Kapittel 17 Does country-of-origin image and ethnocentrism positively influence the home market brand equity for the brand Dybvik ?

The study analyses the effects of consumers’ country-of-origin image perception (COOI) and ethnocentrism on brand equity for the bachalau brand Dybvik. A survey of 333 respondents is analysed by using PLS-SEM in order to answer the following research question: Does COOI and ethnocentrism positively influence the home market brand equity for the brand Dybvik? Both variables positively effects the perceived brand equity, but in different degree on the various underlying dimensions of


INTRODUCTION
Brand building has become a priority for many organizations with the presumption that it yields brand equity and long-term financial advantages (Keller 2013;Yasin, Noor & Mohamad 2007;Yoo & Donthu 2001).To manage this brand building process managers must develop a thorough understanding of the formative factors of brand equity.In this paper brand equity is conceptualized in accordance with Aaker (1991) and Keller (2001), using a consumer or marketing perspective as opposed to a financial one.Thus brand equity is defined as «the value consumers associate with a brand, as reflected in the dimensions of brand awareness, brand associations, perceived quality and brand loyalty» (Aaker 1991, p. 15).
The purpose of this paper is to analyse the effects of COO and CE on brand equity for the bachalau brand Dybvik.This brand is only available at the Norwegian market.Bacalhau, as it is known in countries such as Portugal and Brazil, is salted and dried fish mainly produced from cod or saithe, but can also be produced from ling, tusk and haddock.A Norwegian common term for this product is «klippfisk».The following research question is addressed: Does country-of-origin image and ethnocentrism positively influence the home market brand equity for the brand Dybvik?Eight hypotheses are underpinned and tested.As far as we know, no studies has hitherto addressed this context (home market and bachalau) and included both COO and CE as drivers of brand equity.Thus this paper contributes to increase the knowledge in this topic area.Regarding food and COO it should be mentioned that the interest of country-of-origin labelling (COOL) is on the increase, e.g. in the US where COOL is mandatory for several commodities such as wild and farm-raised fish and shellfish.A number of US consumer groups have voiced their support for COOL, stating that consumers deserve to know the origin of their food (Lewis and Grebitus, 2016).Thus COO is closely linked to food safety, at least for some food products.
The remainder of the paper is organized as follows.The next section addresses the theoretical framework of the study as well as hypotheses.Then, the context, the data, and the research methodology are briefly discussed, followed by a presentation of the results.The paper ends with a discussion of the findings and their implications for managers, presents some limitations and suggestions for further research, and offers a conclusion.

LITERATURE BRAND
Brands can be described as the main form of competitive positioning and differentiation tool in the business-to-consumer marketing context (Lindgreen, Beverland & Farrelly 2010).A brand is differentiated from other products designed to satisfy the same need, and symbolizes the essence of the customers' perceptions of a firm name, a logo, a symbol, an identity or a trademark.Further, it signals to the customers and the producers the source of a certain product, implying protection from competitors that would attempt to provide similar or identical products.Brands have been crucial for building relationships with consumers, thus assuring longterm business success for decades (Tuškej, Golob & Podnar 2013).The brand name, and what it represents, may be one of the most important assets for a firm.

BRAND EQUITY
Brand equity has proven to be a primary source of competitive advantage and future earnings, implying that brand equity may be perceived as representing intellectual capital of the brand builder.The concept denoted «Brand equity» was introduced in the marketing literature in the 1980s (Rajh 2005), and «refers to the incremental utility or value added to a product by its brand name» (Yoo & Donthu 2001, p. 1).One of the earliest and most frequently cited definitions of brand equity stems from Aaker (1991).He defined brand equity as «the value consum-ers associate with a brand, as reflected in the dimensions of brand awareness, brand associations, perceived quality and brand loyalty» (Aaker 1991, p. 15).He described brand equity as a set of assets (or liabilities), and found brand awareness, brand associations, perceived quality and brand loyalty to be its four most important dimensions from a consumer perspective (Pappu et al. 2006).In this paper brand equity is conceptualized in accordance with Aaker (1991).

Brand awareness
Brand awareness is described as «the ability of a potential buyer to recognize or recall that a brand is a member of a certain product category» (Aaker 1991, p. 61).Brand awareness can be related to four levels: «unaware of brand», «brand recognition», «brand recall» and «top-of-mind brand».The first level applies to consumers that are unaware of a brand.«Brand recognition» refers to the consumer's ability to confirm past exposure to a certain brand when given the brand as a cue («aided recall»).The third level, «brand recall», refers to the degree a consumer can retrieve a certain brand when given the product category, the needs fulfilled by the category, or some other form of cue.Finally, the first brand name the respondent names could be described as «top-of-mind awareness», which means that this brand is ahead of other brands in a consumer's mind.

Brand associations
Brand association is defined as «anything linked in memory to a brand» (Aaker 1991, p. 109).According to Aaker (1996), there are three main categories with respect to brand associations: the perceived (customer) value offered by a product, the image (personality) of the product (brand) and the organizational reputation/ associations.A number of associations exist.These associations can provide value both to the firm and to its customers in different ways such as to (1) help process/ retrieve information, (2) differentiate/position the brand, (3) generate a reason to buy, (4) create positive attitudes/feelings, and (5) provide a basis for extensions.

Perceived quality
Perceived quality has been used interchangeably with the term brand quality (Zeithaml 1988).In this paper the term is defined as the «customer's perception of the overall quality or superiority of a product or service with respect to its intended purpose relative to alternatives» (Aaker 1991, p. 85).Thus perceived quality is considered as the consumer's subjective assessment of a certain product.If the perception of brand quality is high, it can lead customers to select a particular brand over another competing brand, which eventually will lead to an increase in brand equity (Kim & Hyun 2011;Low & Lamb 2000).Oliver (1997) relates consumer loyalty to a four-stage model consisting of cognitive, affective, conative and action loyalty.Dick & Basu (1994) perceive loyalty as being based on two interrelated components: relative attitude and repeat patronage, where the former is related to cognitive, affective and conative antecedents.Aaker identifies brand loyalty as a behavioral factor.That is, Brand loyalty, long a central construct in marketing, is a measure of the attachment that a customer has to a brand.It reflects how likely a customer will be to switch to another brand, especially when that brand makes a change, either in price or in product features (Aaker 1991, p. 39).

Brand loyalty
Brand loyalty is considered to be qualitatively different from the other dimensions included in the model, as loyalty cannot exist without some previous purchase or use experience.Schooler (1965) is the first to identify the importance of country-of-origin (COO) as a cue in consumer choice behaviour in the international business literature.Nagashima (1970) provides the earliest definition of COO which relates to the image, reputation, and stereotype that businessmen and consumers attach to products of a specific country.He defines COO image as the sum of impressions and associations held about a specific country.Brands from countries that hold a favourable image generally can benefit from already accepted brands in comparison to those from countries with a less favourable image.For example, a consumer will increase his/her purchase intention when the perception of a country's image is positive, because he/ she then will have a high quality perception and overall positive evaluation to a product manufactured in that country (Agrawal & Kamakura 1999;Roth & Romeo 1992).

COUNTRY-OF-ORIGIN IMAGE
The literature contains different definitions of COO such as where the product is made (Nagashima 1970), the country of the company's headquarter (Johansson, Douglas & Nonaka 1985), the country of manufacture or assembly (Han & Terpstra 1988), or the country of design, parts, and assembly (Insch & McBride 1998).This study uses the country of manufacture, implying the use of the «made in» definition.
The COO effect describes the situation in which a consumer's judgement is altered owing to an association between a product, service, or brand and a place.Both Aaker and Keller argue that COO effects can affect a brand's equity by generating associations for the brand implying that consumers' attitudes may be influenced positively (Pappu et al. 2006).COO has been found to exert a particular potent effect on consumer evaluation in situations where there is a strong link between a country and a particular product category (Andéhn, Nordin & Nilsson 2016).This is the situation for Norwegian salted and dried cod (bachalau).
Given that brand equity is conceptualized as a four-dimensional construct consisting of brand awareness, brand associations, perceived quality and brand loyalty, the following hypotheses are offered (Chen et al. 2011;Pappu et al. 2006): H1: Country-of-origin image (COOI) positively influences brand awareness.H2: COOI positively influences brand associations.H3: COOI positively influences perceived quality.H4: COOI positively influences brand loyalty.
Figure 17.1 presents the theoretical model of the study inclusive the supposed links (hypotheses) between the concepts of the model.The hypotheses between COOI and the four brand equity dimensions are addressed above.The hypotheses between ethnocentrism and the four brand equity dimensions are addressed in the next part of the paper.In addition a dummy variable «Recent consumer experience» is included in the model, cf. the discussion above underscoring that loyalty is considered to be qualitatively different from the other dimensions included in the model and cannot exist without some previous purchase or use experience.CONSUMER ETHNOCENTRISM Shimp and Sharma (1987) coin the term «consumer ethnocentrism» (CE) and conducted the first study where they also describe the concept which has its roots in sociology: From the perspective of ethnocentric consumers, purchasing imported products is wrong because, in their minds, it hurts the domestic economy, causes loss of jobs, and is plainly unpatriotic; … To nonethnocentric customers, however, foreign products are objects to be evaluated on their own merits without consideration for where they are made … (Shimp & Sharma 1987, p. 280).
Thus, CE is the phenomenon of a preference for domestic products over international products.CE leads to a nationalistic evaluation of foreign products and services (de Ruyter et al. 1998).CE is found high in countries where consumers feel that doing so will help local products to flourish and develop the home country markets (Kashif et al. 2015).Ethnocentric consumers are reluctant to buy products and services provided by foreign companies because of a sense of loyalty towards their home country.Highly ethnocentric consumers are willing to purchase a brand from their own country even when the foreign brands are objectively superior (Suppelen & Rittenburg 2001).Balabanis & Diamantopoulos (2004) studied the domestic country bias based on effects of COO and CE.Regarding CE their conclusion is: As expected, CE was found to be positively related with preferences (here, British) products and negatively related with preferences for foreign products.However, CE was a much more consistent predictor of the former than of the latter; in other words the CE construct appears to more capable of explaining consumers' positive bias toward home products rather than negative bias against foreign products (Balabanis & Diamantopoulos 2004, p. 91).
Analogous to the concept denoted COOI, CE is an attitude that is supposed to influence all the four dimensions of the brand equity model of the Norwegian bachalau brand denoted Dybvik.This is supported by the literature referred to above.Thus, the following hypotheses are offered: H5: Consumer ethnocentrism (CE) positively influences brand awareness.H6: CE positively influences brand associations.H7: CE positively influences perceived quality.H8: CE positively influences brand loyalty.

THE RELATIONSHIP BETWEEN COUNTRY OF ORIGIN IMAGE AND CONSUMER ETHNOCENTRISM
Some studies address both COO and CE (Balabanis & Diamantopoulos 2004;Chryssochoidis et al. 2007;Jiménez & Martín 2010).Few studies address food (Kashif et al. 2015;Schnettler et al. 2013) or fish (Aure et al. 2017) and only a few the native country (Lewis & Grebitus 2016;Vabø et al. 2017).However, few of the studies address the relationship between COO and CE.Schnettler et al. 2013).Thus, there is a widespread agreement that the concepts denoted consumer ethnocentrism (CE) and country-of-origin (COO) are linked, however, the form of the relationship is not clear.

CONTEXT
The Norwegian klippfisk industry has traditionally had a strong position in international markets.The main market has been and still is Portugal, followed by Brazil.In 2016, the total export value of Norwegian klippfisk was NOK 3.7 billion, which was down NOK 260 million from the peak year of 2015.The home market of klippfisk is relatively small compared to the export market, but is expected to increase.This industry is to a large extent geographically concentrated in Møre and Romsdal county in the North-Western part of Norway, and this area is likely to host what may be the only salted and dried fish cluster in the world (Bjørndal et al. 2017).The raw material used by this industry includes both fresh and frozen fish of the different types, delivered primarily by Norwegian, Icelandic and Russian fishermen.After salting, the fish will undergo a drying process so as to prepare the final product.Producers and exporters from this part of the country contribute considerably to the value added of the salted and the salted and dried cod industry (Bjørndal et al. 2015).
The company Jakob & Johan Dybvik AS was established in 1923.It is located at Fiskarstrand, a small village right outside Aalesund.The company is familyowned, and the knowledge and craft of producing klippfisk has been passed down three generations so far.The brand «Dybvik» was launched in 2009.At that time the brand building work had been going on for more than a decade (product development, assortment, contacts, etc.).Jakob & Johan Dybvik AS has hitherto primarily been focusing on the Norwegian market.

DATA, METHODOLOGY AND DESCRIPTIVE STATISTICS
Table 17.1 presents the descriptive statistics of the 25 items comprising the six concepts included in this survey.Four of the concepts are measured by four items each.Here we just refer to the concepts and where the items are taken from: «Brand awareness» (BAW) (Kim & Kim 2005;Netemeyer et al. 2004;Yoo & Donthu 2001); «Brand association» (BASS) (Aaker 1996;Buil et al. 2008;Pappu et al. 2005;2006); «Perceived quality» (PQUAL) (Pappu et al. 2005;2006); «Brand loyalty» (BLOY) (Kim & Kim 2005;Yoo & Donthu 2001).The concept «Country-of-origin image» (COOI) (Lin & Chen 2006;Martin & Eroglu 1993) is measured by six items and the concept «Ethnocentrism» (ETHNO) (Herche 1992;Shimp and Sharma 1987) is measured by three items.Each of the items was measured on a seven point Likert scale where the respondents were asked to grade the answer of a given statement from «strongly disagree» to «strongly agree», etc.The total sample comprises 333 valid respondents from Sunnmøre (ten locations).The model is estimated by a partial least square structural equations modeling (PLS-SEM) approach based on SmartPLS (Hair et al. 2014b).The objective of PLS-SEM is to maximize the explained variance of the endogenous latent variables of a model, and simultaneously estimate both a measurement model and a structural model.Because PLS-SEM is a non-parametric method that can handle non-normal data and small samples, in addition to its flexibility regarding the number of indicators (items) per construct, it is often the preferred method compared to the alternative covariance-based SEM (Hair, Black, Babin & Andersson 2014a).
According to our theoretical model illustrated in figure 1, the endogenous variables are the four constructs: 1) BAW, 2) BASS, 3) PQUAL and 4) BLOY, whereas the exogenous variables are the two constructs: 1) COOI and 2) ETHNO.In addition, a dummy variable (REXP) is introduced to control for recent con- sumer experiences with the Dybvik brand.This dummy has the value 1 when the consumer has recently tasted the Dybvik clip fish product and 0 otherwise.Table 17.2 describes the measurement result of the six concepts in terms of standardized loadings and measures of convergent validity (Chronbach's alpha and average variance extracted).The factor loadings vary from 0.73 to 0.94.The average variances extracted (AVEs) are high for all the six concepts (varying from 0.62 to 0.87), and well above the recommended level of 0.50.Cronbach's alpha varies from 0.82 to 0.95, which are higher than the recommended lowest level of 0.70.Thus, the findings indicate that all concepts are measured in a reliable way, implying convergent validity for all the six variables.Based on these findings six new constructs are established as «summated scales».Table 17.3 presents descriptive statistics and the square root of AVE (diagonal) for the six new constructs as well as the correlation coefficients between them.This study examines discriminant validity of the six constructs by comparing the correlation coefficient between any two constructs with their respective square rooted AVEs.If their square rooted AVEs are larger than their correlation coefficient, the variables are truly distinct constructs.There may be a problem with respect to «Brand association» and «Perceived quality».The correlation coefficient between the two constructs is 0.908, and the square rooted AVEs regarding the two concepts are as follows: 0.930 («Perceived quality») and 0.898 («Brand association»).The discriminant validity test between these two concepts is thus inconclusive in our data.However, the measures show high reliability and are in accordance with both theory and earlier empirical research.

RESULTS
The PLS model is estimated in three steps.First (in model 1), only the main theoretical variables COOI and ETHNO variables are included as explanatory varia-  # Based on bootstrapping with 1000 samples 1.The effects of interactions between ETHNO and REXP on the four dependent variables were also tested.However, they turned out to be far from significant and are thus omitted from model 3.
In model 1, all the links between the two main independent variables (COOI and ETHNO) and the four dependent variables (BAW, BASS, BLOY and PQUAL) are positive and significant.Variance explained (R 2 ) of the four different dependent variables varies from 0.184 (BAW) to 0.261 (PQUAL).In model 2 we see that the inclusion of the control variable (REXP) increases the variances explained dramatically (increases in the range of 22 to 34 percentage points).We also see that all the links from the basic model (model 1) are still significant, but the magnitudes are in general somewhat lower.It is obvious that the dummy variable for recent consumer experiences (REXP) picks up substantial parts of the explained variances, and thus the dummy variable has a much larger effect on the four dependent variable than the two basic explanatory variables.In model 3 the interactions of REXP and COOI are included in addition to the direct effects of REXP.
All the regression coefficients of the main independent variables and the REXP dummy coefficients are positive and significant, and quite similar to the coefficients in model 2. Of the four interaction terms, three of them are positive and significant.The only insignificant interaction term is between REXP•COOI and BAW.This means that for consumers with recent experience with the Dybvik brand the effects from COOI on BASS, BLOY and PQUAL are all higher than the similar effects for consumers with no recent experience with the brand.Interactions between ethnocentrism and recent experience with the brand have no strong theoretical foundation.Ethnocentrism is a more static phenomenon than the more experience based country-of-origin image (de Ruyter et al. 1998;Shimp & Sharma 1987;Vabø et al. 2017).This is also confirmed in our data set as all the interaction terms between the dummy variable 'recent brand experience' and 'ethnocentrism' were insignificant and thus omitted from the final model 3.This result has also some methodological consequences.Normally, when moderation effects of a categorical variable is expected, a multi-group SEM is recommended (Hair, Hult, Ringle & Sarstedt 2014b).A multi-group SEM is, however equivalent with a SEM model including all possible interaction terms, i.e. in our analysis also including all the interaction terms between recent brand experience and ethnocentrism.As long as these interactions have no theoretical underpinning and are insignificant, a multi-group SEM will not be the appropriate method to use.

DISCUSSION AND CONCLUSION
The purpose of this paper is to analyse the effects of country of origin and ethnocentrism on brand equity for the bachalau brand Dybvik.This brand was launched by Jakob & Johan Dybvik AS in 2009 and is only available in Norway.The fol-lowing research question is addressed: Does country-of-origin image and ethnocentrism positively influence the home market brand equity for the brand Dybvik? Brand equity is conceptualized as reflected in the dimensions of brand awareness, brand associations, perceived quality and brand loyalty.Eight hypotheses are underpinned and tested.
The first four hypotheses address the relationships between country-of-origin image and the four brand equity dimensions (brand awareness, brand associations, perceived quality and brand loyalty) and the remaining four hypotheses the relationships between ethnocentrism and the same brand equity dimensions.Three models are addressed.In model 1, only the main theoretical variables (country-oforigin image (COOI) and ethnocentrism (ETHNO)) are included as explanatory variables for brand awareness (BAW), brand associations (BASS), perceived quality (PQUAL) and brand loyalty (BLOY).Model 2 is expanded by including a control variable taking account of the consumers' recent experience with the Dybvik brand (REXP).Model 3 is further expanded by allowing interaction effects between the explanatory variables of model 2, however, only the interaction effect between consumers' recent experience with the Dybvik brand (REXP) and country-of-origin image (COOI) was significant and thus included in model 3.
Regarding the relationships between COOI and the four brand equity dimensions, the following four hypotheses were tested: (H1) COOI positively influences brand awareness, (H2) COOI positively influences brand associations, (H3) COOI positively influences perceived quality, and (H4) COOI positively influences brand loyalty.Regarding the relationships between consumer ethnocentrism (CE) and the four brand equity dimensions the following four hypotheses were tested: (H5) CE positively influences brand awareness, (H6) CE positively influences brand associations, (H7) CE positively influences perceived quality, and (H8) CE positively influences brand loyalty.In all the three models, the findings (strongly) support all the eight hypotheses.All coefficient estimates are positive and significant.Thus, it may be asserted that both COOI and CE positively influences the four brand equity dimensions and thus implicitly brand equity.The conclusion seems to be that the higher the level of country-of-origin image and ethnocentrism, the higher the level of brand equity.Thus, both the explanatory variables are drivers of brand equity.
With respect to variance explanations, the three models vary.In model 1 the variance explanations of the four dependent variables (brand awareness, brand associations, perceived quality and brand loyalty) vary from 18 to 26 percent.In model 2 the variance explanations are much higher, i.e. from 42 to 54 per cent.Thus, the inclusion of the variable representing recent consumer experience (REXP) increases the variance explanation of model 2 compared with model 1.Inclusion of the interaction effect between COOI and recent consumer experience (REXP•COOI) in model 3 results in an additional increase of the variance explanation of the dependent variables.This is attributable to the effects of the additional variables included in model 2 and model 3.This is also easily seen in table 4 by looking at the coefficient estimates of recent consumer experience (REXP) and the interaction effect between COOI and recent consumer experience (REXP•COOI), particularly the first one.Thus recent experience with the brand is an important driver of brand equity.It should be underscored that the inclusion of this variable strengthened the findings regarding the hypotheses as long as all the coefficient estimates are significant when controlling for the additional variables included in model 2 and model 3.The inclusion of REXP also results in a more representative sample for prediction.
Of the four interaction terms, three of them are positive and significant.The only insignificant interaction term is between REXP•COOI and BAW.This means that for consumers with recent experience with the Dybvik brand the effects from country-of-origin on brand associations, brand loyalty and perceived quality are all higher than the similar effects for consumers with no recent experience with the brand.
Another interesting result from model 3 is the different effects of COOI and CE on the various brand assets.COOI has a much stronger effect on brand associations (0.258) than CE (0.133), whereas CE has a much stronger effect on brand loyalty (0.258) than COOI (0.111).The COOI effect on brand loyalty will, however, be strengthened when consumers acquire experience with the brand.This gives some interesting managerial implications.If brand equity is weak on brand associations, the best way to increase brand equity is to improve the COOI of the consumers.If, on the other hand, brand equity is weak on brand loyalty, the best way to increase brand equity is to target the ethnocentric consumers.
Further research should include respondents from other regions: are there differences among Norwegian regions with respect to ethnocentrism and COOI?In addition, other products, contexts and related topic areas should be addressed.For instance, country-of-origin labelling (COOL) is on the increase, e.g. in the US where COOL is mandatory for several commodities such as wild and farm-raised fish and shellfish.An interesting problem area seems to be the relationships between country-of-origin image (COOI), country-of-origin labelling (COOL) and consumers' perception of food safety.
Building brand equity has become a priority for many organizations.A strong brand equity is supposed to result in competitive advantages and future earnings.
To manage this brand building process managers must develop a thorough understanding of the formative factors of brand equity.This study offers country-of-origin image and ethnocentrism as formative factors of brand equity.The context is the bachalau brand Dybvik that is only available at the Norwegian market.The findings indicate that both country-of-origin image and ethnocentrism are important drivers of brand equity.Even if these findings are based on a survey for one bachalau brand, the results should probably be of interest for marketers of other food brands owing to the fact that there are significant links between brand equity and business performance.

FIGURE
FIGURE 17.1 The theoretical model of the study (Jiménez & Martín 2010;Lewis & Grebitus 2016;domestic country bias based on effects ofCOO and  CE.Regarding links between consumer CE and COO their conclusion is: «In short, the value of the CE construct … appears to be variable depending on both the product category under consideration and the specific COO involved»(Balabanis & Diamantopoulos 2004, p. 91).The conclusions of other contributors are much the same(Jiménez & Martín 2010;Lewis & Grebitus 2016;

TABLE 17
TABLE 17.3 The six constructs: Descriptive statistics and discriminant validity (Square root of average variance extracted (diagonal) and correlations).
BASS, BLOY and PQUAL, respectively.Next, the model is expanded (model 2) by including a control variable taking account of the consumers' recent experience with the Dybvik clip fish (REXP).Next, the model is further expanded by allowing interaction effects between REXP and COOI.The results are shown in table 17.4 1 .