an important expenditure for UK consumers and a domestic alternative is available. According to the findings of Balabanis & Diamantopoulos (2004), due to consumer ethnocentrism, the product rankings for domestic products vary across product categories. Thus, these studies confirm the relevance of product categories. Herche (1992) attributes this unexpected discrepancy to the consumers’ limited knowledge concerning the country of origin of certain products, because their place of manufacture is rarely visible or pronounced. Furthermore, he suggests that products with a larger economic impact of transaction evoke stronger ethnocentric sentiments due to the increased emotional involvement. Another possible explanation is, as already suggested on different occasions, the availability of acceptable domestic alternatives, which if not guaranteed will force even the ethnocentric consumers to purchase imported goods (Balabanis & Diamantopoulos 2004)

Moderating factors

Furthermore, Shimp and Sharma (1987) investigated whether consumer ethnocentrism operates uniformly across all consumers. Referring to Brehm’s (1966) psychological reactance theory, they assumed that certain population segments, whose quality of life and economic status are especially threatened by foreign competition, would be likely to exhibit disproportionately higher ethnocentric tendencies. Such segments would be, for example, lower social classes or residents of geographic areas where foreign competition is particularly fierce. This assumption was proved empirically showing that older working class-individuals manifested stronger ethnocentric tendencies since these individuals are particularly threatened by the prospect of losing jobs to foreign competition. Similarly the sample stemming from the Detroit area exhibited ethnocentric sentiments, which could be explained by the decline in American manufacturers’ share of the domestic automobile market, in that time hitting this area particularly severe

Similarly, Sinkovics (1999) considered moderating factors such as the relative importance of the product and economic threat imposed by the foreign producing country upon the national economy. In the Austrian context no such effect could be demonstrated, reducing the prognostic power of ethnocentrism regarding the attitude towards imported goods

The demographic factors i.e. age, gender, educational level, income and number of stays abroad have been assessed throughout the studies (Shimp & Sharma 1987; Netemeyer et al. 1991; Herche 1992; Sinkovics 1999). All these elements have been confirmed to have a significant positive influence on an individual’s ethnocentric sentiments, implying that an elderly woman of low education and income, with little travelling experience abroad, is statistically expected to display ethnocentric traits

Hence, the CETSCALE might not measure a simple static trait, but rather a complex dynamic state varying in its impact on buying habits based on influences, such as consumers’ home country, product categories, availability of domestic alternatives and demographics

II.2.5. Antecedents of consumer ethnocentrism

Another area of interest with respect to consumer ethnocentrism are its conceptual antecedents. Balabanis et al. (2001) investigated the impact of patriotism, nationalism and internationalism¹³ as antecedents to consumer ethnocentrism cross- nationally. All three concepts are centred on one’s general attitude towards one’s country and those towards other countries. Similarly the purpose of the study conducted by Balabanis and Diamantopoulos (2004) was to examine the relationship between country-of-origin, consumer ethnocentrism and domestic country bias¹⁴ in an integrative manner. Balabanis et al. (2001) administered a 10-item short version of the CETSCALE¹⁵ and scales of patriotism, nationalism, internationalism (Kosterman & Feshbach 1989) were run on samples of 303 Turkish consumers and 480 Czech consumers

In the study of Balabanis et al. (2004), given the aim of examination of preference patterns of customers for domestic products and specific foreign goods, the 10-item version of the CETSCALE, a country preference rating, a current competitiveness index and cultural distance index were administered to a sample in the United Kingdom. 465 consumers participated with a self-completion questionnaire. Regarding the Balabanis et al.’s (2001) study, the findings suggested that the over and above consideration of the potential influence factors, patriotism, nationalism and internationalism, brought about significant change in the proportion of the variance explained. Furthermore, it was revealed that they had a collective impact on consumer ethnocentrism. However, the results demonstrated that neither patriotism nor nationalism had a consistent influence on consumer ethnocentrism since their effects varied from country to country due to historically determined differences in terms of national and cultural identities (Balabanis et al. 2001). In both samples, internationalism did not affect consumer ethnocentrism and Balabanis et al. (2001) attributed the inward looking tendency to severe internal economic development problems. It can therefore be deduced from the results that patriotism and nationalism did not automatically translate into bias for foreign products in all countries

Vida, Dmitrovic, Reardon (2005) concentrated on the determination of the factors underlying consumers’ preferences for domestic products, namely consumer ethnocentrism, cosmopolitanism, national identity, country-of-origin and their interrelation. For that purpose Vida et al. (2005) combined items of the CETSCALE, CYMYC Scale, and National Identity Scale¹⁶ in their study. The research took place in four major cities in Serbia and the two largest cities in Montenegro with 600 individuals from Serbia and 300 individuals from Montenegro. Vida et al. (2005) examined six hypotheses¹⁷ of which only the following is relevant for this chapter: H5) consumer national identity is positively related to consumer ethnocentrism (not significant). Thus national identity as an antecedent of consumer ethnocentrism and the evaluation of domestic products is not clearly established

The discussion of the efforts in research has shown that the CETSCALE developed by Shimp and Sharma (1987) exhibits substantial predictive power regarding consumption and in terms of product evaluation (e.g. Netemeyer et al. 1991; Herche 1992; Vida et al. 2004) and product choice (e.g. Sinkovics 1999; Balabanis et al. 2001 and 2004). Whereas for the possession of domestic products no such relation could be confirmed (Sinkovics 1999; Vida et al. 2004). Furthermore several empirical studies demonstrated that the predictive power depends to a large extent on the product categories under examination (Herche 1992; Sinkovics 1999; Balabanis, Diamantopoulos 2004), which in turn can be explained by the availability of a national product alternative (e.g. Sinkovics 1999; Douglas et al. 2003) and the relative importance of the product to be purchased (e.g. Netemeyer et al.1991; Herche 1992). Although it has been shown that the scale is applicable in a cross- national context (Netemeyer et al.1991; Sinkovics 1999; Douglas et al. 2003), a cautious application in terms of adaptation to country specific particularities and appropriate translation is fundamental (Douglas et al. 2003). Furthermore consumer ethnocentrism has been successfully incorporated in a nomological network on a theoretical basis regarding nationalism, patriotism and internationalism but the attempt to empirically prove consumer ethnocentrism’s relation to its conceptual antecedents has not yielded uniform results across the countries under examination (Balabanis et al. 2001). Lastly, consumer ethnocentrism is a good predictor of domestic product preferences (Balabanis & Diamantopoulos 2004)

III. A literature review of consumer cosmopolitanism

III.1. The concept of cosmopolitanism

On an etymological basis the term „ cosmopolitan “, is derived from the Greek words kosmos (= world) and politis (= citizen). Hence, cosmopolitanism describes a citizenship of the world. The Oxford Encyclopaedic English Dictionary adds to this definition the notion of being “free from national limitations and prejudices”. Thus, being a citizen of the world means having a cultural disposition which is not limited to the concerns of the immediate locality, but which recognizes global belonging, involvement and responsibility and can integrate these broader concerns into every day life practices (Tomlinson 1999). Nowadays, this idea of citizenship has specific implications in various spheres from the political, the social to the economic, especially in the context of globalization. Relevant for this purpose is the cultural concept which has gained considerable attention in the social science literature

The concept of cosmopolitanism¹⁸ was first introduced by Merton (1957) and

Gouldner (1957) independently. Subsequently, two distinct research streams using cosmopolitanism as a personal orientation and an organizational orientation, respectively, evolved. The construct has been applied in the literature of innovativeness (e.g. Helsen, Jedidi & Desarbo 1993; Jeffres et al. 2004) and information diffusion (e.g. Gatignon, Eliashberg & Robertson 1989). The research in the field of organizational orientation has been amplified (e.g. Grimes & Berger 1970; Robertson & Wind 1983). The sociological literature furthermore provides an ever increasing body of publications on cosmopolitanism focusing on the clarification of the conceptual antecedents of cosmopolitanism and its behavioural consequences (e.g. Hannerz 1990; Turner 2002; Roudomethof 2005). In the marketing literature the behavioural consequences of cosmopolitanism are primarily considered as they also refer to manifestations in consumption behaviour. Cannon et al. (1994) were the first to explicitly explore the notion of cosmopolitanism in a

[...]

---

¹ The term internal glocalization in this context refers to the idea of exposure to other cultures on a regular basis without crossing borders by the means of communicational tools, the encounter with immigrants, tourists, refugees, cultural artefacts and commercial establishments

² Pseudopatriotism is described as a “blind attachment to certain national cultural values, uncritical conformity with the prevailing group ways, and rejection of other nations as out-groups” (Adorno et al. 1950 quoted in: Balabanis, Diamantopoulos, Mueller & Melewar 2001, p.160)

³ In Kosterman and Feshbach’s (1989) study internationalism emerged as a distinct dimension and not just as the polar end of nationalism, as it had been presumed so far

⁴ However, initially Shimp and Sharma (1987) included items capturing patriotism and politico- economic conservatism from the Adorno-scale in the CETSCALE but later eliminated for reasons of consistency

⁵ which represents the first facet

⁶ The Carolinas study consisted of a sub-sample of the Four-areas study of 417 respondents

⁷ The four countries were France, Germany, USA and Japan

⁸ However, Douglas et al. (2003) suggested that the construct was unexpectedly not uni-dimensional but displayed two factors with an eigenvalue > 1. One represented the core items with respect to the definition of consumer ethnocentrism and the other included the items related to the purchase of foreign items when no domestic alternative is available. Douglas and Nijssen (2003) conducted exhaustive analysis regarding convergent and divergent validity: the low correlation of both criterion and control variables with the second factor (domestic alternative availability) further confirmed a distinct element relating to non-availability of domestic products

⁹ Items 2,4,5,6,7,8,11,13,16,17 of the original 17 item CETSCALE

¹⁰ All these scales have been previously developed by Parameswaran and Yaprak (1987)

¹¹ The relevant product categories were cars, TV-sets and similarly products in general from each of the four countries: US, Germany, France, and Japan

¹² The principal components factor analysis suggested that the construct was surprisingly not uni- dimensional but displayed 2 factors with an eigenvalue > 1, one representing the core items and the other included the items related to the purchase of foreign items when no domestic alternative was available

¹³ Internationalism is defined as a positive feeling for other nations and their people, concern about nation's welfare, empathy for the people of other nations (Kosterman & Feshbach 1989)

¹⁴ Domestic country bias is defined as bias against foreign products and in favour of domestic ones,manifest in product perception and buying intentions. (Balabanis & Diamantopoulos 2004). It is furthermore retained that country-of-origin of a product is an important determinant of consumer bias against it

¹⁵ Items 2, 4, 5, 6, 7, 8, 11, 13, 16, 17 of the original CETSCALE (Shimp & Sharma 1987);

¹⁶ The researchers claimed that the scale reliabilities ranged from 0.72 to 0.99 for the scales after purification and attainment of uni-dimensionality

¹⁷ H1) ethnocentric consumers exhibit purchase behaviour in support of the domestic economy (significant), H2) higher evaluation of domestic products leads to purchase behaviour in support of the domestic economy as opposed to consumers with a less favourable evaluation of products (significant), H3) direct, positive, significant relationship between consumers' evaluation of domestic products and their consumer ethnocentrism (significant), H4) consumer cosmopolitanism is negatively related to consumer ethnocentrism (significant), H6) consumer national identity is positively related to consumer evaluation of domestic-made products relative to foreign offerings (not significant). Their results once more confirmed that consumer ethnocentrism is positively linked to the evaluation of domestic products and similarly is significantly positively linked to domestic purchase behaviour. The authors further claim a negative relationship between consumer ethnocentrism and consumer cosmopolitanism

¹⁸ Riefler and Diamantopoulos (2005) provided an excellent review of the hitherto published studies on cosmopolitanism and consumer cosmopolitanism