Sub-national consumer ethnocentrism and the importance of the origin of food products: An exploratory analysis

Purpose This study’s main objective is to analyse the role of the consumer’s ethnocentrism as a potential segmentation basis and to detect product origin-sensitive groups. The relationship between the consumer’s regional ethnocentrism, local and regional identity and corresponding valuation and purchase of food products from a region is also examined. Design/methodology


Introduction
The literature has addressed the role of the origin of food products from different perspectives. Firstly, international marketing literature has discussed over the last five years whether a product's origin is important to consumers. However, the choice is limited to two alternatives: national vs. foreign products (Diamantopoulos et al., 2011;Samiee, 2011), and it appears as if the sub-national origin is irrelevent to the consumer. Secondly, the literature on regional products and geographical indications explicitly recognises that one reason for the consumer's choice of these products is precisely the product's origin. This origin is associated with unique product features, but so far limited attention has been given to the possible influence of normative factors, such as consumer ethnocentrism (CET), or to affective factors, such as regional identity, when analysing the respective purchase decision process (van Ittersum et al., 2007;Fernández-Ferrín et al., 2017). Finally, in the literature on local products, the product's place of origin or rather the distance between this location and the consumer's place of residence was found to be of great importance in the consumer's choice of these products, with criteria such as the production methods used or the size of the production company also showing an impact on this choice (Lang et al., 2014;Martinez et al., 2010). In the literature on local products, however, the debate has primarily focused on the choice between local and nonlocal products, with little attention paid to the brand, which is an important element in the international marketing literature and should not be ignored when trying to better understand the consumer's choice.
As Balabanis and Siamagka (2017, p. 168) point out, research on consumer ethnocentrism effects has focused on predicting purchasing attitudes and intentions (but without adequately addressing purchasing behaviour by using "different forms of aggregate measures rather than specific outcomes (e.g. purchase of a specific brand)". These authors also indicate that "research on the effects of consumer ethnocentrism on specific behaviours remains scarce, revealing a research gap and creating an opportunity to theoretically and empirically examine the relationship between CET and behavioural outcomes".
In addition to addressing this gap, this paper aims to respond to another recent call for more research on these subtle conditioning factors of consumers' purchase of products associated with a certain geographical origin. As Fischer and Zeugner-Roth (2017, p. 202) argue, "recent criticisms of Country-Of-Origin (COO) research argue that the (cognitive) COO label only matters for certain (affective and normative) consumer segments such as those that are highly ethnocentric or that identify strongly with their home country…yet to date, little empirical research demonstrates this effect". Van Ittersum (1999, p. 49) further asserts that the "ethnocentric consumers have a significantly higher preference for products from their own area of residence than for products made elsewhere" and that 'area of residence' may be as diverse as their own country (Shimp and Sharma, 1987), their region (Fernández-Ferrín and Bande-Vilela, 2013), or their locality (Lantz and Loeb, 1998).
This study is based on social identity theory (Tajfel and Turner, 1986), which states that people tend to identify with certain groups of individuals and to differentiate themselves from other groups. In the literature, this theory has explained consumer choice between domestic and foreign products and has also served as a basis for understanding consumer preference for local products (Lantz and Loeb, 1998).
To address the above unresolved issues in the literature, a theoretical framework is constructed that integrates concepts from different areas of study to better understand the regional consumer ethnocentrism role as an antecedent of attitudes, valuation and purchase of specific food products in two different regions. It is proposed that consumers' regional ethnocentrism may result in increased importance being attributed to the product's origin and also in a more active search for consumer information, specifically when reading product labels; on the other hand, the relationship between consumer ethnocentrism, initially suggested as an antecedent of the decision to purchase national vs. foreign products (but also recently employed in several sub-national contexts), and a series of variables related to the valuation of and intention to purchase specific food products from their region and from other regions shall be analysed. These two objectives are specified in the following research questions (RQ): RQ1: How important is the product's origin in relation to the consumer's purchase decision and to what degree do consumers seek information on the origin of food products on labels?
RQ2: What is the relationship between the level of consumer ethnocentrism, the importance attributed to the product's origin and the tendency to read product labels to know its origin? RQ3: What is the profile of the most ethnocentric consumers, their attitude towards buying products from their own versus other regions, the significance of these products for them and their valuation of local and regional food products?
This research question can be divided into four sub-questions: RQ3a: What is the profile of the most ethnocentric consumers?
RQ3b: What is their attitude towards buying products from their own versus other regions?
RQ3c: What do their region's products mean to them?
RQ3d: What is their valuation and purchase of specific food products from their region?

Consumer ethnocentrism
Consumer ethnocentrism is based on the more general concept of ethnocentrism, a term coined by Sumner (1906), which refers to the way in which individuals identify themselves as members of a group (ingroup) while at the same time differentiating themselves from other groups (outgroups). This concept "may apply to any social group and it mixes neatly with the social identity theory concept of ingroup favoritism" (Lantz and Loeb, 1996, p. 375).
The term 'consumer ethnocentrism' was first introduced by Shimp (1984) and Shimp and Sharma (1987, p. 280) to refer to "the beliefs held by American consumers about the appropriateness, indeed morality, of purchasing foreign-made products. 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; products from other countries (i.e., outgroups) are objects of contempt to highly ethnocentric consumers". These authors developed a measurement scale, CETSCALE, which proved to have good psychometric properties when used to measure American consumers' ethnocentric trends. Netemeyer, Durvasala and Lichtenstein (1991) confirmed these properties by using a reduced version of the scale in three other countries (France, Japan and Germany) in addition to the United States. From then on, the use of the scale became generalised and different reduced versions were used. Three decades later a meta-analysis by Guo and Zhuo (2017) based on 60 previous studies confirmed that consumer ethnocentrism has a positive influence on domestic product judgments and willingness to buy domestic products. Alsughayir (2013) and Vabø et al. (2017) arrived at similar conclusions.
CETSCALE has also been applied at a sub-national level, showing that levels of regional consumer ethnocentrism are related to a greater valuation of local or regional products (Ouellet, 2007;Poon, Evangelista and Albaum, 2010), a greater preference for products from the region (van Ittersum, 1999;Yildiz et al., 2018), a greater purchase intention (Fernández-Ferrín and Bande-Vilela, 2013) and a greater rejection of products from outside the region (Fernández-Ferrín and Bande-Vilela, 2015). In terms of social identity theory, consumers have a certain identity associated with their group, for example a national or regional identity, and the CETSCALE, according to Lantz and Loeb (1996), would measure the economic manifestation of that identity.
Some recent investigations warn of inconsistencies in these studies regarding (1) the delimitation of the consumer ethnocentrism concept, which is defined as an attitude, a belief system or a personality trait (Papadopoulos et al., 2018), and (2) the effects on consumer behaviour because of the limited number of studies analysing their effects on effective purchase behaviours of specific food products (Balabanis and Siamagka, 2017).

The importance of the product's origin and consumer ethnocentrism
The importance assigned to product origin in consumer decision making is closely associated with the COO effect concept, one of the most studied concepts within the field of international marketing (Jain, 2007). COO is generally suggested to act as an extrinsic cue, along with price or brand and assists the consumer with the purchase choice of products, especially when other information regarding intrinsic attributes is missing. Iacob (2014), following a systematic literature review, concludes that there is inconclusive evidence of the impact of COO on consumer perceptions, in particular with regard to brands and their relationship with the concept of consumer ethnocentrism. Thøgersen et al. (2017, p. 547) note that the COO effect is not uniform but "complex, explained by the underlying processes of cue utilization and halo effects, contingent on a number of antecedents (e. g. ethnocentrism, cultural orientation, economic development, geographical closeness and familiarity, product-country fit) and moderated by both individual-based and product factors".
Some work based on meta-analysis leads to similar conclusions. Peterson and Jolibert (1995) conclude that "country-of-origin effects are only somewhat generalizable" (p. 883) and highly context-dependent. Samiee et al. (2016)  Similarly, Fischer and Zeugner-Roth (2017) conclude that COO influences consumer valuations and purchase intentions through three processes which constantly interact: a cognitive one, designed as 'product ethnicity', which is related to the country image and implies that consumers have certain expectations about a product, depending on the favourable or unfavourable image of a particular country as a producer of a certain category of products; a normative one, related to consumers' moral considerations as to whether it is appropriate to purchase products that have not been produced in their own country, and which has given rise to numerous studies focused on the 'consumer ethnocentrism' concept (Shimp and Sharma, 1987); and finally, an affective one, related to positive or negative feelings towards one's own country or foreign countries, which has given rise to numerous studies around the 'national identity' concept (Abraha et al., 2015;Zeugner-Roth et al., 2015). Balcombe et al. (2016) further add that the importance of COO information depends very much on the food category considered.
All of the foregoing research seems to indicate that the effect of perceptions regarding the product's origin is not uniform but depends on the consumer's socio-demographic characteristics, cognitive factors such as country image, the product category under consideration, normative variables such as the consumer's ethnocentrism and affective variables such as social identity; so it is perfectly possible that for some consumers the product's origin has little importance in their purchasing decisions, while for others it may well be a decisive factor. For Samiee (2011) the information available to consumers on the origin of the product is also not uniform; on the contrary, "in cases where country-of-origin is salient and diagnostic, the country-of-origin sensitive consumer segment can look for this information on a product's label" (p. 478), which is the most important source of information about a product's origin. Companies can correspondingly adjust the COO information on food labels according to their target (Balcombe et al., 2016). Papadopoulos et al. (2018, p. 736) analysed the research and the complexity of the COO effects concept. They decided to use the equivalent term 'place effects', with the understanding that "place does not begin or end at the country/nation level, but rather refers to everything from a seat at a dinner table to a room and onwards to the universe…place is not just a spot on the map; it encompasses and reflects the identities, memories, heritage, experiences and characteristics that are associate with it and the meanings ascribed to it…place does not just "exist" -it is constructed in the minds of observers".
This approach is not new. In the last two decades, literature has raised the need to address the importance of product origin from a sub-national perspective. As van Ittersum et al. (2003, p. 215) assert "the influence of a product's place of origin on a product evaluation has mainly been studied from a country-of-origin perspective", but regions of origin may have unique aspects that require studying the effects of the product's origin at a sub-national level. The literature on local or regional products takes a very different approach to the literature in international marketing: the product's origin is considered an important sign of quality, the analysis focusing almost exclusively on food products, while the product brand is often ignored. A noteworthy exception is the research conducted by Fernández-Ferrín et al. (2017, p. 72) who find that the effective purchase of local food brands is influenced by brand valuation, considered a private motivation, that is "intrinsic, hedonistic, or appropriable" and by community identity, considered a public motivation, that is "extrinsic, ethical, or altruistic". Telligman et al. (2017) argue that the place of origin is not always the most important quality indicator and that there are large differences in the importance given to origin. Their review of previous work leads them to conclude that in countries like Italy or France, consumers value these aspects more than in other countries like Spain or Belgium.
The consumers' attitude to regional products, i.e. of products "whose quality and/or fame can be attributed to its region of origin and which is marketed using the name of the region of origin" (van Ittersum et al., 2007, p. 2), and which are sometimes protected (as in the case of protected designation of origin products), depends both on their attitude towards the region of origin and on the image of regional certification labels.
Although the use of geographical indications (GI) is very irregular, with a much greater weight in the Mediterranean countries than in northern Europe and in Europe than in the United States, there is a belief that they are beneficial to the consumer because they provide information about the product's origin and ensure a certain level of quality that can be derived from factors such as soil quality, climate or regional expertise of production. In addition, these GIs are associated with increased support for local producers and protection of traditional cultural values (Teuber, 2011).

Samples
Some previous studies highlight the moderating role of product category and geographical context on the effects of consumer ethnocentrism on purchase decision (Fernández-Ferrín and Bande-Vilela, 2013, Shu et al., 2013). They found that consumer ethnocentrism effects are product-and country/region-specific (Balabanis and Diamantopoulos, 2004;Fernández-Ferrín et al., 2018). For this reason -and in order to measure ethnocentrism at a sub-national level, which is less present in the academic literature -different categories of food products were considered whereas data collection in two regions in Spain: the Basque Country and Galicia, was conducted. These two regions were selected because they are historic communities with strong levels of regionalism/nationalism (Brown, 2018;Mees, 2019) and rich gastronomic traditions. In both, much importance is given to fine cuisine. Galicia is known for the richness of its gastronomic products, such as veal, octopus, seafood, potatoes or wine (Fernández Ferrín et al., 2015;Rey Barahona, 1996). The Basque Country is the birthplace of renowned chefs (Aranceta-Bartrina and Perez-Rodrigo, 2019), such as Martín Berasategi, who has been awarded ten Michelin stars.
3.1.1. Sample 1 Using a combination of the snowball sampling procedure and convenience sampling, an electronic questionnaire was administered to a sample of 227 consumers residing in the Basque Country. After eliminating the substantially incomplete responses (17 cases) as well as outliers, detected using the Mahalanobis distance criterion for the variable 'regional ethnocentrism' (9 cases), 201 cases were obtained and included in the analysis. The age range was from 20 to 81 years, with a mean equal to 44 years (M = 44.1, SD = 11.4); the proportion of women (55 percent) surpassed that of men. Most respondents (more than 92 percent) were born in 21 different populations in the Autonomous Community of the Basque Country, with Vitoria-Gasteiz and Bilbao being the most represented populations in the sample.

Sample 2
Using the same sampling procedure as for Sample 1, the second sample used in the study was composed of 176 consumers residing in Galicia. A total of 157 responses were included in the analysis after excluding 15 incomplete responses and 4 cases that can be considered as outliers. The age range in this case was from 19 to 66 years, with a mean equal to 33 years (M = 33.4, SD = 12.2) and the proportion of men, at 57.6 percent, was greater than the proportion of women. All the respondents resided in the Autonomous Community of Galicia and the majority (147) were also born in 38 different municipalities of the four provinces of the region (A Coruña, Lugo, Ourense and Pontevedra).

Measurements
A reduced version of CETSCALE (Shimp and Sharma, 1987) was used to measure the ethnocentric tendencies of consumers with respect to the boundaries of their region, which proved to have good psychometric properties (Klein et al., 2005) and was previously used in sub-national contexts (Fernández-Ferrín and Bande-Vilela, 2013;Fernández-Ferrín et al., 2018). To measure the importance of product origin and the propensity to search for product information on labels, two indicators were developed for research: "For me, product origin is very important in the decision to buy a product" and "I usually read product labels to know a product's origin".
The scale for measuring rejection of non-regional products is an adaptation of the scale used by Klein et al. (1998). The variables 'consumer use of regional brands as quality signals', 'consumer use of regional brands as self-identity signals' and 'involvement with regional brands' were measured using scales adapted from the versions used by Strizhakova et al. (2012). The measurement scale for regional identity is adapted from that used by Lantz and Loeb (1998) and the measurement scale for local identity is based on the indicators used by Tu et al. (2012).
In each of the two territories (Basque Country and Galicia), four products manufactured in the region were selected and two were selected from outside the region. The search was made for products that are well known and usually consumed in these territories, including diverse categories, namely potatoes, cheese, wine and milk in the Basque Country, and dry-cured ham, milk, coffee and beer in the case of Galicia. Table 1 shows these products which were presented to respondents for assessment. Basque consumers valued four products produced within the political-administrative limits of the region, and two consumed at a national level. In the case of Galician consumers, four products from the region, and two from other regions, were evaluated. All of these products were presented to respondents under their brand name.
The scale of measurement for product assessment is adapted from Chryssochoidis et al. (2007) and Ouellet (2007). Effective purchase was measured by the consumer's estimate of the percentage of the product's consumption compared to total consumption in this food category. With the exception of this last scale, measured from 0 to 100%, the rest of the measurements use seven-point Likert scales with extremes at 1 (completely disagree) and 7 (completely agree).
Insert Table 1 about here

Results
With respect to the reliability of the measurement scales, the Cronbach's alpha analysis reflects very high levels of internal consistency in the variables' measurement scales (see Table 1 The frequencies in the responses to the two questions regarding the importance of origin and the propensity to search for origin information on labels show that the answers are highly segmented across the range of values; i.e. both the importance of origin and the level of interest in knowing it vary greatly among the consumers analysed. In both samples of Basque and Galician consumers, the median value is 4 for the importance of origin and 5 for the search for product origin information on labels. Considering that the scale of measurement has seven points, these results indicate that for about one half of the sample (both in the Basque Country and Extremadura) the origin of the product has a reduced and intermediate importance, compared to the other half, which considers it very important in their decision to purchase food products. Similarly, half of the sample has a medium or low propensity to seek information on the origin and the other half is much more active in looking for information about the product's origin.

Research Question 2:
The relationship between the level of consumer ethnocentrism, the importance attributed to the product's origin and the tendency to read product labels to know its origin The second research question is addressed through a mediation analysis with multiple regressions, using the PROCESS statistical macro for SPSS (version 2.16.2) and following the process recommended by Hayes (2013). The results of these analyses, which are shown in Table  2, firstly indicate that Basque consumers' regional ethnocentrism is positively and significantly related to the importance they attach to the origin of the product in the purchase decision (a = 0.67, p < .01). This variable has a tendency to lead to an increased information search about product origin on the product label (b = 0.78, p < .01). Similar results can be observed among Galician consumers: the higher the levels of regional ethnocentrism, the greater the importance of the origin of the product (a = 0.40, p < .01) and the greater the importance of origin, the greater the active search is for information about the origin on product labels (b = 0.81, p < .01). It is also interesting to observe that the consumers' regional ethnocentrism is related to the search for product origin information only in an indirect way, through a greater importance attached to this origin. Table 2 also shows the direct, indirect and total effects of regional ethnocentrism. The results support the existence of indirect effects, since the confidence intervals calculated for the sample of Basque and Extremadura consumers exclude the value 0. The data, however, does not support the existence of direct effects of regional ethnocentrism on the search for product origin information. In short, ethnocentric consumer tendencies influence the search for product origin information and do so through a greater attention to product origin.
Insert Table 2 about here 4.3. Research Question 3: Profile of the most ethnocentric consumers, their attitude towards buying products from their own versus other regions, the significance of these products for them and their valuation of local and regional food products.
To address the third research question, a cluster analysis was conducted in several stages. The variables used to form the groups are the six indicators of the scale of regional ethnocentrism, which presents a high value of internal consistency (Cronbach's alpha = 0.92 in both samples of Basque and Galician consumers).
The variables used in this cluster analysis are expressed in the same measurement unit, so a hierarchical cluster analysis with squared Euclidean distance (Ward) was performed. The dendrograms resulting from the analysis of the Basque and Galician consumer data allow a clear identification of two consumer groups. Consequently, the membership of these two groups was kept as a variable and the initial centroids were obtained, which were then considered in a kmeans (non-hierarchical) analysis and, by means of an iterative process, led to two new centroids in each sub-sample maximizing high homogeneity within each group.
The Brown-Forsythe test shows that there are significant differences between the two groups of consumers according to their degree of regional ethnocentrism for each of the six indicators considered. The correlation between the variables that measure group membership, corresponding to the two methods employed (hierarchical and k-means), was very high (r = 0.97; p<0.01 in the Basque sample and r = 0.93; p < .01 in the Galician sample).
To compare the two consumer segments derived from their scores on the regional ethnocentrism scale, several Brown-Forsythe analyses were conducted. The results, included in Table 3, indicate that both in the Basque Country region and in Galicia, consumers with a higher regional ethnocentrism: (1) have a greater tendency to reject products produced outside the region; (2) tend to consider regional brands to a greater extent as both a sign of quality and selfidentity; (3) show a greater involvement with regional brands; (4) manifest significantly higher values of local and regional identity; and (5) attach greater importance to the origin of the As far as the valuation of food products is concerned (RQ3d), participating consumers reviewed six products, four of which were produced in their region and two from outside of it. In general, Basque consumers rate the four food products produced in their region relatively highly, but the division of the sample into two segments according to levels of regional ethnocentrism allows us to observe a series of interesting effects (see Table 3). First, the more ethnocentric consumers give a significantly higher score than the less ethnocentric consumers to three of the four products under consideration; secondly, while in the less ethnocentric consumer group there is no difference between the assessment of milk produced in the region and the other two nonregional milk products, in the group of consumers with high ethnocentricity, there are significant differences between regional and non-regional milk alternatives in favour of the product from the region (mean difference Milk 1 from the region-Milk 2 from other regions = 0.94, p< 0.01; mean difference Milk 1 from the region-Milk 3 from other regions = 0.78, p< 0.01); thirdly, the most ethnocentric consumers show significantly higher consumption of cheese, wine and milk with regional origin and significantly lower consumption of a non-regional milk brand.
Insert Table 3 about here With regard to the results obtained in the sample of Galician consumers in relation to the valuation and purchase of food products produced inside and outside the region, the following should be highlighted: (1) brands in the region are generally more valued than those outside the region for both the more and the less ethnocentric sub-samples; (2) significant differences are only observed between the two sub-groups of consumers in the assessment of a regional origin brand: Dry-cured Ham and Sausages and in the effective purchase of another product of regional origin: Milk; in the other products analysed, both the valuation and the effective purchase are similar in both segments; (3) the most valued and consumed product of all those analysed is Beer from the region, which represents between 70 and 75 percent of all beer consumed by the Galician consumers participating in the study (see Table 3).
The results also show the absence of significant relationships between the levels of regional ethnocentrism of Basque and Galician consumers and the demographic variables gender and age (see Table 3).

Discussion and conclusions
One of the main objectives of this study was to analyse the role of consumer ethnocentrism as a variable of segmentation and detection of origin-sensitive groups. Analyses of the responses of two consumer samples in two different geographical contexts (Basque Country and Galicia) have allowed us to observe that perceptions about the importance of the origin of food products are quite varied. This polarity regarding the importance given to product origin is consistent with the debate in the international marketing literature about the relevance of the products' origin within consumer behaviour. For some authors, many consumers find the brand name to be a more effective stimulus than the origin of the product (Usunier, 2011) and as a result, producers usually reserve the front of the product for the brand name and logo, while information about the product's origin is relegated to the back of the product, thus being presented in a less evident way. This study's results, however, indicate that origin is not only important for a considerable segment of consumers, but that greater interest in knowing the origin leads consumers to search for origin-related information on product labels.
These results have important public, social and economic implications. Public administrations should require producers, through greater regulation, to provide more information on the origin of the products they commercialise, given the observed consumer interest in knowing a product's origin and the consumer propensity to search product labels for this information.
These requirements would also be in the local producers' interest, enhancing their products' appeal amongst ethnocentric consumers while strengthening local economy and the aggregate effect of the regional food brand (which could potentially benefit all regional food producers).
On the other hand, these results indicate that for a significant part of consumers (especially those with greater ethnocentric tendencies) a product's regional origin is important and greater access to regional brands, which for them represents signals of quality and self-identity, could affect their quality of life, especially in regions where gastronomy is a valuable cultural heritage.
Our results, also in line with the literature, support not only the existence of more originsensitive groups (Samiee, 2011) but also the existence of variables that help identify these consumer segments (Fischer and Zeugner-Roth, 2017;Peterson and Jolibert, 1995). The mediation analyses carried out show that the consumer's regional ethnocentrism is initially related to a greater importance being given to a product's origin, and through this variable the consumer has a tendency to read product labels to determine the place of origin The results allow us to observe a pattern that is repeated in both territories: the more ethnocentric the consumer, the more important the origin of the product will be and, consequently, the more active they will be in searching for information on that origin.
Secondly, the consumer's regional ethnocentrism is related to a greater rejection of food products from outside the region. Literature supports this effect, but only in the context of measuring willingness to buy foreign products and consumer ethnocentrism at a national level (Klein et al., 1998). This study's results show the existence of these effects also when ethnocentrism is measured at a sub-national level and when considering products from other regions. Consumers demand product origin information about the products they consume, including those which originate from their region; public authorities should ensure access to this information.
Thirdly, the most ethnocentric consumers at the regional level are also those who show strongest levels of identity with their closest environment: local and regional. Social identity is another key concept in origin-sensitive consumer segmentation, as some authors suggest in the field of international marketing (Abraha et al., 2015;Zeugner-Roth et al., 2015). At the subnational level, local identity has been found to be a predictor of the valuation and purchase of local food products (Fernández-Ferrín et al., 2017, 2018. Our results accordingly support the close relationship between ethnocentrism and social identity at a sub-national level. The preference for regional brands may be additionally enhanced by social identity reasons and also by sustainability concerns in a time of increasing awareness regarding the climate crisis, encouraging people adhere to lifestyle changes, such as 'buying local' (Telligman et al., 2017). Through this kind of behaviour, consumers could thus reveal sustainability concerns regarding their social, cultural, economic and ecological consumption effects. In future studies it might be interesting to further distinguish these dimensions, which could overlap to form an overall preference.
Fourthly, this study reveals that in some cases consumer ethnocentrism leads to a greater valuation of food products whose origin is close to the consumer, but the relationship is not clear and highly dependent on the product or category, particular food brands, and also on the geographical context analysed. Thus, while the more ethnocentric Basque consumers give a significantly higher rating to three of the four regional food products presented to them (cheese, wine and milk), the Galician consumers with a higher level of regional ethnocentrism value only one of the four products proposed (milk). This high variability in the effects of ethnocentrism on different products and contexts has already been pointed out in previous studies (Fernández-Ferrín and Bande-Vilela, 2013;Shu et al., 2013) and highlights the need to consider different types of product categories when studying the influence of this variable on consumer purchasing behaviour.

Further research
Further research could identify other mediating and moderating variables that may condition the effect of consumer ethnocentrism, such as the perceived necessity of the product, its availability, the frequency of purchase of the product or the conditions under which the product is consumed. It would also be interesting to study the possible influence of ethnocentric consumer trends on attitudes and valuations towards other types of sustainable products such as organic products or fair trade products. Another avenue of research could be to incorporate into the analysis measures of consumer moral virtues, such as integrity, and to examine the possible interaction with ethnocentric tendencies. Since this work has focused on two geographical regions within the same country, it would also be appropriate to conduct studies in other countries and considering other product categories. Finally, as this study is based on reported consumer behaviour it would be interesting to determine if reported behaviour translates to actual purchasing behaviour.  Table 2. Model coefficients for the mediation analysis and direct and indirect effects of regional consumer ethnocentrism on propensity to search labels for source information, through the importance of the origin of the product.  *Food products in the Basque Country sample: (1) Potatoes; (2) Cheese; (3) Wine; (4) Milk 1; (5) Milk 2; and (6) Milk 3. ** Food products in the Galician sample: (1) Dry-cured Ham and Sausages; (2) Milk; (3) Coffee; (4) Beer 1; (5) Beer 2; and (6) Beer 3. *** A Chi-squared test is used to analysing the relationship between gender and membership of the two clusters formed from ethnocentric tendencies, given that the two variables are categorical (dichotomous).  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46  47  48  49  50  51  52  53  54  55  56  57  58  59  60