The Impact of Country of Origin on Brand Equity: An Analysis of The Wine Sector

The purpose of this study is to explore the country of origin ́s effects on brand equity dimensions. This research selected wine as the product category and data were collected from Portuguese and Canadian consumers. Our conceptual framework incorporates the influence of country of origin on brand equity dimensions, composed by brand loyalty, brand associations, brand awareness and perceived quality, as well as the brand equity subsequent effect on purchase intention. The hypotheses were tested using Partial Least Squares Structural Equation Modelling (PLS-SEM). The results of the Portuguese sample indicate that the country of origin affects positively all the brand equity dimensions. The Canadian sample results show that country of origin affects brand loyalty and perceived quality, but there is no significant effect on brand associations and brand awareness dimensions.


INTRODUCTION
Consumers' concerns during the purchasing process comprise, not only product quality and price, but also other factors such as, the product's country of origin. The home country effect is a persistent concern in international marketing (Koschate-Fischer et al., 2012). Countries recognized as good producers in a product category generally leverage the acceptance of their products compared to countries less recognized in that product category (Yasin et al., 2007).
Country of origin can act as an indicator of quality, exerting an important impact on consumer's purchase intention (Lin and Chen, 2006) and consequently, on the consumer's perception of brand equity (Rezvani et al., 2012). This study intends to analyze the impact of country of origin on brand equity, placing the focus of the analysis on the Portuguese and Canadian markets.
Different types of conceptual categorizations have been developed to classify wine producing countries, but the classification that separates the geography of the wine industry into two "worlds", has been widely applied (Rodrigues et al., 2020). According to this taxonomy, Portugal is classified as an "old world" wine producing country, while Canada is a "new world" wine producing country (Johnson and Robinson, 2013). Remaud and Couderc (2006) consider that "old world" wine producing countries have historically emphasized the origin of grapes and promoted their specified regions' designations, while "new world" has mainly concentrated on supporting labeling through strong proprietary brands and grape variety. The wine producer country classification "old world" vs. "new world" can be combined with the categorization into "traditional producer" vs. "nontraditional producer" (Thorpe, 2009). While the categories "old world" vs. "new world" consider the country's history as a wine producer, the classification "traditional producer" vs. "non-traditional producer" refers to the country's relevance in world wine exports. Therefore, a non-traditional exporter indicates a country with a reduced presence as an exporter in the international wine market. According to these categorizations, the differences between the two countries under investigation are evident, since while Portugal is classified as an old world traditional producing wine country, Canada is considered as a new world non traditional producing wine country (Rodrigues et al., 2020).
Portugal is the eleventh major wine producer in the world, with a total volume of wine production of approximately 6.1 million hectoliters (OIV, 2019). Between 2014 and 2018, Portuguese wine production increased 29% in volume. Portugal is the fourth major per capita wine consumer in the world, with a per capita consumption of 51.3 liters in 2018, and the eleventh major wine consumption market in the world, reaching a total of 5.5 million hectoliters in that year (OIV, 2019). Moreover, Portugal is the ninth world top exporter of wine, exporting around three million hectoliters of wine in 2019 (Statista, 2020). The export volumes of Portuguese wines have been relatively stable from 2011 to 2019. Additionally, ProWein business report (ProWein, 2018) considered Portugal as one of the wine origins with stronger demand from international marketers.
Canada has a wine consumption per capita of 16.9 liters (Statista, 2020). The per capita wine consumption in the Canadian market increased 14.2% between 2010 and 2018. The wine consumption in the Canadian market has been showing a growing trend over the past few years, with an increase of 32.7% between 2011 and 2019, in terms of sales. In 2019, Canadian wine sales reached approximately 2.28 billion Canadian dollars, while imported wine sales achieved a total of 5.18 billion Canadian dollars. Canada is the twelfth major wine consumption market in the world, reaching a total of 4.9 million hectoliters in 2018 (OIV, 2019). Moreover, Canada is the sixth world top importer of wine, with approximately a total of 4.2 million hectoliters (OIV, 2019). Between 2014 and 2018, Canadian wine imports demonstrated a growth of 10% in volume. In the ProWein business report (ProWein, 2018), Canada was considered one of the wine markets more attractive and with lower risk for wine producers.
Based on a sample of Portuguese and Canadian consumers, the current study intends to investigate the influence of the wine's country of origin on the different brand equity dimensions and purchase intention. The next section contains the literature review, followed by hypotheses development. Next, the methodology and results sections are presented. This study is concluded with the discussion, managerial implications, and suggestions for future research.

Country of origin
Country of origin can be defined as the country to which the consumer associates a product or brand as its source, regardless of where the product is produced (Herz and Diamantopolous, 2013). The country of origin's effect is mentioned by several authors (e.g. Sauer et al., 1991;Suh et al., 2015) as the influence exerted on consumer evaluation of a product or service, depending on the stereotypes formed in relation to the country of origin.
Country of origin can be considered as a product attribute (Cordell, 1992;Hong et al., 1989). Consumers often develop stereotypes about the products of a particular country and their respective attributes (Yasin et al., 2007). The greater the contact and knowledge about a country or its products, the more objective can be the consumer's perception regarding the country of origin effect (Lee and Lee, 2009). Country of origin may function as an indicator of product quality or as an appeal to references that consumers may have developed regarding products of a particular origin (Insch and Floreck, 2009). Consequently, a positive image of the country can be used to highlight the relationship between the product and its origin. Furthermore, a corporate brand's country of origin can influence the reputation signals' effectiveness (Cowan and Guzman, 2020). The Impact of Country of Origin on Brand Equity: An Analysis of The Wine Sector The influence of country of origin on consumers' product evaluations can occur on three levels, namely, cognitive, affective and normative (Martinez-Carrasco, Brugarolas and Martinez-Poveda, 2005). Regarding the cognitive perspective, country of origin can be used as a signal for quality attributes. The affective level is activated by emotional benefits related to national pride and social status. The normative perspective is associated with personal and social norms that consumers possess related to the country of origin. For instance, the purchase of national products can be perceived as the correct way of conduct to support the national economy.
In the wine sector, communication of its origin is mandatory, both in Canada (Food andDrugs Act, 1985 andConsumer Packaging andLabeling Act, 1985) and in the European Union (Protected Designation of Origin, 2011), thus increasing the relevance of the country of origin's effect in this product category (Thøgersen et al., 2017).
The positive impact of the country of origin occurs in many industries. However, previous research has defended that country of origin effects may differ between product categories (Trinh, Corsi, and Lockshin, 2019). For instance, country of origin seems to exert weaker effects for low-involvement products, such as wheat, potatoes, or butter (Magnusson, Haas, and Zhao, 2008). On the contrary, country of origin effects tend to be stronger for high-involvement products, such as wine, since consumers invest more effort into collecting information about those products (Li, Leung, Wyer, 1993).
Country of origin is used as crucial element in promotion campaigns in the wine industry (Trinh et al., 2019). Previous studies have shown that the country of origin of wines can have important implications for both export and marketing activities of the wine industry. For instance, Rodrigues et al. (2020) have shown that the country of origin of wines is extremely relevant to the wine traders' mental representation. Tzimitra- Kalogianni et al. (1999) concluded that the wine origin is one of the most relevant wine attributes in the Greek market, together with label, aroma, taste, and clarity. Keown and Casey (1995) evidenced that country of origin is the most relevant criterion for wine selection in Northern Ireland. Steiner (2000) highlighted the importance of origin in the selection of French wines in the UK wine market. Therefore, in the wine industry, country of origin can be considered a crucial element in the consumers' wine choice.

Ethnocentrism
The concept of ethnocentrism has been mentioned in international business for about half a century (Michailo-va et al., 2017). Ethnocentrism can be defined as: "a view of things in which one's own group is the center of everything. Each group nourishes its own pride and vanity, boasts itself superior, exalts its own divinities, and looks with contempt on outsiders" (Sumner, 1906: 13).
Ethnocentric consumer trends affect the emotional dimension related to the process of buying foreign products and creates the assumption that this choice somehow threatens domestic industry and national security (Herche, 1992). Thus, ethnocentric consumers believe that buying foreign-produced products is morally wrong and does not contribute to the local economy (El Banna et al., 2018), therefore opting to buy domestic products and disregarding foreign products. With increasing globalization, the concept of ethnocentrism becomes relevant for understanding the moral concern resulting from the consumption of domestic versus foreign goods (Siamagka et al., 2015).
In the wine industry, consumer ethnocentrism has been considered as being capable to influence both the perceived quality and the purchase intentions of regional wines (García-Gallego, Chamorro-Mera, and García-Galán, 2015). These results confirm that communication campaigns that promote a wine region's image are valuable, not only as an external promotion tool, but also as an encouragement for domestic consumption, particularly by consumers with stronger ethnocentrism levels. Martinez-Carrasco et al. (2005) found that consumer ethnocentrism creates strong preferences for local wines. Therefore, consumer ethnocentrism can act as a crucial consumption motivation during the process of wine selection. Ethnocentrism creates in the individual a sense of identity and stimulates the understanding that national symbols and values are objects of attachment and pride (Le Vine and Campbell, 1972), motivating the purchase of domestic products and creating a negative attitude towards foreign products (Verlegh and Steenkamp, 1999).
Nowadays, consumers' choices encounter a large range of domestic and foreign products. Ethnocentric tendencies constitute one of the factors that influence the purchase decision between domestic and foreign products. In the wine industry, consumer ethnocentrism can exert a strong and positive influence on attitudes related to domestic wine purchase (Tomić Maksan, Kovačić, and Cerjak, 2019).

Brand equity
Brand equity can be defined as a "set of brand assets and liabilities linked to a brand, its name and symbol, that add to or subtract from the value provided by a prod-uct or service to a firm and/or to that firm's customers" (Aaker, 1991: 27). Consequently, brand equity translates the value of a product, a service, or a corporate brand (Kim, Choe, and Petrick, 2018). A brand is considered to benefit from a positive customer-based brand equity effect when consumers respond more favorably to an element of the brand marketing mix, compared to their responses to a similar marketing mix element from an unnamed or a fictitiously named version of the product or service (Keller, 1993). Brand equity is of critical importance since it conditions customer lifetime value and leverages financial market outcomes (Heitmann, et al. 2020).
The assets and liabilities that generate brand equity can be grouped into the following dimensions: brand loyalty, brand awareness, perceived quality, brand associations and other proprietary brand assets (Aaker, 1991). These dimensions of brand equity have long been of crucial interest to business managers, marketing, and consumer behavior researchers (Chen, Su, and Lin, 2011;Kim et al., 2018). In the following sections, the main antecedents of customer-based brand equity (brand loyalty, brand awareness, perceived quality, brand associations) are explained in more detail.

Brand loyalty
Brand loyalty is often the core of brand equity (Aaker, 1991), as consumer loyalty reinforces the brand over competitors. This relationship is strengthened as brand loyalty acts as a barrier to entry for new competitors, since the cost of attracting new customers is higher than retaining existing consumers.
Brand loyalty is not a unanimous concept in the literature, but two key approaches to behavioral loyalty and attitudinal loyalty are highlighted (Roustasekehravani et al., 2015). Behavioral loyalty is based on consumer behavior. Thus, the consumer who systematically buys the same brand is considered loyal. The subjectivity and complexity of explaining why the behavior occurs, is pointed out as the main disadvantage of this approach, since the company is objectively unaware of the factors that determine the consumer's brand loyalty (Odin et al., 2001). On the other hand, attitudinal loyalty delimits and identifies the determining factors of loyal consumer attitudes. Therefore, research focuses on the consumer's psychological commitment to purchase, without necessarily considering whether the purchase takes place. Premium price can be perceived as a basic indicator of brand loyalty, as it indicates that consumers are willing to pay an extra percentage for that brand, although other brands with similar characteristics may be available at a lower price (Aaker, 1991).

Brand awareness
Brand awareness can be defined as the ability of consumers to recognize or recall that a brand belongs to a particular product category (Romaniuk et al., 2017). Brand awareness is strongly related to the strength of a brand's presence in the minds of consumers, resulting in their ability to identify the brand under various market conditions (Świtała et al., 2018). For Aaker (1991), brand awareness is created by repeatedly and memorably exposing brand elements such as the name, slogan, logo, or packaging to consumers. This exposure contributes to establish brand roots in consumers' memory and to strengthen brands' links with the product category.
The strengthening of brand awareness paves the way for a victorious brand, due to the distinction from other brands present in the market (Ahmed et al., 2017). For example, Constellation, one of the largest wine companies in the world, invests heavily in the acquisition of wine brands to create and increase brand awareness (Atkin et al., 2017). This strategy recognizes the important role that brand awareness plays in the distinction between products and in the consequent purchase, therefore consumers can buy a bottle of wine by the brand name, even when they have little knowledge about that particular wine (Foroudi, 2018).

Brand associations
Brand associations relate a memory to a brand (Aaker, 1991). Associations are starting points for buying decisions, as consumers rely on them to retain information emanating from brand actions, to generate a buying reaction and to create positive behaviors, allowing openness for brand diversification into other types of products. Thus, companies try to associate their brand with a certain attribute, recognized by the consumer, to make it difficult for new brands to enter the market (Akkucuk et al., 2016).
Brand associations are important for both businesses and consumers, as brand associations influence brand differentiation, brand positioning, brand extension, consumer information processing, consumer behavior, purchase intention and consumer satisfaction (Bawa and Saha, 2016).

Perceived quality
Perceived quality is the consumer's subjective assessment of the excellence or generic superiority of a product (Zeithaml, 1988). The degree of perceived quality increases with the long-term brand relationship experi-ence, as through this, consumers recognize brand differentiation and superiority. Perceived quality is subjective since it depends on the perception and discernment of the consumer involved. Saleem et al. (2015) characterize perceived quality as a psychological assessment of the product based on consumer perceptions. This assessment is based on product characteristics, whether intrinsic or extrinsic. The intrinsic attributes, which are related to the product itself, may be the aroma, the palate, or the color, in the case of wine. Extrinsic characteristics are related to properties that are not physically part of the product, such as packaging or region of origin (Sáenz-Navajas et al., 2016). The intrinsic determinants of wine quality are the most important for brand value creation, however, they are also the most difficult to control. Thus, the combined valuation of both characteristics is predominant (Danner et al., 2016).
Perceived quality is a particularly important factor in the wine product category, as consumers perceptions are quite heterogeneous and strongly influenced by their level of specialization or knowledge (Sáenz-Navajas et al., 2016). Therefore, perceived quality can be understood as a brand value proposition, which makes consumers endogenous to value creation (Liu et al., 2017).

HYPOTHESES
The country of origin can affect the purchase behavior (Dmitrovic, Vida, and Reardon, 2009). Consumers form multiple stereotyped national images regarding a product's country of origin (Ahmed and d'Astous, 2007;Laroche et al., 2005), and these perceptions influence their purchase behavior (García-Gallego, et al., 2015;Papadopoulos, and Heslop, 2002;Pharr, 2005). A country of origin with a strong presence generally has a positive impact on purchase intention, as it acts as an indicator of product quality (Aichner et al., 2017). Therefore, when a wine is thought to be originating from a country with high standards, a long tradition and or with strong wines brands, it will be considered superior and more appealing compared to a similar wine from a country not recognized as a wine producer. Certain brands adopt names that refer to nationalities recognized by the product category, in order to leverage their product based on the effect of the country of origin, such as the French brewer Brasserie Fischer that launched its tequila flavored beer by adopting a name, Desperados, and a Mexican image. Based on the effect of Portugal as a country of origin in the wine sector, the following hypotheses are proposed: H 1a : The country of origin Portugal relates positively with purchase intention for Portuguese consumers.
H 1b : The country of origin Portugal relates positively with purchase intention for Canadian consumers.
Brand equity has a positive and important impact on the preference for a brand. Hoeffler et al. (2003) suggest that strong brands get a greater preference from consumers because with a certain level of product knowledge, consumers buy the brands with the highest value (Aaker, 1991). Brand equity is reflected in the preference for the brand, and thus we can infer that the preference for the brand is reflected in purchase intention (Bougenvile and Ruswanti, 2017). Thus, the following hypotheses are proposed: H 2a : Brand equity relates positively with purchase intention for Portuguese consumers.
H 2b : Brand equity relates positively with purchase intention for Canadian consumers.
Consumers' national identities have been showing a growing impact on their consumption behavior (Dmitrovic et al, 2009;Quellet, 2007). Stronger competitive rivalries in domestic markets may stimulate nationalistic purposes in consumers' purchasing motivations (Shankarmahesh, 2006). Since consumers judge products taking into account the country with which they are associated, consumers may reveal, in specific conditions, a preference for a domestical alternative (Granzin and Olsen, 1998). Ethnocentric consumers tend to believe that it is unpatriotic to purchase foreign products and to consider national manufactured products as being superior (Shimp and Sharma, 1987). Consequently, consumer ethnocentrism strongly influences the construction of consumer attitudes towards domestic and foreign products (Kim and Pysarchik, 2000). In the current study, the ethnocentrism of Portuguese and Canadian consumers and their relationship with "Portugal" country of origin effect was analyzed, assuming that the greater the degree of Portuguese consumers' ethnocentrism, more favorable will be the associations that they will create in relation to their own country of origin. In turn, it is considered that the greater the degree of ethnocentrism of Canadian consumers the less favorable their associations will be in relation to the country of origin Portugal. Therefore, the following hypotheses are proposed: H 3a : Portuguese consumers' ethnocentrism relates positively with the country of origin Portugal.
H 3b : Canadian consumers' ethnocentrism relates negatively with the country of origin Portugal.
The country of origin can provide to consumers important quality connotations, consequently affecting the perceived quality of a product (Klein, Ettenson, and Morris, 1998;Verlegh and Steenkamp, 1999). The coun-try image influences consumers' product evaluations , regarding attributes such as the product quality and reliability (Laroche et al., 2005). Consequently, consumers make use of product-country images as a cue for inferring product quality (Haubl and Elrod 1999;Verlegh, Steenkamp, and Meulenberg, 2005). Consumers tend to have favorable perceptions regarding brand quality when the brand is known to come from a country strongly associated with a certain product category (Elliot and Cameron, 1994). Thus, it is expected that associations to a country, well recognized in a product category, are transferred to the perceived quality of brands originated from that country. Hence, the following hypotheses are proposed: H 4a : The country of origin Portugal relates positively with perceived quality for Portuguese consumers.
H 4b : The country of origin Portugal relates positively with perceived quality for Canadian consumers.
Country of origin can generate and emphasize secondary associations to a given brand, through connections to places, people, and moments (Pappu et al., 2006). Since associations to a country of origin are considered as secondary associations to the brand (Keller, 1993), it is assumed that the country of origin influences brand associations. Thus: H 5a : The country of origin Portugal relates positively with brand associations for Portuguese consumers.
H 5b : The country of origin Portugal relates positively with brand associations for Canadian consumers.
Consumers are known to associate the brand with its country of origin. Previous studies confirm a strong relationship between country of origin and brand awareness (Sanyal et al., 2011). In this sense, the awareness of a wine-producing country is transferred to wines from that country, thus influencing the consumer. Therefore, brands from the same country share a common base of awareness related to the country of origin (Pappu et al, 2007). Hence, the following hypotheses are proposed: H 6a : The country of origin Portugal relates positively with brand awareness for Portuguese consumers.
H 6b : The country of origin Portugal relates positively with brand awareness for Canadian consumers.
Country of origin can affect brand loyalty through its positive image and consumer familiarity with the country's products (Septyanti and Hananto, 2017). In parallel, Paswan et al. (2003) verified that consumers tend to be loyal to a country, just as they are loyal to brands. Therefore, consumer satisfaction with products from a particular country can influence consumer loyalty to that country's brands (Pappu et al., 2007). Thus: H 7a : The country of origin Portugal relates positively with brand loyalty for Portuguese consumers.
H 7b : The country of origin Portugal relates positively with brand loyalty for Canadian consumers.
Perceived quality can increase brand preference and induce consumers to choose a brand over another (Liu et al., 2017). Aaker (1991) also identifies perceived quality as a distinctive factor that offers the consumer a reason to buy or not a certain product. In this way, a growth in perceived quality can translate into an increase in brand equity. Therefore, the following hypotheses are proposed: H 8a : Perceived quality relates positively with brand equity for Portuguese consumers.
H 8b : Perceived quality relates positively with brand equity for Canadian consumers.
Brand associations are fundamental to differentiate the brand from its competitors (Aaker, 1991). Krishnan (1996) concluded that brands with strong brand equity tend to possess more positive brand associations than others with weak brand equity. Positive brand associations, influence consumers' choice, benefit the brand's image and enhance brand equity (Faircloth, Capella, and Alford, 2001;Pouromid and Iranzadeh, 2012;Sasmita and Suki, 2015;Yasin et al., 2007). In the context of the wine product category, associations with a brand may be represented by status or by a moment that makes that wine special, giving it an identity that translates into a positive association with that wine brand. Consequently, brand associations can benefit brand equity. Thus: H 9a : Brand associations relate positively with brand equity for Portuguese consumers.
H 9b : Brand associations relate positively with brand equity for Canadian consumers.
The greater the brand awareness the more likely the brand is to be considered in purchase situations (Yasin et al., 2007). Consumers usually prefer to buy brands that are familiar to them. As a result, increasing brand awareness can lead to an increase in brand equity (Pouromid and Iranzadeh, 2012;Keller, 2008). This leads to following hypotheses: H 10a : Brand awareness relates positively with brand equity for Portuguese consumers.
H 10b : Brand awareness relates positively with brand equity for Canadian consumers.
Consumers' brand loyalty reflects a repetitive purchasing behavior of the brand, which is positively associated to brand equity (Sasmita and Suki, 2015;Zhang, van Doorn, and Leeflang, 2014). Consequently, brand loyalty is considered as an important source of consumer-based brand equity (Aaker, 1991;Kim, Kim, and An, 2003). When consumers are loyal to a brand, even when brands with enhanced characteristics are available, it means that the brand has value for the consumer (Agrawal and Kamakura, 1999). Therefore, stronger lev-els of brand loyalty translate into higher levels of brand equity. Therefore: H 11a : Brand loyalty relates positively with brand equity for Portuguese consumers.
H 11b : Brand loyalty relates positively with brand equity for Canadian consumers.
Given the hypotheses previously deducted, the conceptual model proposed is presented in Figure 1.

Survey development
An online survey was launched based on the literature review. The country of origin measure (Table  2 and Table 3) was adapted from Yasin et al. (2007) and is composed by eight items. Ethnocentrism was measured with a five-item scale version of Shimp and Sharma's (1987) CETSCALE, adapted from Strizhakova and Coulter (2015) and Alden et al. (2013). The purchase intention measure was adapted from Xie et al. (2015) and is composed of three items. Brand associations, brand awareness and brand loyalty were measured with three-item scales, adapted from Yasin et al. (2007) and Yoo et al. (2000). Perceived quality was measured with a six-item scale adapted from Yoo et al. (2000). Brand equity was measured with a seven-item scale adapted from Yasin et al. (2007). All constructs were measured on a Likert scale, with the response anchors from 1, indicating "strongly disagree, to 5 "indicating" strongly agree".
The survey was prepared in English and then translated into Portuguese, following the back-translation process. Prior to the questionnaire released, a pre-test was sent to Canadian and Portuguese correspondents to identify possible improvements in the scope of ques-tions understanding and adaptation to the reality of each country.
Through the application of an exploratory factor analysis, common method bias (Podsakoff and Organ, 1986) was analyzed. The results of the exploratory factor analysis revealed that there is no single factor that explains most of the results variance. In order to test for non-response bias, the early and late respondents (comprising the first 75% and the last 25% to answer the questionnaire, respectively) were compared across all variables in the model (Armstrong and Overton, 1977). No significant differences were found between the initial respondents and the late respondents.

Sample
This study was developed in order to analyze the influence of the country of origin in the different dimensions of brand equity, in the wine sector, considering the perceptions of Portuguese and Canadian consumers. The Portuguese domestic wine market currently demonstrates a high growth potential and Canada was identified as a strategic market by ViniPortugal (2019) in the strategic plan for the internationalization of the wine sector in Portugal.
Following a quantitative approach and a non-probability sampling procedure, an online survey was implemented in Portugal and Canada to collect the data. To reach a diversified and wide sample, the survey was implemented across universities, wine associations and wine importers. Participants were invited to participate in the study by email. The survey was addressed to individuals over the age of eighteen with Portuguese and Canadian nationalities.
A total of 208 valid responses were collected in Portugal and 63 valid responses in Canada, in the year 2018. In the Portuguese sample, 50.5% of respondents are male and 49.5% are female, which indicates a balanced sample in terms of gender. In the Canadian sample 57.1% are male and 42.9% are female.
The most representative age group in both samples is 28 to 37 years old, with 45.2% and 42.9% of Portuguese and Canadian respondents, respectively. In the Portuguese sample, 40.9% of respondents have a bachelor's degree and in the Canadian sample, this percentage is 49.2%. The Portuguese sample consumes an average of 1 bottle of wine per month and presents Portugal and France as preferred countries of origin. The Canadian sample consumes an average of 2 bottles of wine per month and presents the USA and Italy as preferred countries of origin for wine purchase. Table 1 presents the demographic characteristics of the sample.

RESULTS
Data were analyzed with Partial Least Squares Structural Equation Modelling (PLS-SEM) in Smart PLS 3.0. (Ringle, Wende, and Becker, 2015), a technique widely used in behavioral sciences. We selected this method because PLS-SEM is particularly recommended to predict and explain the variance of key constructs through different explanatory constructs (Hair et al., 2012). Consequently, this technique should be selected when prediction is one of the main focus of the research (Hair et al., 2020). Moreover, PLS-SEM is recommended to test complex models with a large number of latent variables (Hair et al., 2020). Furthermore, PLS-SEM is strongly appropriate to test structural models with small sample sizes (Barclay and Smith, 1997). In the evaluation process of the partial least squares structural equation model, the steps identified by Hulland (1999) were followed. The analysis was started with the assessment of the model reliability and validity, followed by the structural model evaluation.
To analyze the measurement model the individual items' evaluation was conducted. All constructs possess items with loadings superior to 0.6 (Bagozzi and Yi, 2012), ensuring convergent validity -Tables 2 and 3. Alpha Cronbach values exceed the desirable value of 0.60, which demonstrates that the scales are reliable and the used measures have content validity (Hair et al., 2009). The composite reliability for each construct is above the desirable value 0.7 (Bagozzi and Yi, 2012). The Average Variance Extract (AVE) in each construct is above the reference value of 0.5 (Fornell and Larcker, 1981) ensuring convergent validity of the constructs.
The assessment of discriminant validity followed Fornell and Larcker's (1981) recommendations. The fact that the shared variance among any two constructs (the square of their intercorrelation) is less than the average variance explained in the items by the construct (Table  4 and Table 5) supports the existence of discriminant validity.
To evaluate the structural model, it is recommended an assessment through the use of R 2 , the level of explained variance of each endogenous variable, which should be superior to 10% (Falk and Miller, 1992). As Table 2 and Table 3 evidence, this condition is verified for each endogenous variable.
Regarding the Portuguese sample, Hypothesis 1a is supported by the results as shown by the following values: (β)=0.099; p<0.05, t-value=1.878 (Table 6). Therefore, the country of origin Portugal is positively related with purchase intention for Portuguese consumers. The results also demonstrate that brand equity relates positively with purchase intention for Portuguese consumers, supporting Hypothesis 2a (β =0.784; p<0.001, t-value =17.999). Moreover, Portuguese consumers' ethnocentrism relates positively with the country of origin Portugal, supporting Hypothesis 3a (β = 0.525; p<0.001, t-value = 10.742). The study also demonstrates that the country of origin Portugal relates positively with perceived quality for Portuguese consumers, supporting Hypothesis 4a (β = 0.714; p<0.001, t-value = 17.309). Results support Hypothesis 5a with the following values (β = 0.510; p<0.001; t-value = 6.578), therefore the country of origin Portugal relates positively with brand associations for Portuguese consumers. Hypothesis 6a is supported (β = 0.472; p<0.001, t-value = 9.693), suggesting that country of origin Portugal relates positively with brand awareness for Portuguese consumers. Regarding Hypothesis 7a, the results demonstrate that the country of origin Portugal relates positively with brand loyalty for Portuguese consumers (β = 0.549; p<0.001; t-value = 9.174). Hypothesis 8a is supported by the results (β = 0.481; p<0.001; t-value = 8.177) thus, perceived quality relates positively with brand equity for Portuguese consumers. Hypothesis 9a is not supported (β = -0.026; t-value = 0.376), therefore results don´t provide evidence for a significant effect of brand associations on brand equity. Moreover, Hypothesis 10a is not supported (β = 0.01; t-value = 0.209), thus results provide evidence for a non-significant effect of brand awareness on brand equity. The positive relationship between brand loyalty and brand equity predicted in Hypothesis 11a is supported by the results (β = 0.459; p<0.001, t-value = 7.077).   Concerning the Canadian sample, the positive relationship between country of origin and purchase intention predicted in Hypothesis 1b is not supported by the results (β =0.131; t-value =0.801) - Table 7. Regarding Hypothesis 2b, the results demonstrate that brand equity relates positively with purchase intention for Canadian consumers, supporting Hypothesis 2b (β =0.656; p<0.001, t-value =8.259). Moreover, Canadian consumers' ethnocentrism relates negatively with the country of origin Portugal, supporting Hypothesis 3b (β = -0.397; p<0.001, t-value = 3.749). The study also demonstrates that the country of origin Portugal relates positively with perceived quality for Canadian consumers, supporting Hypothesis 4b (β = 0.405; p<0.001, t-value = 4.001). Hypothesis 5b is not supported (β = 0.058; t-value = 0.443), therefore results do not provide evidence for a significant effect of the country of origin Portugal on brand associations for Canadian consumers. Moreover, Hypothesis 6b is not supported (β = 0.205; t-value = 1.318), thus results provide evidence for a non-significant effect of the country of origin Portugal on brand awareness for Canadian consumers. Regarding Hypothesis 7b, the results demonstrate that the country of origin Portugal relates positively with brand loyalty for Cana-dian consumers (β = 0.405; p<0.001, t-value = 4.348). Hypothesis 8b is not supported by the results (β = 0.132; t-value = 1.292) thus, the relationship between perceived quality and brand equity is non-significant for Canadian consumers. The positive relationship between brand associations and brand equity predicted in Hypothesis 9b is supported by the results (β = 0.196; p<0.05, t-value = 2.300). Hypothesis 10b is not supported (β = -0.109; t-value = 0.975), thus results provide evidence for a nonsignificant effect of brand awareness on brand equity. The positive relationship between brand loyalty and brand equity predicted in Hypothesis 11b is supported by the results (β = 0.697; p<0.001, t-value = 8.255).

DISCUSSION AND MANAGERIAL IMPLICATIONS
This study analyses the influence of country of origin on the different brand equity dimensions, in the wine sector, considering the perceptions of Portuguese and Canadian consumers.
Results demonstrate that country of origin partially influences purchase intention. Regarding Portuguese consumers, findings show that consumers are sensi-  tive to wine origin during the buying process, preferring to buy national wines. This result is in accordance with previous studies that indicate country of origin as a crucial consumers' decision factor (Yunus and Rashid, 2016). Moreover, Aichner et al. (2017) indicate that a country of origin with strong associations generally exerts a positive impact on purchase intention. However, concerning the Canadian consumers, the positive influence of country of origin on purchase intention was not demonstrated. This result might be explained by the still incipient Portuguese country-product image in the Canadian market, as well as, by the weak presence of the  Results demonstrate that brand equity positively influences purchase intention for both Portuguese and Canadian consumers, corroborating several previous studies (e.g. Hoeffler and Keller, 2003;Aaker, 1991) who report that brands with stronger brand equity are commonly preferred by the consumer.
Considering the accentuation of protectionist policies, we have been witnessing in the world economy, and their consequences (Andrews et al., 2018), the degree of ethnocentrism has been analyzed in order to understand the behavior of Portuguese and Canadian consumers regarding the country of origin effects. The results show that the effects of ethnocentrism on country of origin are emphasized in the Portuguese market, since the degree of ethnocentrism positively enhances the country of origin effect, accentuating the preference of Portuguese consumers for national wine. Moreover, the degree of ethnocentrism of Canadian consumers negatively influences the effect of Portugal country of origin, restricting their preference for imported wines.
The strong relationship between country of origin and perceived quality, defended by previous authors (e.g. Elliott and Cameron, 1994), was demonstrated in both Portuguese and Canadian samples. Consumers of both nationalities are likely to transpose the favorable characteristics that Portugal as a country conveys, to this product category.
The influence of country of origin on consumers' brand associations was partially demonstrated. The results of the Portuguese sample demonstrate the positive effect, therefore, as advocated by Pappu et al. (2006), favorable associations to the country of origin are transferred to the brand. However, the results of the Canadian sample did not support this positive influence, which might be explained by the limited offer of Portuguese wine in Canada, compared to other sources, which may hinder the consumer's ability to transfer possible qualities, memories and benefits to the brand.
The fact that the product is not "massively" available in the Canadian market might explain the unsupported relationship between country of origin and brand awareness in the Canadian sample, contrarily to what is suggested in the literature (Pappu et al., 2007). Additionally, Portugal may hasn´t achieved yet a distinctive level of wine producer country, compared to other countries present in the Canadian market, such as Italy and France. However, in the Portuguese sample, the positive influence of the country of origin on brand awareness is supported, which is in accordance with previous studies (e.g. Pappu et al., 2007). Therefore, the Portuguese mar-ket relies heavily on national demarcated wine regions to underline the country of origin influence.
The positive influence of country of origin on brand loyalty was supported in both samples. This indicates that a favorable country image can increase brand popularity and enhance brand loyalty (Septyanti and Hananto, 2017). Therefore, loyalty to a country may be transferred to loyalty to the country's brands, through the familiarity with a country's products (Paswan et al., 2003).
Regarding the positive influence of Aaker's (1991) four brand equity dimensions, namely perceived quality, brand associations, brand awareness and brand loyalty on brand equity, our results weren´t totally consistent with previous literature. The results partially supported the positive influence of perceived quality on brand equity. In accordance with previous literature (e.g. Liu et al., 2017), the results of the Portuguese sample demonstrate that perceived quality positively influences brand equity. However, the results of the Canadian sample did not support this relationship. One possible explanation for this unsupported relationship, besides the fact that the sample may not be representative of the population, is that Portugal ranks ninth in total wine imports of the Canadian market (Canadian Vintners Association, 2019), both in terms of value and quantity. As defended by Zeithaml (1988), the degree of perceived quality increases with a long-term relationship with the brand, enabling the recognition of brand differentiation and superiority. Thus, compared to other wines available in the Canadian market, Portuguese wines may still evidence some disadvantage in this brand equity dimension.
Moments, episodes, and facts constitute associations that consumers can produce regarding one brand, which will lead them to select one brand over another, thus contributing to the creation and enhancement of brand equity. Although the relationship between brands associations and brand equity has been previously accepted in the literature (Yasin et al., 2007), regarding the wine industry, this study only partially supported it (in the Canadian sample). For Portuguese consumers, the impact of brand associations on brand equity is not significant, which may be explained by the fact that consumers' associations may be essentially connected to the product and not to the brand. Moreover, it is important to emphasize that the survey was drawn up addressing the Wines of Portugal brand, instead of specific national brands.
Previous studies (e.g. Keller, 2008) have demonstrated the importance of brand awareness for brand equity reinforcement. However, this relationship was not supported in this study. In both Portuguese and Canadian samples, the impact of brand awareness on brand equity does not reach the importance of other brand equity dimensions, namely perceived quality, brand associations and brand loyalty.
The relationship between brand loyalty and brand equity is generally accepted in the literature and has been pointed out as preponderant in the creation of brand equity (Aaker, 1996). The results corroborate this relationship in both samples, therefore for wine consumers of both nationalities, brand loyalty positively influences brand equity.
In the specific case of the Portuguese market, the current country's positive phase regarding tourism may be an important tool to promote the wine industry. The promotion of wine tourism can improve the reputation of wine regions and highlight their differentiating characteristics (Frochot, 2003), allowing the consumer to transfer these positive associations to the brands coming from that region/country. The creation of favorable consumer experiences may be a tool to enhance the wine perceived quality and the creation of memories that enable increased brand loyalty and brand equity (Madeira et al., 2019). This suggestion can be applied to both the Portuguese and Canadian markets. In addition to the above benefit, this action may also contribute to change the Canadian consumer habits towards the product category under study, as according to 2018 data on alcoholic beverages (Statistics Canada, 2019), Canadians prefer beer, with wine appearing as a second choice (only in two provinces British Columbia and Quebec is the situation reversed).
The results of the Canadian sample highlight the need to enhance Portugal's reputation as a distinguished and high-quality wine country producer. To this end, joint and coordinated actions should be considered between governmental and private entities. This joint effort will benefit not only Wines of Portugal, but also individual brands wishing to internationalize to this market or to strengthen their presence, leveraged by the country of origin effect.
The results of this study highlight the work that needs to be done for wine producer countries to become reference countries in this product category and to raise brands to another level of recognition and distinction.

Limitations
Our results supported the perspective that country of origin is important for brand equity (Koschate-Fischer et al., 2012), however, country of origin's effects may diverge depending on several characteristics, such as nationality (Johansson et al., 1985) and culture (Bal-anis et al., 2002). Therefore, the results presented may have been influenced by both the diversity and size of the sample. In the Portuguese sample, most individuals were from Leiria district and aged 28 to 37 years, and the Canadian sample was composed mostly of individuals from the province of Alberta and aged 28 to 37 years. At the same time, the size of the samples and the disparity between them may also affect the results.
The study analyzed the wine industry, however, the generalization of the results needs to be conducted with caution, since specific wines have different behaviors, for instance the Portuguese Porto wine has a different behavior compared to the industry average.

Future research
Future investigations may extend the analysis of country of origin's effect on wine brands to other markets. Future studies may also analyze how other factors may influence the country of origin's effects on brand equity, such as the difference between inexperienced and experienced consumers. Additionally, in this study four brand equity dimensions identified by Aaker (1991) were used to measure brand equity. Future studies may examine the influence of other brand equity dimensions within the wine industry.

Final conclusion
Given the above results, the wine market is sensitive to the effects of the country of origin on brand equity, therefore players in the wine industry, such as brands, and producers' associations, should emphasize their origin in their communication strategies. It should be noted that the results suggest that efforts should be made to create brand associations, such as associating the brand to sustainable agriculture policies and cultural events, in order to create favorable consumer memories and to leverage brand equity. Moreover, improving brand image and investing in distinctive packaging, design and memorable logo are possible spontaneous and assisted vehicles to create brand awareness and enhance brand equity.