A comparative study of food values between the United States and Norway

A comparative study of food values between the United States and Norway Abstract We compare the food values in the USA and Norway using the best–worst scaling approach. The food values examined are aimed at capturing the main issues related to food consumption such as naturalness, taste, price, safety, convenience, nutrition, novelty, origin, fairness, appearance, environmental impact and animal welfare. Results show that respondents in both countries have mostly similar food values, with safety being the most important value; while convenience and novelty are the least important values. Specifically, US respondents consider price more important and naturalness less important than Norwegian respondents. 1. Introduction The food systems in Europe and the United States significantly differ in terms of agricultural production practices, agricultural policy and marketing of foods. For example, many discussions have been raised regarding the use of genetically modified organisms (GMOs) and growth hormones in food production since European regulations on these food production issues are notably stricter than in the USA (Chern et al., 2002; Lusk, Roosen and Fox, 2003; Alfnes, 2004; Loureiro and Umberger, 2007; Delwaide et al., 2015). At the same time, food consumption trends in the USA can affect food patterns in Europe and vice versa (Mitchell, 2004), e.g. the local food movement. The development of different forms of Alternative Agri-Food Networks (AAFNs) such as farmers’ markets or Communities Supported Agriculture (CSA), for instance, first occurred in USA in the 1970s and 1980s but these have only recently become more popular in Europe (Martinez et al., 2010; Bazzani and Canavari, 2013). In addition, the adoption of nutrition food labelling is currently a widely discussed topic both in US and European food systems; but while nutritional labels have been regulated by the Food and Drug Administration (FDA) in USA since the early 90s, the European Union (EU) has only very recently introduced uniform or harmonised nutritional food labelling regulations (Nayga, Lipinski and Savur, 1998; Bonsmann and Wills, 2012; Soederberg Miller and Cassady, 2015). Although the presence of ethical and environmental food labels has consistently grown both in Europe and in the USA, the development of sustainable food labels occurred more recently in the USA in comparison to the European food system (Getz and Shreck, 2006; Golden et al., 2010; Grunert, Hieke and Wills, 2014; Ilbery et al., 2005; Louriero and Lotade, 2005). Moreover, the European food system is characterised by the presence of labels indicating specific regions of origin such as protected designation of origin (PDO), protected geographic origin (PGO) or country of origin (COOL) (Loureiro and Umberger, 2007; Aprile, Caputo and Nayga, 2012). Another notable difference is that the US food market is generally less developed in terms of traceability systems than the European food market, although US consumers have increasingly called for foods labelled as produced in the USA (Loureiro and Umberger, 2007; Lim et al., 2013). In order to capture these similarities and differences across European and US food systems, several studies have explored European and US consumers’ attitudes towards food claims, aiming at the development of potential international marketing strategies and policies (Roininen, Lähteenmäki and Tuorila, 1999; Chern et al. 2002; Bech-Larsen and Grunert, 2003; Lusk, Roosen and Fox, 2003; Lusk et al.2004; Loureiro and Umberger, 2007). The existing literature investigating consumers’ food attitudes in Europe and the USA has mainly focused on consumers’ evaluations of food safety claims and their attitudes towards genetically modified (GM) products (Chern et al. 2002; Lusk et al., 2004). The findings in these studies generally suggest that people in Europe are less willing to accept GM foods. For example, Chern et al. (2002) showed that Norwegian consumers were more willing to pay for non-GM vegetable oil and salmon than US consumers. Similarly, Alfnes and Rickertsen (2003), Lusk, Roosen and Fox (2003) and Alfnes (2004) showed that European consumers were willing to pay a higher price for beef from cattle that had not been administered growth hormones and Lusk, Roosen and Fox (2003) showed a higher willingness to pay for cattle that had not been fed with GM corn among Europeans when compared with US consumers. More recently, Rickertsen, Gustavsen and Nayga (2017) assessed consumers’ willingness to pay for GM soybean oil, farmed salmon fed with GM soy and GM salmon. Interestingly, their results suggest a large similarity in WTP in Norway and the USA and across the three products. Additionally, Rozin, Levine and Stoess (1991) investigated factors affecting individuals’ preferences for different kinds of chocolate bars, using students from universities in the USA, Belgium and France as a subject pool. They observed that US students were more health-oriented in making their choices, while Belgian and French students were more pleasure-oriented. Bech-Larsen and Grunert (2003) also showed that US consumers were more willing to buy functional foods than Danish and Finnish consumers, mainly because of health-related motivations. Finally, Basu and Hicks (2008), investigated US and German consumers’ evaluations for fair-trade coffee using a choice experiment approach and found that German respondents were more inequality averse than US consumers. Generally, studies investigating consumers’ preferences in the USA and Europe have limited their analyses to the assessment of consumers’ evaluations for specific food attributes such as GM production, nutritional content, use of growth hormones or sustainability issues. Lusk and Briggeman (2009) (henceforth LB) claimed that individuals’ food choices may be explained by their preferences for more abstract food quality attributes1 which LB identified as intermediary values, that ‘relate specifically to people’s food choices’ (Lusk and Briggeman, 2009: 186). These so-called ‘food values’ can be considered as more stable than consumers’ preferences for a specific set of food attributes on specific food products. According to LB, food values can explain individuals’ food choices across a variety of food products and do not depend on the specific context under investigation. However, to the best of our knowledge, no study has compared the food values between the USA and Europe, which is the aim of our study. In this study, we identify a set of 12 food values, which differ slightly from the set that was used by LB. These values are aimed at capturing the main issues related to food consumption patterns such as naturalness, taste, price, safety, convenience, nutrition, novelty, origin, fairness, appearance, environmental impact and animal welfare. In order to measure individuals’ preferences for food values, we implement a best–worst scaling (BWS) approach. The choice of this approach has been determined by the fact that Lee, Soutar and Louviere (2007) observed that the use of BWS provided better outcomes than other rating methods in measuring human values. In addition, BWS is particularly appropriate in cross-country comparisons, since the use of other forms of rating scales might lead to scalar inequivalence, which is generally caused by divergences in lexicon and response styles across different cultures (Baumgartner and Steenkamp, 2001; Auger, Devinney and Louviere, 2007; Jaeger et al., 2008; Ter Hofstede, Steenkamp and Wedel, 1999; Loose and Lockshin, 2013). For example, Mueller Loose and Lockshin (2013) and Dekhili, Sirieix and Cohen (2011) showed that the BWS method worked well to explore differences across countries in rating a set of attributes on wine and olive oil products, respectively. A potential limitation of BWS could be the lack of complete transitivity in attribute importance and therefore of consistency in dominance relations of attribute importance ranking. However, Lagerkvist (2013), in a study investigating Swedish consumers’ preferences for food quality attributes on beef, explored these issues using different rating methods such as BWS and direct ranking (DR) and showed that estimates at the aggregate level from BWS were more consistent than the estimates from DR both in terms of preference relations and of dominance ordering of attribute importance. We specifically compare the food values in Norway and the USA for several reasons. The Norwegian regulations on the use of biotechnology are quite restrictive and so one would expect more resistance against production methods based on modern biotechnology. In addition, the Norwegian food environment is very different from the US food environment. In contrast to the USA, Norwegian agriculture is dominated by small scale farming. The average farm size in Norway was 23.4 hectares in 2015 according to the Norwegian Institute of Bioeconomy Research (NIBIO) (NIBIO, 2016: 24), and the average dairy herd was 25 dairy cows in 2014 (Budsjettnemda for jordbruket, 2015), while in the USA the average farm size and the average dairy herd were 438 acres and 2,017 dairy cows, respectively, in 2014 (Progressive Dairyman, 2016; Statistica, 2017). Furthermore, the key tenets of the Norwegian agricultural policy are different from those in the USA. There are four main objectives of the Norwegian agricultural and food policy: (i) food security (with emphasis on having high domestic production of agricultural products, especially meat and dairy products), (ii) agricultural production in all parts of the country, (iii) increased value of the agricultural products and (iv) sustainable agriculture (for example through the target that 15 per cent of the production and consumption should be organic before 2020) (NIBIO, 2016: 12, 49). These objectives are supported by one of the highest levels of agricultural subsidies in the world. Producer support estimates were 61 per cent of gross farm receipts for the period of 2007–2009 as compared with only 9 per cent in the USA (OECD, 2010). Moreover, Norway has very high-import tariffs for products such as dairy and meats and, consequently, very little trade with these products (NIBIO, 2016: 54). Opinion polls also show a strong public support for the current state. In a recent poll, 90 per cent of the respondents wanted to maintain Norwegian agriculture on at least the present level (Norsk Landbruk, 2014). Finally, while the average per capita income, measured at purchasing power parities, is quite similar in the two countries, Norway is characterised by a more equal distribution of income. According to the OECD (2017), Norway was the second-most equal OECD country after Iceland in 2014 while the USA was the third-most unequal country. We believe that the differences between food systems in Norway and the USA make these two countries an interesting context to compare food values. Our hypothesis is that differences in agricultural systems might be related to differences in individuals’ food values. To illustrate, the adoption of high-agricultural subsidies, the enhancement of domestic and sustainable food production in Norway might be respectively related to the importance that Norwegian people give to food values such as fairness, origin and environmental impact. Moreover, even though food prices are relatively much higher in Norway than in the USA, the high degree of income equality in Norway may result in less emphasis on food prices and higher emphasis on fairness.2 To sum up, this study advances the literature in two important ways: (i) we adopt the concept of food values and the set of items used by LB to identify which food values are most important among US and Norwegian consumers and (ii) we compare consumers’ preferences for food values in a multi-country setting, considering credence (e.g. food safety and origin) as well as experience attributes (e.g. taste). Results from this study are of value to food marketers and policy makers for two main reasons. First, the comparison of consumers’ preferences between a European country and the USA is currently of particular interest since Europe and the USA are key trading partners and results from this study would help future trade negotiations (Luckstead and Devadoss, 2016). Second, while we do not want to lessen the contribution to the literature of previous studies comparing different countries’ consumer preferences for food attributes on specific food products, the results from our study could be applied to various commodities, and, therefore, could be used as a guide in the development and implementation of marketing strategies and food policies for a broad range of food products. To illustrate, if our results show, for example, a high preference for the food value ‘safety’ both in the USA and Norway, then this could encourage the support of policies aimed at increasing the traceability of food products, while a high rating for ‘naturalness’ could support the production and trade of foods produced without the use of modern technologies or pesticides, no matter what the product under consideration is. 2. Materials and methods This section is dedicated to the description of (i) the data collection, (ii) experimental design, i.e. selection of the food values and implementation of the BWS and (iii) applied econometric approach. 2.1. Data collection Data were collected from an online survey conducted between October and November of 2015 in Norway and the USA. More than 1,000 respondents in each country (1,037 in Norway and 1,025 in the USA) took part in the survey. Respondents were randomly recruited across regions and urban/non-urban areas in both countries by a professional market research agency called Ipsos.3 Respondents were invited to participate in an internet survey and were asked about the aspects they considered more or less important when buying food products. They were assured that any given information was anonymous and that they could quit the survey whenever they wanted to. The survey also contained questions about attitudes towards food claims. The selected samples in Norway and the USA were relatively representative of the national populations in terms of socio-demographic information. In Table 1, we report information related to the distribution of demographic and socio-economic variables in the two samples and of the US and Norwegian populations, respectively. Table 1. Demographic and socio-economic distribution in the USA and Norway   USA    Norway  Sample  Population    Sample  Population  Female (%)  51  51    50  50  Age (years)  40  39    53  39  Education (%)             Less than high school  3  17    3  27   High school  46  55    34  40   University degree  38  18    43  23   Post-university degree  13  10    20  10  Marital status (%)             Married  48  50    54  35   Cohabitant  7  NA    15  NA   Never been married  32  31    16  51   Separated or divorced  12  12    11  9   Widow or widower  1  7    4  5  Number of children in household (%)   No children  55  58    70  72   One child  19  18    11  13   Two children  16  16    12  11   More than two  10  8    7  4  Income (gross annual income) (%)   Less than $ 15,000  12  $53,718 (median)  Less than $12,500  1a  $61,387 (median)   $15,000–29,000  17    $12,500–24,900  2     $30,000–44,000  14    $25,000–37,400  3     $45,000–59,000  13    $37,500–49,900  7     $60,000–74,000  12    $50,000–62,400  10     $75,000–89,000  11    $62,500–74,900  12     $90,000–119,000  10    $75,000–87,400  30     $120,000–49,000  6    $87,500–99,900  17     $150,000 or more  5    $100,000 or more  18    Rural area (%)b  18  19    25  19    USA    Norway  Sample  Population    Sample  Population  Female (%)  51  51    50  50  Age (years)  40  39    53  39  Education (%)             Less than high school  3  17    3  27   High school  46  55    34  40   University degree  38  18    43  23   Post-university degree  13  10    20  10  Marital status (%)             Married  48  50    54  35   Cohabitant  7  NA    15  NA   Never been married  32  31    16  51   Separated or divorced  12  12    11  9   Widow or widower  1  7    4  5  Number of children in household (%)   No children  55  58    70  72   One child  19  18    11  13   Two children  16  16    12  11   More than two  10  8    7  4  Income (gross annual income) (%)   Less than $ 15,000  12  $53,718 (median)  Less than $12,500  1a  $61,387 (median)   $15,000–29,000  17    $12,500–24,900  2     $30,000–44,000  14    $25,000–37,400  3     $45,000–59,000  13    $37,500–49,900  7     $60,000–74,000  12    $50,000–62,400  10     $75,000–89,000  11    $62,500–74,900  12     $90,000–119,000  10    $75,000–87,400  30     $120,000–49,000  6    $87,500–99,900  17     $150,000 or more  5    $100,000 or more  18    Rural area (%)b  18  19    25  19  Sources: The data of the Norwegian population were extracted from Statistics Norway (2017) and the data of the US population were extracted from the United Census Bureau (United Stated Census Bureau, 2017). aExchange rate during the survey (15 October 2015) was USD 1 = NOK 8.00, which was used to convert the Norwegian income figures to USD. bThe standard definition of rural area according to ‘Norway Statistics’ is ‘a hub of buildings that is inhabited by less than 200 persons’, while the definition of rural area in the US Census Bureau is an area which is inhabited by less than 2,500 individuals. In our survey, we defined rural area as a settlement with a population lower than 1,000 individuals. Gender distribution was fairly similar in both samples with about 50 per cent and 51 per cent female respondents in the USA and Norway. The average age of the respondents was substantially higher in Norway (53 years) than in the USA (40 years). The average age of the Norwegian sample is higher than the average age of the Norwegian population. Regarding the education level, both samples are more educated than their respective country populations. Norwegian respondents, on average, had a somewhat higher education level than the US sample, and the Norwegian sample was also characterised by a higher percentage of married people (54 per cent) and cohabitants (15 per cent) than the US sample (48 per cent and 7 per cent, respectively). However, the percentage of married individuals in the Norwegian sample was higher than the Norwegian population, while the US sample was characterised by a slightly lower percentage of married people in comparison to the US population. On the other hand, respondents in the USA tended to have more children in the household compared with Norwegian respondents; however, most respondents in both countries indicated having no children in their household (70 per cent for Norway and 55 per cent for the USA), which closely resemble the statistics of the populations in the two countries. Notably, the majority of the respondents in the USA had an annual income equal or below $59,000 (56 per cent), while only 23 per cent of the Norwegian sample had an annual income equal or below $62,400. This is consistent with the median income of the populations in both countries, indicating that the annual median income is higher than in the USA. Importantly though, the income differences are calculated at market exchange rates that vary considerably over time and are quite different from the exchange rates calculated at rates that reflect the purchasing power. Finally, Table 1 shows that Norwegian and US populations have the same percentage of people residing in rural areas (19 per cent). However, the Norwegian sample included a higher percentage of people living in rural areas (28 per cent) than in the US sample (18 per cent). 2.2. Experimental design 2.2.1. Food values As previously mentioned, we followed the work of LB who specified 11 food values (naturalness, safety, environmental impact, origin, fairness, nutrition, taste, appearance, convenience, price and tradition). LB selected these attributes in an attempt to resemble the 10 values identified by Schwartz (1994). LB noted that some values considered by Schwartz, such as achievement and power, might not have a direct relation with food. However, one of the values identified by Schwartz is ‘stimulation’ that could be related to the excitement that ‘novelty’ could present. With the improvement in food technologies and growing globalisation, consumers are continuously offered new food products (Siro et al. 2008; Lee et al., 2015). In addition, a large body of literature shows that variety seeking plays an important role in consumers’ food choices and eating behaviour (Van Trijp and Steenkamp, 1992; Adamowicz and Swait, 2012; Frewer, Risvik and Schifferstein, 2013). Hence, we included ‘novelty’ in our set of food values. Recent literature also shows that consumers are increasingly interested in animal welfare (Carlsson, Frykblom and Lagerkvist, 2007; Napolitano, et al., 2008; Barber and Gertler, 2009). Animal welfare could also be associated with the Schwartz value of ‘universalism’ which resembles individuals’ ‘understanding, appreciation, tolerance, and protection for the welfare of all people and for nature’ (Schwartz, 1994: 22). Hence, we also included ‘animal welfare’ in our set of food values. However, we excluded ‘tradition’ which LB defined as ‘preserving traditional consumption patterns’ due to the growing globalisation of food markets. Indeed, due to increasing ethnic diversity in US and Norwegian populations, tradition is likely to be interpreted differently across respondents. Moreover, studies investigating the meaning of food tradition in six European countries (including Norway) showed that respondents tended to give different interpretations of food tradition depending on the country they belonged to and they especially tended to associate food tradition with different aspects of food consumption such as origin, locality, processing transformation, habits, naturalness, sensory property and familiarity (Guerrero et al., 2009; Pieniak et al., 2009; Almli, et al., 2011; Verbeke et al., 2016). Thus, the inclusion of ‘tradition’ in our set of food values would have been a confounder or would have been overlapping with other food values in our study. The 12 food values incorporated into our study, the food values in LB and the definitions used in the surveys are exhibited in Table 2. Table 2. Food values with descriptions in parentheses Lusk and Briggeman (2009)  This study  Naturalness (extent to which food is produced without modern technologies)  Naturalness (made without modern food technologies like genetic engineering, hormone treatment and food irradiation)  Safety (extent to which consumption of food will not cause illness)  Safety (eating the food will not make you sick)  Environmental impact (effect of food production on the environment)  Environmental impact (effects of food production on the environment)  Origin (where the agricultural commodities were grown)  Origin (whether the food is produced locally, in USA/Norway or abroad)  Fairness (the extent to which all parties involved in the production of the food equally benefit)  Fairness (farmers, processors and retailers get a fair share of the price)  Nutrition (amount and type of fat, protein, vitamins, etc.)  Nutrition (amount and type of fat, protein, etc.)  Taste (extent to which consumption of the food is appealing to the senses)  Taste (the flavour of the food in your mouth)  Appearance (extent to which food looks appealing)  Appearance (the food looks appealing and appetising)  Convenience (ease with which food is cooked and/or consumed)  Convenience (how easy and fast the food is to cook and eat)  Price (the price that is paid for the food)  Price (price you pay for the food)  Tradition (preserving traditional consumption patterns)      Animal welfare (well-being of farm animals)    Novelty (the food is something new that you have not tried before)  Lusk and Briggeman (2009)  This study  Naturalness (extent to which food is produced without modern technologies)  Naturalness (made without modern food technologies like genetic engineering, hormone treatment and food irradiation)  Safety (extent to which consumption of food will not cause illness)  Safety (eating the food will not make you sick)  Environmental impact (effect of food production on the environment)  Environmental impact (effects of food production on the environment)  Origin (where the agricultural commodities were grown)  Origin (whether the food is produced locally, in USA/Norway or abroad)  Fairness (the extent to which all parties involved in the production of the food equally benefit)  Fairness (farmers, processors and retailers get a fair share of the price)  Nutrition (amount and type of fat, protein, vitamins, etc.)  Nutrition (amount and type of fat, protein, etc.)  Taste (extent to which consumption of the food is appealing to the senses)  Taste (the flavour of the food in your mouth)  Appearance (extent to which food looks appealing)  Appearance (the food looks appealing and appetising)  Convenience (ease with which food is cooked and/or consumed)  Convenience (how easy and fast the food is to cook and eat)  Price (the price that is paid for the food)  Price (price you pay for the food)  Tradition (preserving traditional consumption patterns)      Animal welfare (well-being of farm animals)    Novelty (the food is something new that you have not tried before)  The food values include credence, experience and price attributes. Naturalness and safety are considered credence attributes since they are product characteristics that consumers cannot decipher just by looking at the product without any label information. In addition to naturalness and food safety, credence attributes were included that are related to sustainability and ethical issues such as environmental impact, origin, animal welfare and fairness. Finally, nutrition is a credence attribute related to the nutritional content of the food products. On the other hand, taste and appearance are experience attributes. Convenience and novelty can also be considered experience attributes; consumers can personally experience whether a food product is easy or fast to eat, or whether they have never tried a product before. Finally, price is the search attribute that identifies the money individuals pay to buy a food product. LB’s definitions were slightly modified in our study in order to make them more understandable to respondents in Norway and the USA. To illustrate, for naturalness, we indicated that this is food produced without the use of modern food technologies such as genetic engineering, hormone treatment and food irradiation. 2.2.2. Best-worst scaling The BWS approach was developed by Louviere and Woodworth (1990) and first published by Finn and Louviere (1992). It consists of a series of choice sets where respondents are asked to indicate among a (sub)set of attributes or statements which one they prefer the most (or consider the most important) and which one they prefer the least (or consider the least important). This approach has been defined by researchers as an extension of Thurstone’s (1927) paired comparison method in which respondents are asked to choose the best between paired items. Nowadays, BWS is a popular methodology that has been implemented in several research fields such as psychology, marketing and social and environmental sciences (Auger, Devinney and Louviere, 2007; Cohen, 2009; Scarpa et al., 2011; Lancsar et al., 2013). In food consumption literature, BWS has been mainly used for the estimation of consumers’ valuations for product attributes, as well as consumers’ food attitudes (Jaeger et al. 2008; Cohen, 2009; Lusk and Briggeman, 2009; Lagerkvist, Okello and Karanja, 2012; de-Magistris, Gracia and Albisu, 2014). The growing popularity of the BWS method is due to the fact that it provides several advantages over other common rating-based methods such as the Likert scale. In BWS, individuals can respond to the question only in one way, indicating which value is the most important and which one is the least important. This method forces individuals to make choices among values of the scale and does not allow the possibility to give the same value to all the issues in question. Comparatively, in rating scales, individuals might have their own evaluation for the scale values; for example, a three for one person could represent a four for another person, so they might use the scale differently. Finally, using a BWS approach, researchers can construct individual-level scales of preference/importance for each issue under consideration and accurately compare these scales (Cohen, 2009; Hein et al., 2008; Lusk and Briggeman, 2009). In BWS surveys, researchers have the option to use one of the three response mechanisms, which are generally described as BWS cases (Flynn and Marley, 2014; Rose, 2014; Beck, Rose and Greaves, 2017). In Case 1, the respondents are asked to choose the most preferred (most important) and the least preferred (least important) item among a list of items. In Case 2, items are not presented as a whole; rather for each choice set, respondents are asked to make a selection among a list of associated attributes and attribute levels. In Case 3, for each choice set, respondents are asked to select the best and worst from the alternatives which are described by a number of attributes and attribute levels of the items under investigation. In this study, we chose to use the Case 1 mechanism since this is the most appropriate approach for our research goal, i.e. investigating relative preferences for a list of food values (Flynn and Marley, 2014). When designing BWS experiments, researchers have to take into consideration both the potential number of choice sets and the potential number of the items per choice set. A large number of choice sets might induce fatigue to respondents, while a large number of items per choice set might decrease individuals’ attendance to the different attributes (Louviere et al., 2008; Scarpa et al. 2011; de-Magistris, Gracia and Albisu, 2014). For the allocation of the different items across the choice sets, we used a nearly balanced incomplete block design (NBIBD). The balanced incomplete block design (BIBD) is in general one of the most implemented experimental designs in Case 1 BWS surveys (Lee et al., 2007; Auger, Devinney and Louviere, 2007; Cohen, 2009; Flynn and Marley, 2014). This balance is due to each choice set being characterised by an equal number of items, and each item being repeated the same number of times across the choice sets. In addition, the items are orthogonally allocated, meaning that each item is paired with other items an equal number of times across the choice sets. However, researchers might find difficulties in generating a BIBD with a restricted number of choice sets and attributes per choice set. For this reason, different studies have implemented experimental designs where the orthogonality requirement is relaxed, i.e. partially balanced incomplete designs or nearly balanced designs (Erdem et al., 2012; Hamada, 1973; Street and Street, 1996; Orme, 2005; Jaeger et al. 2008; Thomson, Crocker and Marketo, 2010; Lagerkvist, Okello and Karanja, 2012). Our nearly BIBD consists of 12 choice sets, with each of the choice sets containing a subset of four food values. Each food value was repeated four times across the 12 choice sets and each food value was compared with each other 1.09 times, maximising the D-efficiency score (98.71 per cent) to satisfy the orthogonality property (Kuhfeld, 2005). Another important aspect of the nearly balanced incomplete design is that it also helps to minimise the possibility that preferences for values can be unintentionally inferred by features of the design. This way, violations of transitivity and dominance that may be related to the use of BWS can be reduced (Lagerkvist, 2013; Flynn and Marley, 2014). In Figure 1, we report an example of a choice set. Fig. 1. View largeDownload slide Example of a choice set. Fig. 1. View largeDownload slide Example of a choice set. In every choice set, respondents were asked to indicate which one among the four food values they considered the most important and which one they considered the least important when buying food products. If a respondent tried to choose more than one food value as the most or least important, they were told to choose only one value before they could continue to the next choice set in the online survey. 2.3. Econometric analysis Marley and Louviere (2005) describe the different probabilistic models for the best, worst and best-worst choices, by explaining theoretically the processes that respondents might follow in providing best and worst observations in BWS. These models are distinguished in three overlapping classes: random ranking and random utility models, joint and sequential and ratio scale models. Finally, in their paper, Marley and Louviere (2005) make a larger distinction between sequential and maximum difference (maxdiff) models (Flynn and Marley, 2014). A sequential model assumes that respondents make best and worst choices in a particular order (e.g. best first and then worst), while the maxdiff model, which is a well-established probabilistic model that was introduced by the pioneering work of Finn and Louviere (1992), assumes that respondents simultaneously choose the pair of items that maximises the difference between the best and the worst choices. In this study, we apply the maxdiff model for two main reasons: (i) the maxdiff is the most appropriate probabilistic model for the Case 1 BWS approach and (ii) estimating best and worst values separately can be a source of bias due to potential error variance differences between the best and worst choice observations (Louviere et al., 2015; Flynn and Marley, 2014; Scarpa et al. 2011; Rose, 2014). Data were analysed using a discrete-choice framework. Notably, discrete-choice models are consistent with random utility (McFadden, 1974) and Lancaster consumer theories (Lancaster, 1966). According to random utility theory, the utility for respondent n in choosing alternative j in choice set t, is   Unjt=Vnjt+εnjt (1)where Vnjt is a systematic component that can be observed by the researcher, while ɛnjt is the unobserved error term, which is assumed to be independent of Vnjt. Generally, when respondents are presented with a choice set, they make choices on the basis of the maximisation of the utility they can derive from each alternative of the presented choice set. As such, in making a choice between alternative j and alternative k, respondent n will pick alternative j over alternative k when:   Unjt>Unkt for⁢all⁢ ⁢j≠k. (2) However, in BWS experiments, respondents make choices depending on which pair of alternatives (most important and least important) maximises their utility. Specifically, in each choice set, respondent n chooses the pair of alternatives j and k as the best and worst, respectively, when   Unjt−Unkt>Unlt−Unmt for all⁢j≠l ⁢an⁢d⁢ k≠m. (3) Given that each choice set has J food values (4 in our case), the pair of items chosen by the respondent as best and worst represents a choice from all J(J − 1) possible pairs (12 in this study), which maximises the difference in importance. Following LB, λj, is defined as the observable location of the value j on the scale of importance. Given taste homogeneity this parameter will be constant across respondents. The unobserved level of importance of food value j for respondent n, Inj, is given by   Inj=λj+εnj (4)where ɛnj is the random error term; hence, the probability that the respondent chooses food value j as the best and food value k as the worst will be equal to the probability that the difference among Inj and Ink is larger than any of the J(J − 1) − 1 possible differences among the other food values in the choice set. Using a multinomial logit model (MNL), the probability of respondent n choosing j as the best and k as the worst among pairs of alternative J(J −1) is specified as follows:   Pnjk=eλj−λk∑l=1J∑m=1Jeλl−λm (5)where the choice of respondent n takes the value of 1 for the pair of values chosen by the respondent as best and worst, and the value of 0 for the non-chosen J(J − 1) − 1 pairs of food values. Specifically, λj represents the relative importance of food value j over one of the values, which is normalised to 0. In this way, the dummy variable trap can be avoided. Effects coding was applied: the food value takes the value 1 when the value is described as the best alternative, −1 when the value is described as the worst alternative, and 0 otherwise. The MNL assumes that the error terms are independently and identically distributed (IID) with a Gumbel (Extreme value type I) distribution and implies independence within the alternatives and taste homogeneity across respondents. Heterogeneity in respondents’ food values is likely. When heterogeneity valuations are expected, discrete-choice models such as the random parameters logit (RPL) model should be used. The RPL model allows for random taste variations and accounts for the panel structure of the data (Train, 2003). As such, in contrast to the MNL model, the importance parameter of value j in the RPL model is assumed to be different for each respondent n and was specified as follows:   λ̃nj=λ̅j+σjμnj (6)where λ̅j and σj are the mean and the standard deviation of λj, respectively, and μnj is a random error term that is assumed to be normally distributed with mean zero and unit standard deviation. Substituting equation (6) into equation (5), the RPL can be estimated by maximising a simulated log-likelihood function for μnj (Train, 2003).4 In the standard RPL model, independency across taste parameters is assumed; however, food values are expected to be interdependent. In order to take this interdependency into account,5 the correlation structure of the attribute parameters was assumed to follow a multivariate normal distribution. The estimates from the RPL model might be difficult to interpret because the random error term might vary across respondents, and therefore the mean of the parameter estimates of λj may be confounded with differences in scale. Hence, following LB, we calculated the share of preference, Sj, for each value, which explains how important respondents rate one value j over the other J values:   Sj=eλˆj∑k=1Jeλˆk. (7) Each share can be interpreted as the forecasted probability that the corresponding value is chosen as the most important. If value j has a twice as big preference share as another value, this indicates that the value j is twice as important as the other value. The share of preferences of all the J values must sum to one. 3. Results and discussion In this section, we describe the results obtained from the econometric analysis. Specifically, in our study, the standard MNL model and the RPL model were estimated. From the RPL model estimates, the respondents’ specific preferences for the different food values were calculated using the estimated parameters as priors conditional on actual choices made by each respondent.6 From these posterior estimates, the mean and individual shares of preferences for the 12 values were calculated. 3.1. Model estimates In the BWS approach, the importance of a set of attributes is estimated relative to one of these attributes (Marley and Louviere, 2005; Lusk and Briggeman, 2009; de-Magistris, Gracia and Albisu, 2014). Following LB, we used as the baseline the least important food value, based on the calculation of the percent of times each item was selected best or worst, which in our case is novelty. Estimates from the MNL and RPL models are reported in Table 3. Table 3. Estimates from MNL and RPL models Food value    MNL  RPL      USA  Norway  USA  Norway  Naturalness  Mean  1.598***  2.502***  2.974***  4.491***  (0.036)  (0.042)  (0.091)  (0.107)    SD      2.719***  3.676***      (0.085)  (0.122)  Safety  Mean  2.746***  3.381***  5.139***  6.158***  (0.040)  (0.045)  (0.070)  (0.111)    SD      3.478***  3.177***      (0.101)  (0.117)  Environmental impact  Mean  1.360***  2.131***  2.406***  3.801***  (0.036)  (0.041)  (0.084)  (0.097)    SD      2.421***  3.207***  (0.086)  (0.106)  Origin  Mean  0.918***  1.578***  1.732***  2.738***  (0.037)  (0.040)  (0.079)  (0.102)    SD      2.010***  3.432***  (0.081)  (0.117)  Fairness  Mean  1.228***  2.186***  2.185***  3.939***  (0.036)  (0.041)  (0.091)  (0.098)    SD      2.146***  3.170***  (0.094)  (0.010)  Nutrition  Mean  1.922***  2.404***  3.612***  4.466***  (0.037)  (0.042)  (0.092)  (0.097)    SD      2.606***  3.170***  (0.076)  (0.010)  Taste  Mean  2.113***  2.714***  3.912***  5.133***  (0.038)  (0.043)  (0.095)  (0.086)    SD      2.648***  1.487***  (0.081)  (0.098)  Convenience  Mean  0.748***  0.850***  1.331***  1.496***  (0.033)  (0.036)  (0.069)  (0.084)    SD      1.826***  2.464***  (0.080)  (0.084)  Appearance  Mean  1.114***  1.469***  2.112***  2.670***  (0.036)  (0.040)  (0.076)  (0.077)    SD      2.015***  1.609***  (0.102)  (0.073)  Price  Mean  1.741***  1.780***  3.219***  3.337***  (0.037)  (0.041)  (0.097)  (0.094)    SD      2.855***  2.388***  (0.083)  (0.108)  Animal welfare  Mean  1.544***  2.470***  2.750***  4.452***  (0.036)  (0.042)  (0.091)  (0.100)    SD      2.738***  3.124***  (0.102)  (0.104)  Novelty  Mean  0.000  0.000  0.000  0.000  SD  0.000  0.000  Number of choices    12,300  12,444  12,300  12,444  Log-likelihood    −26,384  −25,057  −22,161  −19,951  BIC    52,897  50,217  45,048  40,628  AIC    52,790  50,135  44,477  40,056  AIC/N    4.292  4.029  3.616  3.219  Food value    MNL  RPL      USA  Norway  USA  Norway  Naturalness  Mean  1.598***  2.502***  2.974***  4.491***  (0.036)  (0.042)  (0.091)  (0.107)    SD      2.719***  3.676***      (0.085)  (0.122)  Safety  Mean  2.746***  3.381***  5.139***  6.158***  (0.040)  (0.045)  (0.070)  (0.111)    SD      3.478***  3.177***      (0.101)  (0.117)  Environmental impact  Mean  1.360***  2.131***  2.406***  3.801***  (0.036)  (0.041)  (0.084)  (0.097)    SD      2.421***  3.207***  (0.086)  (0.106)  Origin  Mean  0.918***  1.578***  1.732***  2.738***  (0.037)  (0.040)  (0.079)  (0.102)    SD      2.010***  3.432***  (0.081)  (0.117)  Fairness  Mean  1.228***  2.186***  2.185***  3.939***  (0.036)  (0.041)  (0.091)  (0.098)    SD      2.146***  3.170***  (0.094)  (0.010)  Nutrition  Mean  1.922***  2.404***  3.612***  4.466***  (0.037)  (0.042)  (0.092)  (0.097)    SD      2.606***  3.170***  (0.076)  (0.010)  Taste  Mean  2.113***  2.714***  3.912***  5.133***  (0.038)  (0.043)  (0.095)  (0.086)    SD      2.648***  1.487***  (0.081)  (0.098)  Convenience  Mean  0.748***  0.850***  1.331***  1.496***  (0.033)  (0.036)  (0.069)  (0.084)    SD      1.826***  2.464***  (0.080)  (0.084)  Appearance  Mean  1.114***  1.469***  2.112***  2.670***  (0.036)  (0.040)  (0.076)  (0.077)    SD      2.015***  1.609***  (0.102)  (0.073)  Price  Mean  1.741***  1.780***  3.219***  3.337***  (0.037)  (0.041)  (0.097)  (0.094)    SD      2.855***  2.388***  (0.083)  (0.108)  Animal welfare  Mean  1.544***  2.470***  2.750***  4.452***  (0.036)  (0.042)  (0.091)  (0.100)    SD      2.738***  3.124***  (0.102)  (0.104)  Novelty  Mean  0.000  0.000  0.000  0.000  SD  0.000  0.000  Number of choices    12,300  12,444  12,300  12,444  Log-likelihood    −26,384  −25,057  −22,161  −19,951  BIC    52,897  50,217  45,048  40,628  AIC    52,790  50,135  44,477  40,056  AIC/N    4.292  4.029  3.616  3.219  Note: ***Indicate significance at the 1 per cent level. Numbers in parentheses are standard errors. Table 3 indicates that we obtained a better fit with the RPL than the MNL model in both samples as shown by the increases in the log-likelihood values and the reductions in the AIC, BIC and AIC/N statistics. In addition, Table 3 shows that the derived standard deviations of the attributes’ parameters are statistically different from zero, and our assumption of heterogeneity in preferences for the 12 values across individuals cannot be rejected. 3.2. Shares of preferences for the 12 food values On the basis of the RPL estimates, we assessed the preferences for the 12 food values by calculating their shares of preference. In Table 4, we report the shares of preference for the different values, from the most to the least important in each country. Table 4. Preference shares and rankings of importance of food values in USA and Norway Rank  USA  Norway  Food value  Share  Food value  Share  1  Safety  0.380*  Safety  0.313*  2  Price  0.115*  Naturalness  0.125*  3  Taste  0.112  Taste  0.112  4  Nutrition  0.088*  Animal welfare  0.098*  5  Naturalness  0.078*  Nutrition  0.094*  6  Animal welfare  0.077*  Price  0.074*  7  Environmental impact  0.039*  Fairness  0.060*  8  Fairness  0.028*  Origin  0.047*  9  Appearance  0.027*  Environmental impact  0.046*  10  Origin  0.026*  Appearance  0.018*  11  Convenience  0.020*  Convenience  0.011*  12  Novelty  0.012*  Novelty  0.002*  Rank  USA  Norway  Food value  Share  Food value  Share  1  Safety  0.380*  Safety  0.313*  2  Price  0.115*  Naturalness  0.125*  3  Taste  0.112  Taste  0.112  4  Nutrition  0.088*  Animal welfare  0.098*  5  Naturalness  0.078*  Nutrition  0.094*  6  Animal welfare  0.077*  Price  0.074*  7  Environmental impact  0.039*  Fairness  0.060*  8  Fairness  0.028*  Origin  0.047*  9  Appearance  0.027*  Environmental impact  0.046*  10  Origin  0.026*  Appearance  0.018*  11  Convenience  0.020*  Convenience  0.011*  12  Novelty  0.012*  Novelty  0.002*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the two samples is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 4 shows that the mean preference shares are statistically different in the two countries, except in the case of taste attribute. However, if we consider the differences in the ranking of the importance of food values across the two countries, respondents’ preferences are quite similar in many aspects. In both countries, safety is clearly the most important value with a share of 38.0 per cent in the USA and 31.3 per cent in Norway. The high importance of safety is in line with the results of LB, who also found that safety was the most important food value in the USA. After safety, there is a group of five values that are fairly close in importance with shares ranging between 11.5 per cent and 7.7 per cent in the USA, and 12.5 per cent and 7.4 per cent in Norway (price, taste, nutrition, naturalness and animal welfare). The remaining values have preference shares ranging between 3.9 per cent and 1.2 per cent in the USA, and 6.0 per cent and 0.2 per cent in Norway. Convenience and novelty are the least important values in both countries. These similarities in values may reflect a convergence in food values between Europe and the USA. Within these broad similarities in the rankings of food values, there are also some notable differences. Price was the second-most important value among the US respondents, which is consistent with the LB study. In contrast, Norwegian respondents considered price as the sixth-most important value. The relatively lower importance of price in Norway may be a reflection of the more equal income distribution. Furthermore, taste was rated as the third-most important value both in the USA and Norway, which again is consistent with the results in LB. Nutrition was predicted as most important for about 9 percent of the respondents in each country. This result is somewhat at odds with past studies that showed consumers in the USA tend to pay more attention to the nutritional content of food products, as compared to European consumers (Rozin, Levine and Stoess, 1991; Bech-Larsen and Grunert, 2003); however, the result may reflect a convergence between the two countries. Additionally, naturalness was the second-most important value of Norwegian respondents, while it was the fifth-most important value of US respondents, which is all consistent with the current literature and not surprising given the differences in food environment. Indeed, several studies have shown that European consumers are generally less willing to consume food that has been produced with technologies such as genetic modification, or with cattle fed with growth hormones (Chern et al. 2002; Alfnes and Rickertsen, 2003; Lusk, Roosen and Fox, 2003). In addition, this result is also consistent with LB who found that naturalness was rated as the fifth-most important food value. Food values concerning ethical aspects of food production such as fairness and animal welfare were ranked as more important by the Norwegian than the US respondents. The higher importance of fairness in Norway is as expected given that the social and economic welfare of farmers are crucial aspects in the Norwegian food system, and the result is also consistent with the high equality in income distribution. The higher importance of animal welfare in Norway may also reflect that animal welfare labelling regulations tend to be more developed in Europe than in the USA (Mitchell, 2001; Napolitano et al., 2010; Vandemoortele and Deconinck, 2014). Environmental impact was ranked as the seventh-most important attribute by the US sample and the ninth-most important by the Norwegian sample, however, the actual preference share was slightly higher in Norway. This result is not unexpected given the higher presence of regulated eco-food labels in the European food system than in the USA (Czarnezki, 2011). Not surprising, origin was rated as somewhat more important by the Norwegian respondents than the US ones. Although existing literature reports that consumers both in the USA and Europe are generally willing to pay a price premium for local or designated origin of food products (Darby et al., 2008; Aprile, Caputo and Nayga, 2012; de-Magistris and Gracia, 2014; Meas et al., 2015), origin is ranked relatively low in both countries. This result is consistent with LB, who found that origin was considered as the least important value in their US study. 3.3. Socio-demographic information and shares of preferences for the 12 food values Overall, the results suggest that US and Norwegian respondents differ mostly in terms of the ranking of price and naturalness. However, Table 1 shows that the US and Norwegian samples differ in terms of some socio-demographic variables, which might explain some of the similarities/differences in preferences for food values across the countries. Specifically, we observed that the two samples differ in terms of age, education, having children or not, income and belonging to rural/urban areas. As such, in order to test whether individuals’ preferences for food values may differ in terms of socio-demographic characteristics, we divided the US and Norwegian samples into different subgroups based on age (young/old), education (low/high), the presence of children in the household (with/without), income level (low/high), whether residing in urban/rural area (urban/rural) and whether the respondent had purchased organic food during the previous 12 months (purchased/not purchased). In case of age, education and income, we determined the grouping based on the median values in the samples, and then divided each sample into two groups. We estimated the RPL model for each subgroup and calculated the respondents’ shares of preferences for the subgroups. In addition, we also report results from t-tests to test whether the preferences for the food values differed among the subgroups within each country (indicated with asterisks in Tables 5–10). Table 5. Shares of preferences and rankings by country and age Rank  USA  Norway  Old (n = 526)  Young (n = 499)  Old (n = 490)  Young (n = 547)  1  Safety  0.411*  Safety  0.352*  Safety  0.287*  Safety  0.362*  2  Price  0.129*  Taste  0.107*  Naturalness  0.147*  Taste  0.131*  3  Taste  0.124*  Price  0.103*  Taste  0.124*  Animal welfare  0.098*  4  Naturalness  0.083  Nutrition  0.094*  Nutrition  0.083  Naturalness  0.087*  5  Animal welfare  0.073  Animal welfare  0.078  Animal welfare  0.073*  Nutrition  0.087  6  Nutrition  0.072*  Naturalness  0.078  Price  0.072  Price  0.082  7  Environmental impact  0.027*  Environmental impact  0.051*  Fairness  0.027*  Fairness  0.052*  8  Appearance  0.022*  Fairness  0.033*  Origin  0.022*  Environmental impact  0.039  9  Fairness  0.022*  Origin  0.031*  Environmental impact  0.022  Origin  0.037*  10  Origin  0.021*  Appearance  0.029*  Appearance  0.021  Appearance  0.015  11  Convenience  0.011*  Convenience  0.026*  Convenience  0.011  Convenience  0.010  12  Novelty  0.005*  Novelty  0.018*  Novelty  0.005  Novelty  0.001  Rank  USA  Norway  Old (n = 526)  Young (n = 499)  Old (n = 490)  Young (n = 547)  1  Safety  0.411*  Safety  0.352*  Safety  0.287*  Safety  0.362*  2  Price  0.129*  Taste  0.107*  Naturalness  0.147*  Taste  0.131*  3  Taste  0.124*  Price  0.103*  Taste  0.124*  Animal welfare  0.098*  4  Naturalness  0.083  Nutrition  0.094*  Nutrition  0.083  Naturalness  0.087*  5  Animal welfare  0.073  Animal welfare  0.078  Animal welfare  0.073*  Nutrition  0.087  6  Nutrition  0.072*  Naturalness  0.078  Price  0.072  Price  0.082  7  Environmental impact  0.027*  Environmental impact  0.051*  Fairness  0.027*  Fairness  0.052*  8  Appearance  0.022*  Fairness  0.033*  Origin  0.022*  Environmental impact  0.039  9  Fairness  0.022*  Origin  0.031*  Environmental impact  0.022  Origin  0.037*  10  Origin  0.021*  Appearance  0.029*  Appearance  0.021  Appearance  0.015  11  Convenience  0.011*  Convenience  0.026*  Convenience  0.011  Convenience  0.010  12  Novelty  0.005*  Novelty  0.018*  Novelty  0.005  Novelty  0.001  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 6. Shares of preferences and rankings by country and education level Rank  USA  Norway  High (n = 531)  Low (n = 494)  High (n = 653)  Low (n = 384)  1  Safety  0.384  Safety  0.363  Safety  0.293  Safety  0.319  2  Nutrition  0.113*  Price  0.139*  Naturalness  0.148*  Animal welfare  0.156*  3  Taste  0.112  Taste  0.120  Taste  0.128*  Taste  0.097*  4  Price  0.096*  Animal welfare  0.088*  Nutrition  0.108*  Naturalness  0.084*  5  Naturalness  0.093*  Naturalness  0.072*  Animal welfare  0.072*  Price  0.077  6  Animal welfare  0.063*  Nutrition  0.055*  Price  0.072  Nutrition  0.074*  7  Environmental impact  0.037*  Environmental impact  0.042*  Fairness  0.056  Fairness  0.061  8  Fairness  0.026  Fairness  0.031  Environmental impact  0.050*  Origin  0.059*  9  Origin  0.025  Origin  0.029  Origin  0.040*  Environmental impact  0.039*  10  Appearance  0.024*  Appearance  0.029*  Appearance  0.017*  Appearance  0.020*  11  Convenience  0.018  Convenience  0.021  Convenience  0.014  Convenience  0.012  12  Novelty  0.005*  Novelty  0.011*  Novelty  0.001*  Novelty  0.003*  Rank  USA  Norway  High (n = 531)  Low (n = 494)  High (n = 653)  Low (n = 384)  1  Safety  0.384  Safety  0.363  Safety  0.293  Safety  0.319  2  Nutrition  0.113*  Price  0.139*  Naturalness  0.148*  Animal welfare  0.156*  3  Taste  0.112  Taste  0.120  Taste  0.128*  Taste  0.097*  4  Price  0.096*  Animal welfare  0.088*  Nutrition  0.108*  Naturalness  0.084*  5  Naturalness  0.093*  Naturalness  0.072*  Animal welfare  0.072*  Price  0.077  6  Animal welfare  0.063*  Nutrition  0.055*  Price  0.072  Nutrition  0.074*  7  Environmental impact  0.037*  Environmental impact  0.042*  Fairness  0.056  Fairness  0.061  8  Fairness  0.026  Fairness  0.031  Environmental impact  0.050*  Origin  0.059*  9  Origin  0.025  Origin  0.029  Origin  0.040*  Environmental impact  0.039*  10  Appearance  0.024*  Appearance  0.029*  Appearance  0.017*  Appearance  0.020*  11  Convenience  0.018  Convenience  0.021  Convenience  0.014  Convenience  0.012  12  Novelty  0.005*  Novelty  0.011*  Novelty  0.001*  Novelty  0.003*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 7. Shares of preferences and rankings by country and the presence of children in the household Rank  USA  Norway  With (n = 457)  Without (n = 568)  With (n = 307)  Without (n = 730)  1  Safety  0.414*  Safety  0.354*  Safety  0.385*  Safety  0.279*  2  Taste  0.100*  Price  0.140*  Naturalness  0.119  Taste  0.123*  3  Nutrition  0.094*  Taste  0.131*  Taste  0.105*  Naturalness  0.117  4  Price  0.089*  Animal welfare  0.082  Price  0.093*  Animal welfare  0.116*  5  Naturalness  0.087*  Nutrition  0.076*  Nutrition  0.088  Nutrition  0.101  6  Animal welfare  0.071  Naturalness  0.072*  Animal welfare  0.062*  Price  0.065*  7  Environmental impact  0.041  Environmental impact  0.035  Fairness  0.046*  Fairness  0.061*  8  Origin  0.027  Fairness  0.029*  Environmental impact  0.042  Origin  0.054*  9  Fairness  0.024*  Appearance  0.028  Origin  0.033*  Environmental impact  0.048  10  Appearance  0.023  Origin  0.023  Appearance  0.020  Appearance  0.017  11  Convenience  0.019  Convenience  0.019  Convenience  0.006*  Convenience  0.015*  12  Novelty  0.011  Novelty  0.011  Novelty  0.001*  Novelty  0.002*  Rank  USA  Norway  With (n = 457)  Without (n = 568)  With (n = 307)  Without (n = 730)  1  Safety  0.414*  Safety  0.354*  Safety  0.385*  Safety  0.279*  2  Taste  0.100*  Price  0.140*  Naturalness  0.119  Taste  0.123*  3  Nutrition  0.094*  Taste  0.131*  Taste  0.105*  Naturalness  0.117  4  Price  0.089*  Animal welfare  0.082  Price  0.093*  Animal welfare  0.116*  5  Naturalness  0.087*  Nutrition  0.076*  Nutrition  0.088  Nutrition  0.101  6  Animal welfare  0.071  Naturalness  0.072*  Animal welfare  0.062*  Price  0.065*  7  Environmental impact  0.041  Environmental impact  0.035  Fairness  0.046*  Fairness  0.061*  8  Origin  0.027  Fairness  0.029*  Environmental impact  0.042  Origin  0.054*  9  Fairness  0.024*  Appearance  0.028  Origin  0.033*  Environmental impact  0.048  10  Appearance  0.023  Origin  0.023  Appearance  0.020  Appearance  0.017  11  Convenience  0.019  Convenience  0.019  Convenience  0.006*  Convenience  0.015*  12  Novelty  0.011  Novelty  0.011  Novelty  0.001*  Novelty  0.002*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 8. Shares of preferences and rankings by country and income level Rank  USA  Norway  Low (n = 441)  High (n = 584)  Low (n = 358)  High (n = 679)  1  Safety  0.373*  Safety  0.382*  Safety  0.258*  Safety  0.339*  2  Price  0.160*  Taste  0.123*  Animal welfare  0.128*  Taste  0.123  3  Taste  0.096*  Nutrition  0.101*  Taste  0.122  Naturalness  0.113  4  Animal welfare  0.080*  Price  0.089*  Naturalness  0.119  Nutrition  0.087*  5  Naturalness  0.076  Naturalness  0.084  Nutrition  0.112*  Price  0.083*  6  Nutrition  0.068*  Animal welfare  0.073*  Price  0.070*  Animal welfare  0.083*  7  Environmental impact  0.038  Environmental impact  0.039  Fairness  0.070*  Fairness  0.061*  8  Fairness  0.028  Appearance  0.026  Environmental impact  0.052*  Origin  0.059*  9  Appearance  0.026  Origin  0.026  Origin  0.051*  Environmental impact  0.039*  10  Origin  0.025  Fairness  0.026  Appearance  0.017  Appearance  0.020  11  Convenience  0.021  Convenience  0.019  Convenience  0.014  Convenience  0.012  12  Novelty  0.011  Novelty  0.012  Novelty  0.001*  Novelty  0.003*  Rank  USA  Norway  Low (n = 441)  High (n = 584)  Low (n = 358)  High (n = 679)  1  Safety  0.373*  Safety  0.382*  Safety  0.258*  Safety  0.339*  2  Price  0.160*  Taste  0.123*  Animal welfare  0.128*  Taste  0.123  3  Taste  0.096*  Nutrition  0.101*  Taste  0.122  Naturalness  0.113  4  Animal welfare  0.080*  Price  0.089*  Naturalness  0.119  Nutrition  0.087*  5  Naturalness  0.076  Naturalness  0.084  Nutrition  0.112*  Price  0.083*  6  Nutrition  0.068*  Animal welfare  0.073*  Price  0.070*  Animal welfare  0.083*  7  Environmental impact  0.038  Environmental impact  0.039  Fairness  0.070*  Fairness  0.061*  8  Fairness  0.028  Appearance  0.026  Environmental impact  0.052*  Origin  0.059*  9  Appearance  0.026  Origin  0.026  Origin  0.051*  Environmental impact  0.039*  10  Origin  0.025  Fairness  0.026  Appearance  0.017  Appearance  0.020  11  Convenience  0.021  Convenience  0.019  Convenience  0.014  Convenience  0.012  12  Novelty  0.011  Novelty  0.012  Novelty  0.001*  Novelty  0.003*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 9. Shares of preferences and rankings by country and living in urban/rural area Rank  USA  Norway  Rural (n = 188)  Urban (n = 837)  Rural (n = 257)  Urban (n = 780)  1  Safety  0.369  Safety  0.376  Safety  0.317  Safety  0.315  2  Price  0.132  Taste  0.117  Naturalness  0.159*  Taste  0.118*  3  Taste  0.124  Price  0.110  Animal welfare  0.122*  Naturalness  0.109*  4  Naturalness  0.099*  Nutrition  0.088*  Taste  0.084*  Nutrition  0.106*  5  Animal welfare  0.087  Naturalness  0.085*  Origin  0.077*  Animal welfare  0.090*  6  Nutrition  0.065*  Animal welfare  0.072  Fairness  0.065  Price  0.083*  7  Environmental impact  0.033  Environmental impact  0.040  Nutrition  0.061*  Fairness  0.057  8  Fairness  0.026  Fairness  0.028  Price  0.049*  Environmental impact  0.051*  9  Appearance  0.023  Appearance  0.027  Environmental impact  0.032*  Origin  0.038*  10  Origin  0.020  Origin  0.026  Appearance  0.021  Appearance  0.019  11  Convenience  0.015  Convenience  0.021  Convenience  0.011  Convenience  0.013  12  Novelty  0.006*  Novelty  0.012*  Novelty  0.001*  Novelty  0.002*  Rank  USA  Norway  Rural (n = 188)  Urban (n = 837)  Rural (n = 257)  Urban (n = 780)  1  Safety  0.369  Safety  0.376  Safety  0.317  Safety  0.315  2  Price  0.132  Taste  0.117  Naturalness  0.159*  Taste  0.118*  3  Taste  0.124  Price  0.110  Animal welfare  0.122*  Naturalness  0.109*  4  Naturalness  0.099*  Nutrition  0.088*  Taste  0.084*  Nutrition  0.106*  5  Animal welfare  0.087  Naturalness  0.085*  Origin  0.077*  Animal welfare  0.090*  6  Nutrition  0.065*  Animal welfare  0.072  Fairness  0.065  Price  0.083*  7  Environmental impact  0.033  Environmental impact  0.040  Nutrition  0.061*  Fairness  0.057  8  Fairness  0.026  Fairness  0.028  Price  0.049*  Environmental impact  0.051*  9  Appearance  0.023  Appearance  0.027  Environmental impact  0.032*  Origin  0.038*  10  Origin  0.020  Origin  0.026  Appearance  0.021  Appearance  0.019  11  Convenience  0.015  Convenience  0.021  Convenience  0.011  Convenience  0.013  12  Novelty  0.006*  Novelty  0.012*  Novelty  0.001*  Novelty  0.002*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the two subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 10. Shares of preferences and rankings by country and organic food purchases Rank  USA  Norway  Not purchased (n = 393)  Purchased (n = 577)  Not purchased (n = 491)  Purchased (n = 458)  1  Safety  0.373†  Safety  0.367†  Safety  0.311*†  Safety  0.294*†  2  Price  0.160*†  Naturalness  0.131*†  Taste  0.154*†  Naturalness  0.167*†  3  Taste  0.096*†  Nutrition  0.111*  Price  0.116*†  Animal welfare  0.115*†  4  Animal welfare  0.062†  Taste  0.083*  Nutrition  0.090†  Nutrition  0.105  5  Nutrition  0.053*†  Animal welfare  0.076†  Naturalness  0.087*†  Taste  0.084*  6  Naturalness  0.031*†  Price  0.070*†  Animal welfare  0.087*†  Fairness  0.075*†  7  Appearance  0.030*  Environmental impact  0.045*  Origin  0.046†  Environmental impact  0.069*  8  Environmental impact  0.027*  Fairness  0.030*†  Fairness  0.043*†  Origin  0.046†  9  Fairness  0.023*†  Origin  0.030*†  Appearance  0.024*  Price  0.032*†  10  Origin  0.018*†  Appearance  0.022*†  Environmental impact  0.023*  Appearance  0.007*†  11  Convenience  0.014*  Convenience  0.021*†  Convenience  0.015*  Convenience  0.006*†  12  Novelty  0.006*  Novelty  0.014*†  Novelty  0.003*  Novelty  0.001*†  Rank  USA  Norway  Not purchased (n = 393)  Purchased (n = 577)  Not purchased (n = 491)  Purchased (n = 458)  1  Safety  0.373†  Safety  0.367†  Safety  0.311*†  Safety  0.294*†  2  Price  0.160*†  Naturalness  0.131*†  Taste  0.154*†  Naturalness  0.167*†  3  Taste  0.096*†  Nutrition  0.111*  Price  0.116*†  Animal welfare  0.115*†  4  Animal welfare  0.062†  Taste  0.083*  Nutrition  0.090†  Nutrition  0.105  5  Nutrition  0.053*†  Animal welfare  0.076†  Naturalness  0.087*†  Taste  0.084*  6  Naturalness  0.031*†  Price  0.070*†  Animal welfare  0.087*†  Fairness  0.075*†  7  Appearance  0.030*  Environmental impact  0.045*  Origin  0.046†  Environmental impact  0.069*  8  Environmental impact  0.027*  Fairness  0.030*†  Fairness  0.043*†  Origin  0.046†  9  Fairness  0.023*†  Origin  0.030*†  Appearance  0.024*  Price  0.032*†  10  Origin  0.018*†  Appearance  0.022*†  Environmental impact  0.023*  Appearance  0.007*†  11  Convenience  0.014*  Convenience  0.021*†  Convenience  0.015*  Convenience  0.006*†  12  Novelty  0.006*  Novelty  0.014*†  Novelty  0.003*  Novelty  0.001*†  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. †The hypothesis that the mean of the corresponding values are the same across the subgroups within the organic purchasers and non-purchaser is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. In Tables 5 and 6, we report the shares of preferences, respectively, for young and old respondents, and for respondents with a high and low education level in each country. From Tables 5 and 6, we observe that the shares of preferences for the different food values tend to be similar across the age and education groups in both countries. Only in the USA did we observe a difference in the rank of the attribute nutrition across higher and lower educated groups. Nutrition is rated as the second-most important value by higher educated people, while it is rated as the sixth-most important value by lower educated people. We would expect that the presence of children in the household would also conspicuously influence respondents’ preferences for the attribute nutrition (Drichoutis, Lazaridis and Nayga, 2006). However, Table 7 shows that this is not the case in either sample. Indeed, nutrition is rated in the USA as the third-most important value by respondents with children in the household and the fifth-most important by respondents without children. In the Norwegian sample, nutrition is equally rated by the two subgroups. Table 7 actually shows that there are no substantial differences in the rating of the importance of the food values across respondents with and without children in the household. However, an interesting result is that price is ranked fourth by respondents living with children both in the USA and Norway. In regards to the price attribute, the difference in the importance of price may be explained by a higher income level and a more equal income distribution in Norway. In Table 8, we report the preference shares of low- and high-income respondents in both countries. In the USA, lower income respondents considered price to be slightly more important than higher income respondents, and price is the second-most important food value for lower income respondents and the fourth-most important for higher income respondents. On the other hand, price is rated as the fifth-most important food value by higher income Norwegian respondents and the sixth-most important value by lower income respondents. These results suggest that price preferences between the income subgroups within each country tend to be similar. Finally, we tested whether residing in rural/urban area had a significant impact on respondents’ preferences for the food values. Table 9 shows that in Norway, the ranking of origin notably changes depending on whether the respondent resides in rural or urban area: origin is on average the fifth-most important attribute for individuals living in rural areas and the ninth-most important for individuals living in urban areas. However, in the case of the USA, we do not observe this difference in the ranking between the two subgroups. According to these results, we might conclude that socio-demographic variables scarcely explain the differences/similarities in preferences for food values between the two countries. However, LB found that the preferences for the different food values particularly differed among organic food purchasers and non-purchasers. Specifically, LB observed that price and naturalness (the two attributes which Norwegian and US respondents valued most differently in our survey) were the most differently rated food values between consumers who purchased organic foods and consumers who did not purchase organic foods in the USA. In Table 10, we report the mean shares of preference of US and Norwegian respondents who have purchased and not purchased organic food during the 12 months before the survey.7 Consistent with LB, the importance of price was rated differently across the organic food purchaser/non-purchaser subgroups. In the case of the USA, organic non-purchasers rated price as the second-most important value, while organic purchasers rated price as the sixth-most important food value. Notably, Norwegian respondents who did not buy organic food, rated price as the third-most important food value, while organic purchasers only rated price as the ninth-most important value. In Table 10, we also report results from t-tests to test whether the preferences for the food values differed between organic purchasers in the USA and Norway and organic non-purchasers in the USA and in Norway (indicated with ‘†’ in the table). Although the mean shares of preferences were statistically different for most of the values both in the case of Norwegian and American organic purchasers and organic non-purchasers, the food values become more similar across the samples when mean shares of preferences between organic purchasers and non-purchasers in the two countries were compared. Generally, organic purchasers gave more importance to naturalness and food values related to sustainability issues, while organic non-purchasers gave more importance to attributes such as appearance and especially price. Specifically, price was rated as one of the least important attributes by organic purchasers, while one of the most important attributes by the organic non-purchasers both in the USA and in Norway. 4. Conclusions To the best of our knowledge, our study is the first attempt to estimate consumers’ preferences for food values in a multi-country setting. We used a BWS approach in order to compare consumers’ preferences for food values in the USA and Norway. We included 12 food values: naturalness, taste, price, safety, convenience, nutrition, novelty, origin, fairness, appearance, environmental impact and animal welfare. Our results show that there were large similarities in preferences for food values among US and Norwegian respondents. For instance, safety was clearly the most important value in both countries. Similarly, taste was rated as the third-most important value and convenience and novelty were rated as the two least important values in both countries. There were also notable differences in the evaluation of the importance of price. Price was considered to be the second-most important value by US respondents but only the sixth-most important value by the Norwegian respondents. In addition, naturalness was rated as the second-most important value by Norwegian respondents but rated only as the fifth-most important value by US respondents. This difference is in line with the existing literature that shows European consumers are generally willing to pay a higher price for non-GM or hormone free foods, as compared with US consumers (Chern et al. 2002; Lusk, Roosen and Fox, 2003). Specifically, Norwegian respondents gave more importance to food values related to ethical aspects of food production such as fairness and animal welfare; however, it is important to note that animal welfare was rated among the six-most important food values in both countries. This finding confirms that the addition of animal welfare to LB’s set of attributes is important in assessing food values. This result also suggests that animal welfare labelling could be a potential strategy for the marketing of food products both in the USA and in Europe. For example, given that origin is rated less important than animal welfare, this might suggest that imported meat and dairy products might have a better market share if produced respecting the welfare of the animals. On the other hand, novelty, i.e. the other new food value that we have introduced besides animal welfare, was the least important value for both samples. We included novelty in order to capture the value related to enthusiasm, excitement for new products, but this outcome might suggest that novelty in food products may be associated with food neophobia, which could make individuals reluctant towards food products they do not know (Lähteenmäki and Arvola, 2001; Camarena, Sanjuán and Philippidis, 2011; Mielby et al., 2012). Nutrition was ranked as one of the most important food values in both countries, which should encourage policy makers in Europe to use regulated nutritional labelling on food products. Furthermore, origin was rated as one of the least important values in both countries. This result is at odds with the existing literature that shows that consumers both in the USA and Europe are usually willing to pay a price premium for locally or nationally produced food products, even over other attributes such as organic production, fair trade or low carbon emission (Basu and Hicks, 2008; Darby et al. 2008; Hu, Woods and Bastin, 2009; Onozaka and Mcfadden, 2011; Campbell, Mhlanga and Lesschaeve, 2013; de-Magistris and Gracia, 2014; Gracia, Barreiro-Hurlé and Galán, 2014). As LB suggest, the low importance of origin may be explained by the fact that in preference elicitation methods such as conjoint analyses or choice experiments, differences in consumers’ preferences are estimated in a range of specific attributes levels, which might not be the levels that endogenously come to mind for the consumer. As such, the use of specific attribute levels could be a source of bias in revealing individuals’ preferences. In addition, the use of a specific food product might also influence consumers’ perceptions for different food attributes. For example, Scarpa, Philippidis and Spalatro (2005) observed that consumers’ evaluations for locally produced and organic food claims varied according to the product under consideration. In this study, we did not specify any food product or food attribute level. To test the robustness of our findings, more research on comparison of consumers’ preferences for food values across countries and also across different product groups is warranted. On the other hand, regarding the naturalness, consumers who purchase organic food gave more importance to this value, as compared with non-purchasers in both countries. Moreover, our results show that organic purchasers and organic non-purchasers in both countries rated naturalness similarly; the same applied for price. Hence, these findings suggest that purchasing attitudes might also be an important factor when assessing consumers’ preferences for food values in different countries. Specifically, we observed that while an attitudinal variable, such as buying organic food, substantially affected how individuals rated the 12 food attributes, socio-demographic information, including the belonging to a certain country, did not have such a relevant impact on respondents’ preferences for the food values. These results might then confirm the intuition of LB in defining these more abstract food attributes as more stable, intrinsic meta-preferences which drive consumers’ food choices. As such, our results suggest that food values are an important factor in explaining behavioural reasoning in food consumption. Finally, our findings have important implications for food marketing and policy. For example, since safety is the top food value in both the USA and Norway, food producers need to constantly be cognisant of potential food safety incidents in their products. The recent insecticide contamination in eggs in Europe is a prime example of how incidents like these can significantly affect the livelihood of food producers and erode the confidence of consumers. The implications for food safety policy are equally compelling obviously given the immense importance that consumers are putting on the safety of the products they consume. In terms of the food marketing benefits, the relative rankings of the food values in our study can be useful and informative in terms of the new products that could be developed by the food industry and also in terms of product differentiation using food labels, particularly for the credence food values that are valued more by consumers (e.g. nutrition and animal welfare). Our results can also potentially be used for ongoing trade negotiations between Europe and the USA. The similarities in food values among consumers in both countries are large, with a clear emphasis on safety, suggesting that consumers in both countries highly value the safety of food products, both domestically produced and imported products. Furthermore, price, taste, nutrition, naturalness and animal welfare are considered to be the five-most important food values after safety in both countries. Consumers in both the USA and Europe would benefit from increased trade in agricultural products that could potentially lower prices and increase the variation in the attributes of food, as long as the products are safe and adequately labelled for the key attributes related to nutrition, naturalness and animal welfare. Supplementary data Supplementary data are available at European Review of Agricultural Economics online. Funding This work was supported by the Norwegian Research Council (grant # 233800/E50) and partially supported by the National Research Foundation of Korea (Grant #NRF-2014S1A3A2044459). Acknowledgements A special thank goes to Jayson Lusk for his support in the design of the Best Worst Scaling and in the models estimation. The authors also want to thank Tiziana de-Magistris, Diana Danforth and Vincenzina Caputo for the suggestions and support in the data analysis. Finally, the authors want to thank the anonymous reviewers for their insightful comments. Any remaining errors are the responsibility of the authors. Footnotes 1 We use the term ‘food values’ in order to be consistent with the terminology used by LB. However, it is important to point out as suggested to us by a reviewer that many of these food values are actually ‘food quality attributes’ and not higher constructs such as ‘values’, which are cognitive representation of concepts of beliefs (Schwartz and Bilsky, 1987). 2 It is difficult to find comparable statistics for households’ income distribution in USA and Norway. However, according to the OECD database (OECD, 2016a), the GDP per capita calculated at 2011 purchasing power parities was 172 and 138 in Norway and in the USA, respectively, as compared with an OECD average of 100. For household consumption expenditures, the numbers were 121 for Norway and 146 for the USA. The Gini index in 2011 was 0.25 for Norway and 0.39 for the USA (OECD, 2016b). 3 For more information: http://ipsos-mmi.no/ 4 We used 1000 Halton draws for the simulation. 5 In both samples, most of the estimates in the Cholesky matrix were statistically significant, indicating that correlation across the parameters exists across both models (results are available upon request). 6 These posterior estimates are precisely the means of the parameter distributions, which are conditional on the actual choices of each respondent. These estimates might not be the same as the actual respondent’s coefficients (Train, 2003); however, when respondents face several choice scenarios, the difference between the values can be very small. 7 In our survey, we asked respondents: ‘Have you purposely purchased organic foods during the last 12 months?’ We also gave them the opportunity to answer: ‘I do not know.’ In defining the ‘organic food purchasers’ and the ‘organic food non-purchasers’ subgroups, we did not include the observations of respondents who replied: ‘I do not know.’ References Adamowicz, W. L. and Swait, J. D. ( 2012). Are food choices really habitual? Integrating habits, variety-seeking, and compensatory choice in a utility-maximizing framework. American Journal of Agricultural Economics  95( 1): 17– 41. Google Scholar CrossRef Search ADS   Alfnes, F. ( 2004). Stated preferences for imported and hormone-treated beef: application of a mixed logit model. European Review of Agriculture Economics  31( 1): 19– 37. Google Scholar CrossRef Search ADS   Alfnes, F. and Rickertsen, K. ( 2003). European consumers’ willingness to pay for U.S. beef in experimental auction markets. American Journal of Agricultural Economics  85( 2): 396– 405. Google Scholar CrossRef Search ADS   Almli, V. L., Verbeke, W., Vanhonacker, F., Næs, T. and Hersleth, M. ( 2011). General image and attribute perceptions of traditional food in six European countries. Food Quality and Preference  22( 1): 129– 138. Google Scholar CrossRef Search ADS   Aprile, M. C., Caputo, V. and Nayga, R. M.Jr. ( 2012). Consumers’ valuation of food quality labels: the case of the European geographic indication and organic farming labels. International Journal of Consumer Studies  36( 2): 158– 165. Google Scholar CrossRef Search ADS   Auger, P., Devinney, T. M. and Louviere, J. J. ( 2007). Using best-worst scaling methodology to investigate consumer ethical beliefs across countries. Journal of Business Ethics  70( 3): 299– 326. Google Scholar CrossRef Search ADS   Barber, S. L. and Gertler, P. J. ( 2009). Empowering women to obtain high quality care: evidence from an evaluation of Mexico’s conditional cash transfer programme. Health Policy and Planning  24( 1): 18– 25. Google Scholar CrossRef Search ADS PubMed  Basu, A. K. and Hicks, R. L. ( 2008). Label performance and the willingness to pay for Fair Trade coffee: a cross-national perspective. International Journal of Consumer Studies  32( 5): 470– 478. Google Scholar CrossRef Search ADS   Baumgartner, H. and Steenkamp, J. B. E. ( 2001). Response styles in marketing research: a cross-national investigation. Journal of Marketing Research  38( 2): 143– 156. Google Scholar CrossRef Search ADS   Bazzani, C. and Canavari, M. ( 2013). Alternative agri-food networks and short food supply chains: a review of the literature. Economia Agro-Alimentare  2: 11– 34. Google Scholar CrossRef Search ADS   Bech-Larsen, T. and Grunert, K. G. ( 2003). The perceived healthiness of functional foods. Appetite  40( 1): 9– 14. Google Scholar CrossRef Search ADS PubMed  Beck, M. J., Rose, J. M. and Greaves, S. P. ( 2017). I can’t believe your attitude: a joint estimation of best worst attitudes and electric vehicle choice. Transportation  44( 4): 753– 772. Google Scholar CrossRef Search ADS   Bonsmann, S. S. G. and Wills, J. M. ( 2012). Nutrition labeling to prevent obesity: reviewing the evidence from Europe. Current Obesity Reports  1( 3): 134– 140. Google Scholar CrossRef Search ADS PubMed  Budsjettnemda for jordbruket. ( 2015). Resultatkontroll for gjennomføringen av landbrukspolitikken (Result control for the agricultural policy). Report from the Agricultural Budget Commision, Oslo. Camarena, D. M., Sanjuán, A. I. and Philippidis, G. ( 2011). Influence of ethnocentrism and neo-phobia on ethnic food consumption in Spain. Appetite  57( 1): 121– 130. Google Scholar CrossRef Search ADS PubMed  Campbell, B. L., Mhlanga, S. and Lesschaeve, I. ( 2013). Perception versus reality: Canadian consumer views of local and organic. Canadian Journal of Agricultural Economics  61( 4): 531– 558. Google Scholar CrossRef Search ADS   Carlsson, F., Frykblom, P. and Lagerkvist, C. J. ( 2007). Consumer willingness to pay for farm animal welfare: mobile abattoirs versus transportation to slaughter. European Review of Agricultural Economics  34( 3): 321– 344. Google Scholar CrossRef Search ADS   Chern, W. S., Rickertsen, K., Tsuboi, N. and Fu, T. T. ( 2002). Consumer acceptance and willingness to pay for genetically modified vegetable oil and salmon: a multiple-country assessment. AgBioForum  5( 3): 105– 112. Cohen, E. ( 2009). Applying best‐worst scaling to wine marketing. International Journal of Wine Business Research  21( 1): 8– 23. Google Scholar CrossRef Search ADS   Czarnezki, J. J. ( 2011). Future of food eco-labelling: organic, carbon footprint, and environmental life-cycle analysis. Stanford Environmental Law Journal  30( 3): 3– 50. Darby, K., Batte, M. T., Ernst, S. and Roe, B. ( 2008). Decomposing local: a conjoint analysis of locally produced foods. American Journal of Agricultural Economics  90( 2): 476– 486. Google Scholar CrossRef Search ADS   de-Magistris, T. and Gracia, A. ( 2014). Do consumers care about organic and distance labels? An empirical analysis in Spain. International Journal of Consumer Studies  38( 6): 660– 669. Google Scholar CrossRef Search ADS   de-Magistris, T., Gracia, A. and Albisu, L. M. ( 2014). Wine consumers’ preferences in Spain: an analysis using the best-worst scaling approach. Spanish Journal of Agricultural Research  12( 3): 529– 541. Google Scholar CrossRef Search ADS   Dekhili, S., Sirieix, L. and Cohen, E. ( 2011). How consumers choose olive oil: the importance of origin cues. Food Quality and Preference  22( 8): 757– 762. Google Scholar CrossRef Search ADS   Delwaide, A. C., Nalley, L. L., Dixon, B. L., Danforth, D. M., Nayga, R. M., Van Loo, E. J. and Verbeke, W. ( 2015). Revisiting GMOs: are there differences in European consumers’ acceptance and valuation for cisgenically vs transgenically bred rice? PLoS ONE  10( 5): 1– 16. Google Scholar CrossRef Search ADS   Drichoutis, A. C., Lazaridis, P. and Nayga, R. M.Jr. ( 2006). Consumers’ use of nutritional labels: a review of research studies and issues. Academy of Marketing Science Review  9: 1– 25. Erdem, S., Rigby, D. and Wossink, A. ( 2012). Using best–worst scaling to explore perceptions of relative responsibility for ensuring food safety. Food Policy  37( 6): 661– 670. Google Scholar CrossRef Search ADS   Finn, A. and Louviere, J. J. ( 1992). Determining the appropriate response to evidence of public concern: the case of food safety. Journal of Public Policy and Marketing  11( 2): 12– 25. Flynn, T. and Marley, T. ( 2014). Best-worst scaling: theory and methods. In: S. Hess and A. Daly (eds), Handbook of Choice Modelling . Cheltenham, UK: Edward Elgar, 178– 201. Frewer, L. J., Risvik, E. and Schifferstein, H. (eds) ( 2013). Food, People and Society: A European Perspective of Consumers’ Food Choices . Berlin: Springer. Getz, C. and Shreck, A. ( 2006). What organic and fair trade labels do not tell us: towards a place-based understanding of certification. International Journal of Consumer Studies  30: 490– 501. Google Scholar CrossRef Search ADS   Golden, J. S., Vermeer, D., Clemen, B., Noyes, C. and Akella, A. ( 2010). An overview of ecolabels and sustainability certifications in the global marketplace. Corporate Sustainability Initiative – Nicholas Institute for Environmental Policy Solutions at Duke University  1( 10): 1– 99. Gracia, A., Barreiro-Hurlé, J. and Galán, B. L. ( 2014). Are local and organic claims complements or substitutes? A consumer preferences study for eggs. Journal of Agricultural Economics  65( 1): 49– 67. Google Scholar CrossRef Search ADS   Grunert, K. G., Hieke, S. and Wills, J. ( 2014). Sustainability labels on food products: consumer motivation, understanding and use. Food Policy  44: 177– 189. Google Scholar CrossRef Search ADS   Guerrero, L., Guàrdia, M. D., Xicola, J., Verbeke, W., Vanhonacker, F., Zakowska-Biemans, S. and Scalvedi, M. L. ( 2009). Consumer-driven definition of traditional food products and innovation in traditional foods. A qualitative cross-cultural study. Appetite  52( 2): 345– 354. Google Scholar CrossRef Search ADS PubMed  Hamada, N. ( 1973). On the p-rank of the incidence matrix of a balanced or partially balanced incomplete block design and its applications to error correcting codes. Hiroshima Mathematical Journal  3( 1): 153– 226. Hein, K. A., Jaeger, S. R., Carr, B. T. and Delahunty, C. M. ( 2008). Comparison of five common acceptance and preference methods. Food Quality and Preferences  19: 651– 661. Google Scholar CrossRef Search ADS   Hu, W., Woods, T. and Bastin, S. ( 2009). Consumer acceptance and willingness to pay for blueberry products with nonconventional attributes. Journal of Agricultural and Applied Economics  41( 1): 47– 60. Google Scholar CrossRef Search ADS   Ilbery, B., Morris, C., Buller, H., Maye, D. and Kneafsey, M. ( 2005). Product, process and place: an examination of food marketing and labelling schemes in Europe and North America. European Urban and Regional Studies  12( 2): 116– 132. Google Scholar CrossRef Search ADS   Jaeger, S. R., Jørgensen, A. S., Aaslyng, M. D. and Bredie, W. L. P. ( 2008). Best–worst scaling: an introduction and initial comparison with monadic rating for preference elicitation with food products. Food Quality and Preference  19( 6): 579– 588. Google Scholar CrossRef Search ADS   Kuhfeld, W. F. ( 2005). Experimental Design, Efficiency, Coding, and Choice Designs. Marketing Research Methods in SAS: Experimental Design, Choice, Conjoint, and Graphical Techniques . Iowa State University, IA: SAS Institute Inc., 53– 241. Lagerkvist, C. J. ( 2013). Consumer preferences for food labelling attributes: Comparing direct ranking and best–worst scaling for measurement of attribute importance, preference intensity and attribute dominance. Food Quality and Preference  29( 2): 77– 88. Google Scholar CrossRef Search ADS   Lagerkvist, C. J., Okello, J. and Karanja, N. ( 2012). Anchored vs. relative best–worst scaling and latent class vs. hierarchical Bayesian analysis of best–worst choice data: investigating the importance of food quality attributes in a developing country. Food Quality and Preference  25( 1): 29– 40. Google Scholar CrossRef Search ADS   Lancsar, E., Louviere, J., Donaldson, C., Currie, G. and Burgess, L. ( 2013). Best–worst discrete choice experiments in health: methods and an application. Social Science & Medicine  76( 1): 74– 82. Google Scholar CrossRef Search ADS   Lancaster, K. J. ( 1966). A new approach to consumer theory. The Journal of Political Economy  74( 2): 132– 157. Google Scholar CrossRef Search ADS   Lee, S. H., Han, D. B., Caputo, V. and Nayga, R. M. ( 2015). Consumers’ valuation for a reduced salt product: a nonhypothetical choice experiment. Canadian Journal of Agricultural Economics  63( 4): 563– 582. Google Scholar CrossRef Search ADS   Lee, J. A., Soutar, G. N. and Louviere, J. ( 2007). Measuring values using best–worst scaling: the LOV example. Psychology & Marketing  24( 12): 1043– 1058. Google Scholar CrossRef Search ADS   Lim, K. H., Hu, W., Maynard, L. J. and Goddard, E. ( 2013). US consumers’ preference and willingness to pay for country‐of‐origin‐labeled beef steak and food safety enhancements. Canadian Journal of Agricultural Economics  61( 1): 93– 118. Google Scholar CrossRef Search ADS   Loose, S. M. and Lockshin, L. ( 2013). Testing the robustness of best worst scaling for cross-national segmentation with different numbers of choice sets. Food Quality and Preference  27( 2): 230– 242. Google Scholar CrossRef Search ADS   Loureiro, M. L. and Umberger, W. J. ( 2007). A choice experiment model for beef: what US consumer responses tell us about relative preferences for food safety, country-of-origin labeling and traceability. Food Policy  32( 4): 496– 514. Google Scholar CrossRef Search ADS   Louviere, J. J., Street, D., Burgess, L., Wasi, N., Islam, T. and Marley, A. A. J. ( 2008). Modeling the choices of individual decision-makers by combining efficient choice experiment designs with extra preference information. Journal of Choice Modelling  1( 1): 128– 164. Google Scholar CrossRef Search ADS   Louviere, J. J., Flynn, T. N. and Marley, A. A. J. ( 2015). Best Worst Scaling: Theory, Methods and Applications . Cambridge, England: Cambridge University Press. Google Scholar CrossRef Search ADS   Louviere, J. J. and Woodworth, G. G. ( 1990). Best–worst scaling: a model for largest difference judgments. Working Paper, Faculty of Business, University of Alberta, Edmonton. Luckstead, J. and Devadoss, S. ( 2016). Impacts of the transatlantic trade and investment partnership on processed food trade under monopolistic competition and firm heterogeneity. American Journal of Agricultural Economics  98( 5): 1389– 1402. Google Scholar CrossRef Search ADS   Lusk, J. L. and Briggeman, B. C. ( 2009). Food values. American Journal of Agricultural Economics  91( 1): 184– 196. Google Scholar CrossRef Search ADS   Lusk, J. L., House, L. O., Jaeger, S. R., Moore, M., Morrow, J. L. and Traill, W. B. ( 2004). Effect of information about benefits of biotechnology on consumer acceptance of genetically modified food: evidence from experimental auctions in the United States, England, and France. European Review of Agricultural Economics  31( 2): 179– 204. Google Scholar CrossRef Search ADS   Lusk, J. L., Roosen, J. and Fox, J. A. ( 2003). Demand for beef from cattle administered growth hormones or fed genetically modified corn: a comparison of consumers in France, Germany, the United Kingdom, and the United States. American Journal of Agricultural Economics  85( 1): 1– 58. Google Scholar CrossRef Search ADS   Lähteenmäki, L. and Arvola, A. ( 2001). Food neophobia and variety seeking – consumer ear or demand for new food products. In: L. J. Frewer, E. Risvik and H. Schifferstein (eds), Food, People and Society . Berlin: Springer, 161– 175. Google Scholar CrossRef Search ADS   Marley, A. A. J. and Louviere, J. J. ( 2005). Some probabilistic models of best, worst, and best-worst choices. Journal of Mathematical Psychology  49( 6): 464– 480. Google Scholar CrossRef Search ADS   Martinez, S., Hand, M., Da Pra, M., Pollack, S., Ralston, K., Smith, T., Vogel, S., Clark, S., Lohr, L., Low, S. and Newman, C. ( 2010). Local Food Systems: Concepts, Impacts, and Issues . Washington, DC: US Department of Agriculture, Economic Research Service. McFadden, D. ( 1974). Conditional logit analysis of qualitative choice behavior. In: P. Zarembka (ed.), Frontiers in Econometrics . New York: Academic Press, 105– 142. Meas, T., Hu, W., Batte, M. T., Woods, T. A. and Ernst, S. ( 2015). Substitutes or complements? Consumer preference for local and organic food attributes. American Journal of Agricultural Economics  97( 4): 1044– 1071. Google Scholar CrossRef Search ADS   Mielby, L. H., Nørgaard, M. K., Edelenbos, M. and Thybo, A. K. ( 2012). Affective response of adolescents toward fruit‐ and vegetable-based snacks and the role of neophobia, gender and age. Journal of Sensory Studies  27( 6): 425– 438. Google Scholar CrossRef Search ADS   Mitchell, L. ( 2001). Impact of consumer demand for animal welfare on global trade. Changing Structure of Global Food Consumption and Trade. Department of Agriculture, Economic Research Service, Market and Trade Economics Division, Agriculture and Trade Report WRS-01-1. Washington, DC: US Department of Agriculture, Economic Research Service. Mitchell, L. ( 2004). US and EU consumption comparisons. In Normile and Leetmaa (eds), US-EU Food and Agriculture Comparisons. ERS, USDA, WRS-04-04, 49–65. Napolitano, F., Braghieri, A., Piasentier, E., Favotto, S., Naspetti, S. and Zanoli, R. ( 2010). Effect of information about organic production on beef liking and consumer willingness to pay. Food Quality and Preference  21( 2): 207– 212. Google Scholar CrossRef Search ADS   Napolitano, F., Pacelli, C., Girolami, A. and Braghieri, A. ( 2008). Effect of information about animal welfare on consumer willingness to pay for yogurt. Journal of Dairy Science  91( 3): 910– 917. Google Scholar CrossRef Search ADS PubMed  Nayga, R. M., Lipinski, D. and Savur, N. ( 1998). Consumers’ use of nutritional labels while food shopping and at home. Journal of Consumer Affairs  32( 1): 106– 120. Google Scholar CrossRef Search ADS   NIBIO. ( 2016). Utsyn over norsk landbruk, tilstand og utviklingstrekk 2016 (Norwegian agriculture: state and trends). NIBIO Bok 2(3), Oslo. Norsk landbruk. ( 2014). http://www.norsklandbruk.no/norsk-landbruk/ni-av-ti-vil-opprettholde-det-norske-jordbruket/. Accessed 22 March 2017. OECD. ( 2010). Agricultural policies in OECD countries: at a glance. http://www.oecd.org/tad/agricultural-policies/agriculturalpoliciesinoecdcountriesataglance.htm. Accessed 18 May 2016. OECD. ( 2016a). OECD.Stat. Expenditures at national prices in national currencies (millions). http://stats.oecd.org/viewhtml.aspx?datasetcode=PPP2011&lang=en. Accessed 26 August 2016. OECD. ( 2016b). OECD.Stat, Income distribution and poverty. http://stats.oecd.org/Index.aspx?DataSetCode=IDD. Accessed 26 August 2016. OECD. ( 2017). Inequality. http://www.oecd.org/social/inequality.htm. Accessed 22 March 2017. Onozaka, Y. and Mcfadden, D. T. ( 2011). Does local labeling complement or compete with other sustainable labels? A conjoint analysis of direct and joint values for fresh produce claim. American Journal of Agricultural Economics  93( 3): 693– 706. Google Scholar CrossRef Search ADS   Orme, B. K. ( 2005). Accuracy of HB estimation in MaxDiff experiments. Sequim, WA: Sawtooth Software. 98382(360). www.sawtoothsoftware.com. Accessed 13 April 2016. Pieniak, Z., Verbeke, W., Vanhonacker, F., Guerrero, L. and Hersleth, M. ( 2009). Association between traditional food consumption and motives for food choice in six European countries. Appetite  53( 1): 101– 108. Google Scholar CrossRef Search ADS PubMed  Progressive Dairyman. ( 2016). 2016 U.S. dairy statistics. http://www.progressivedairy.com/site/stats/us-dairy-stats. Accessed 13 May 2017. Rickertsen, K., Gustavsen, G. W. and Nayga, R. M.Jr. ( 2017). Consumer willingness to pay for genetically modified vegetable oil and salmon in the United States and Norway. AgBioForum  20( 2): 1– 11. Roininen, K., Lähteenmäki, L. and Tuorila, H. ( 1999). Quantification of consumer attitudes to health and hedonic characteristics of foods. Appetite  33( 1): 71– 88. Google Scholar CrossRef Search ADS PubMed  Rose, J. M. ( 2014). Interpreting discrete choice models based on best–worst data: a matter of framing. Transportation Research Board, Annual Meeting, 12–16 January, Washington DC. Rozin, P., Levine, E. and Stoess, C. ( 1991). Chocolate craving and liking. Appetite  17( 3): 199– 212. Google Scholar CrossRef Search ADS PubMed  Scarpa, R., Notaro, S., Louviere, J. and Raffaelli, R. ( 2011). Exploring scale effects of best/worst rank ordered choice data to estimate benefits of tourism in alpine grazing commons. American Journal of Agricultural Economics  93( 3): 809– 824. Google Scholar CrossRef Search ADS   Scarpa, R., Philippidis, G. and Spalatro, F. ( 2005). Product-country images and preference heterogeneity for Mediterranean food products: a discrete choice framework. Agribusiness  21( 3): 329– 349. Google Scholar CrossRef Search ADS   Schwartz, S. H. and Bilsky, W. ( 1987). Toward a universal psychological structure of human values. Journal of Personality and Social Psychology  53: 550– 562. Google Scholar CrossRef Search ADS   Schwartz, S. H. ( 1994). Are there universal aspects in the structure and contents of human values? Journal of Social Issues  50( 4): 19– 45. Google Scholar CrossRef Search ADS   Siro, I., Kapolna, E., Kapolna, B. T. A. and Lugasi, A. ( 2008). Functional food. Product development, marketing and consumer acceptance – a review. Appetite  51( 3): 456– 467. Google Scholar CrossRef Search ADS PubMed  Soederberg Miller, L. M. and Cassady, D. L. ( 2015). The effects of nutrition knowledge on food label use. A review of the literature. Appetite  92: 207– 216. Google Scholar CrossRef Search ADS PubMed  Statistica. ( 2017). ‘Average size of farms’. https://www.statista.com/statistics/196106/average-size-of-farms-in-the-us-since-2000/. Accessed 6 May 2017. Statistics Norway. ( 2017). ‘Population’. https://www.ssb.no/en/befolkning. Accessed 25 April 2017. Street, D. J. and Street, A. P. ( 1996). Partially Balanced Incomplete Block Design. In: C. J. Colbourn and J. H. Dinitz (eds), Encyclopedia of Biostatistics . Boca Raton, FL: CRC Press. Ter Hofstede, F., Steenkamp, J.-B. E. M. and Wedel, M. ( 1999). International market segmentation based on consumer-product relations. Journal of Marketing Research  36( 1): 1– 17. Google Scholar CrossRef Search ADS   Thomson, D. M. H., Crocker, C. and Marketo, C. G. ( 2010). Linking sensory characteristics to emotions: an example using dark chocolate. Food Quality and Preference  21( 8): 1117– 1125. Google Scholar CrossRef Search ADS   Thurstone, L. L. ( 1927). A law of comparative judgment. Psychological Review  34( 4): 273– 286. Google Scholar CrossRef Search ADS   Train, K. ( 2003). Discrete Choice Methods with Simulation . Cambridge, UK: Cambridge University Press. Google Scholar CrossRef Search ADS   United Stated Census Bureau ( 2017). ‘Population’. www.census.gov. Accessed 26 April 2017. Van Trijp, H. C. and Steenkamp, J. B. E. ( 1992). Consumers’ variety seeking tendency with respect to foods: measurement and managerial implications. European Review of Agricultural Economics  19( 2): 181– 195. Google Scholar CrossRef Search ADS   Vandemoortele, T. and Deconinck, K. ( 2014). When are private standards more stringent than public standards? American Journal of Agricultural Economics  96( 1): 154– 171. Google Scholar CrossRef Search ADS   Verbeke, W., Guerrero, L., Almli, V. L., Vanhonacker, F. and Hersleth, M. ( 2016). European consumer definition and perception of traditional foods. In: K. Kristbergsson and J. Oliveira (eds), Traditional Foods: General and Consumer Aspects . Boston, MA: Springer, 3– 16. Google Scholar CrossRef Search ADS   Author notes Review coordinated by Ada Wossink © Oxford University Press and Foundation for the European Review of Agricultural Economics 2018; all rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Review of Agricultural Economics Oxford University Press

A comparative study of food values between the United States and Norway

Loading next page...
 
/lp/ou_press/a-comparative-study-of-food-values-between-the-united-states-and-5uvpRTorqI
Publisher
Oxford University Press
Copyright
© Oxford University Press and Foundation for the European Review of Agricultural Economics 2018; all rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0165-1587
eISSN
1464-3618
D.O.I.
10.1093/erae/jbx033
Publisher site
See Article on Publisher Site

Abstract

Abstract We compare the food values in the USA and Norway using the best–worst scaling approach. The food values examined are aimed at capturing the main issues related to food consumption such as naturalness, taste, price, safety, convenience, nutrition, novelty, origin, fairness, appearance, environmental impact and animal welfare. Results show that respondents in both countries have mostly similar food values, with safety being the most important value; while convenience and novelty are the least important values. Specifically, US respondents consider price more important and naturalness less important than Norwegian respondents. 1. Introduction The food systems in Europe and the United States significantly differ in terms of agricultural production practices, agricultural policy and marketing of foods. For example, many discussions have been raised regarding the use of genetically modified organisms (GMOs) and growth hormones in food production since European regulations on these food production issues are notably stricter than in the USA (Chern et al., 2002; Lusk, Roosen and Fox, 2003; Alfnes, 2004; Loureiro and Umberger, 2007; Delwaide et al., 2015). At the same time, food consumption trends in the USA can affect food patterns in Europe and vice versa (Mitchell, 2004), e.g. the local food movement. The development of different forms of Alternative Agri-Food Networks (AAFNs) such as farmers’ markets or Communities Supported Agriculture (CSA), for instance, first occurred in USA in the 1970s and 1980s but these have only recently become more popular in Europe (Martinez et al., 2010; Bazzani and Canavari, 2013). In addition, the adoption of nutrition food labelling is currently a widely discussed topic both in US and European food systems; but while nutritional labels have been regulated by the Food and Drug Administration (FDA) in USA since the early 90s, the European Union (EU) has only very recently introduced uniform or harmonised nutritional food labelling regulations (Nayga, Lipinski and Savur, 1998; Bonsmann and Wills, 2012; Soederberg Miller and Cassady, 2015). Although the presence of ethical and environmental food labels has consistently grown both in Europe and in the USA, the development of sustainable food labels occurred more recently in the USA in comparison to the European food system (Getz and Shreck, 2006; Golden et al., 2010; Grunert, Hieke and Wills, 2014; Ilbery et al., 2005; Louriero and Lotade, 2005). Moreover, the European food system is characterised by the presence of labels indicating specific regions of origin such as protected designation of origin (PDO), protected geographic origin (PGO) or country of origin (COOL) (Loureiro and Umberger, 2007; Aprile, Caputo and Nayga, 2012). Another notable difference is that the US food market is generally less developed in terms of traceability systems than the European food market, although US consumers have increasingly called for foods labelled as produced in the USA (Loureiro and Umberger, 2007; Lim et al., 2013). In order to capture these similarities and differences across European and US food systems, several studies have explored European and US consumers’ attitudes towards food claims, aiming at the development of potential international marketing strategies and policies (Roininen, Lähteenmäki and Tuorila, 1999; Chern et al. 2002; Bech-Larsen and Grunert, 2003; Lusk, Roosen and Fox, 2003; Lusk et al.2004; Loureiro and Umberger, 2007). The existing literature investigating consumers’ food attitudes in Europe and the USA has mainly focused on consumers’ evaluations of food safety claims and their attitudes towards genetically modified (GM) products (Chern et al. 2002; Lusk et al., 2004). The findings in these studies generally suggest that people in Europe are less willing to accept GM foods. For example, Chern et al. (2002) showed that Norwegian consumers were more willing to pay for non-GM vegetable oil and salmon than US consumers. Similarly, Alfnes and Rickertsen (2003), Lusk, Roosen and Fox (2003) and Alfnes (2004) showed that European consumers were willing to pay a higher price for beef from cattle that had not been administered growth hormones and Lusk, Roosen and Fox (2003) showed a higher willingness to pay for cattle that had not been fed with GM corn among Europeans when compared with US consumers. More recently, Rickertsen, Gustavsen and Nayga (2017) assessed consumers’ willingness to pay for GM soybean oil, farmed salmon fed with GM soy and GM salmon. Interestingly, their results suggest a large similarity in WTP in Norway and the USA and across the three products. Additionally, Rozin, Levine and Stoess (1991) investigated factors affecting individuals’ preferences for different kinds of chocolate bars, using students from universities in the USA, Belgium and France as a subject pool. They observed that US students were more health-oriented in making their choices, while Belgian and French students were more pleasure-oriented. Bech-Larsen and Grunert (2003) also showed that US consumers were more willing to buy functional foods than Danish and Finnish consumers, mainly because of health-related motivations. Finally, Basu and Hicks (2008), investigated US and German consumers’ evaluations for fair-trade coffee using a choice experiment approach and found that German respondents were more inequality averse than US consumers. Generally, studies investigating consumers’ preferences in the USA and Europe have limited their analyses to the assessment of consumers’ evaluations for specific food attributes such as GM production, nutritional content, use of growth hormones or sustainability issues. Lusk and Briggeman (2009) (henceforth LB) claimed that individuals’ food choices may be explained by their preferences for more abstract food quality attributes1 which LB identified as intermediary values, that ‘relate specifically to people’s food choices’ (Lusk and Briggeman, 2009: 186). These so-called ‘food values’ can be considered as more stable than consumers’ preferences for a specific set of food attributes on specific food products. According to LB, food values can explain individuals’ food choices across a variety of food products and do not depend on the specific context under investigation. However, to the best of our knowledge, no study has compared the food values between the USA and Europe, which is the aim of our study. In this study, we identify a set of 12 food values, which differ slightly from the set that was used by LB. These values are aimed at capturing the main issues related to food consumption patterns such as naturalness, taste, price, safety, convenience, nutrition, novelty, origin, fairness, appearance, environmental impact and animal welfare. In order to measure individuals’ preferences for food values, we implement a best–worst scaling (BWS) approach. The choice of this approach has been determined by the fact that Lee, Soutar and Louviere (2007) observed that the use of BWS provided better outcomes than other rating methods in measuring human values. In addition, BWS is particularly appropriate in cross-country comparisons, since the use of other forms of rating scales might lead to scalar inequivalence, which is generally caused by divergences in lexicon and response styles across different cultures (Baumgartner and Steenkamp, 2001; Auger, Devinney and Louviere, 2007; Jaeger et al., 2008; Ter Hofstede, Steenkamp and Wedel, 1999; Loose and Lockshin, 2013). For example, Mueller Loose and Lockshin (2013) and Dekhili, Sirieix and Cohen (2011) showed that the BWS method worked well to explore differences across countries in rating a set of attributes on wine and olive oil products, respectively. A potential limitation of BWS could be the lack of complete transitivity in attribute importance and therefore of consistency in dominance relations of attribute importance ranking. However, Lagerkvist (2013), in a study investigating Swedish consumers’ preferences for food quality attributes on beef, explored these issues using different rating methods such as BWS and direct ranking (DR) and showed that estimates at the aggregate level from BWS were more consistent than the estimates from DR both in terms of preference relations and of dominance ordering of attribute importance. We specifically compare the food values in Norway and the USA for several reasons. The Norwegian regulations on the use of biotechnology are quite restrictive and so one would expect more resistance against production methods based on modern biotechnology. In addition, the Norwegian food environment is very different from the US food environment. In contrast to the USA, Norwegian agriculture is dominated by small scale farming. The average farm size in Norway was 23.4 hectares in 2015 according to the Norwegian Institute of Bioeconomy Research (NIBIO) (NIBIO, 2016: 24), and the average dairy herd was 25 dairy cows in 2014 (Budsjettnemda for jordbruket, 2015), while in the USA the average farm size and the average dairy herd were 438 acres and 2,017 dairy cows, respectively, in 2014 (Progressive Dairyman, 2016; Statistica, 2017). Furthermore, the key tenets of the Norwegian agricultural policy are different from those in the USA. There are four main objectives of the Norwegian agricultural and food policy: (i) food security (with emphasis on having high domestic production of agricultural products, especially meat and dairy products), (ii) agricultural production in all parts of the country, (iii) increased value of the agricultural products and (iv) sustainable agriculture (for example through the target that 15 per cent of the production and consumption should be organic before 2020) (NIBIO, 2016: 12, 49). These objectives are supported by one of the highest levels of agricultural subsidies in the world. Producer support estimates were 61 per cent of gross farm receipts for the period of 2007–2009 as compared with only 9 per cent in the USA (OECD, 2010). Moreover, Norway has very high-import tariffs for products such as dairy and meats and, consequently, very little trade with these products (NIBIO, 2016: 54). Opinion polls also show a strong public support for the current state. In a recent poll, 90 per cent of the respondents wanted to maintain Norwegian agriculture on at least the present level (Norsk Landbruk, 2014). Finally, while the average per capita income, measured at purchasing power parities, is quite similar in the two countries, Norway is characterised by a more equal distribution of income. According to the OECD (2017), Norway was the second-most equal OECD country after Iceland in 2014 while the USA was the third-most unequal country. We believe that the differences between food systems in Norway and the USA make these two countries an interesting context to compare food values. Our hypothesis is that differences in agricultural systems might be related to differences in individuals’ food values. To illustrate, the adoption of high-agricultural subsidies, the enhancement of domestic and sustainable food production in Norway might be respectively related to the importance that Norwegian people give to food values such as fairness, origin and environmental impact. Moreover, even though food prices are relatively much higher in Norway than in the USA, the high degree of income equality in Norway may result in less emphasis on food prices and higher emphasis on fairness.2 To sum up, this study advances the literature in two important ways: (i) we adopt the concept of food values and the set of items used by LB to identify which food values are most important among US and Norwegian consumers and (ii) we compare consumers’ preferences for food values in a multi-country setting, considering credence (e.g. food safety and origin) as well as experience attributes (e.g. taste). Results from this study are of value to food marketers and policy makers for two main reasons. First, the comparison of consumers’ preferences between a European country and the USA is currently of particular interest since Europe and the USA are key trading partners and results from this study would help future trade negotiations (Luckstead and Devadoss, 2016). Second, while we do not want to lessen the contribution to the literature of previous studies comparing different countries’ consumer preferences for food attributes on specific food products, the results from our study could be applied to various commodities, and, therefore, could be used as a guide in the development and implementation of marketing strategies and food policies for a broad range of food products. To illustrate, if our results show, for example, a high preference for the food value ‘safety’ both in the USA and Norway, then this could encourage the support of policies aimed at increasing the traceability of food products, while a high rating for ‘naturalness’ could support the production and trade of foods produced without the use of modern technologies or pesticides, no matter what the product under consideration is. 2. Materials and methods This section is dedicated to the description of (i) the data collection, (ii) experimental design, i.e. selection of the food values and implementation of the BWS and (iii) applied econometric approach. 2.1. Data collection Data were collected from an online survey conducted between October and November of 2015 in Norway and the USA. More than 1,000 respondents in each country (1,037 in Norway and 1,025 in the USA) took part in the survey. Respondents were randomly recruited across regions and urban/non-urban areas in both countries by a professional market research agency called Ipsos.3 Respondents were invited to participate in an internet survey and were asked about the aspects they considered more or less important when buying food products. They were assured that any given information was anonymous and that they could quit the survey whenever they wanted to. The survey also contained questions about attitudes towards food claims. The selected samples in Norway and the USA were relatively representative of the national populations in terms of socio-demographic information. In Table 1, we report information related to the distribution of demographic and socio-economic variables in the two samples and of the US and Norwegian populations, respectively. Table 1. Demographic and socio-economic distribution in the USA and Norway   USA    Norway  Sample  Population    Sample  Population  Female (%)  51  51    50  50  Age (years)  40  39    53  39  Education (%)             Less than high school  3  17    3  27   High school  46  55    34  40   University degree  38  18    43  23   Post-university degree  13  10    20  10  Marital status (%)             Married  48  50    54  35   Cohabitant  7  NA    15  NA   Never been married  32  31    16  51   Separated or divorced  12  12    11  9   Widow or widower  1  7    4  5  Number of children in household (%)   No children  55  58    70  72   One child  19  18    11  13   Two children  16  16    12  11   More than two  10  8    7  4  Income (gross annual income) (%)   Less than $ 15,000  12  $53,718 (median)  Less than $12,500  1a  $61,387 (median)   $15,000–29,000  17    $12,500–24,900  2     $30,000–44,000  14    $25,000–37,400  3     $45,000–59,000  13    $37,500–49,900  7     $60,000–74,000  12    $50,000–62,400  10     $75,000–89,000  11    $62,500–74,900  12     $90,000–119,000  10    $75,000–87,400  30     $120,000–49,000  6    $87,500–99,900  17     $150,000 or more  5    $100,000 or more  18    Rural area (%)b  18  19    25  19    USA    Norway  Sample  Population    Sample  Population  Female (%)  51  51    50  50  Age (years)  40  39    53  39  Education (%)             Less than high school  3  17    3  27   High school  46  55    34  40   University degree  38  18    43  23   Post-university degree  13  10    20  10  Marital status (%)             Married  48  50    54  35   Cohabitant  7  NA    15  NA   Never been married  32  31    16  51   Separated or divorced  12  12    11  9   Widow or widower  1  7    4  5  Number of children in household (%)   No children  55  58    70  72   One child  19  18    11  13   Two children  16  16    12  11   More than two  10  8    7  4  Income (gross annual income) (%)   Less than $ 15,000  12  $53,718 (median)  Less than $12,500  1a  $61,387 (median)   $15,000–29,000  17    $12,500–24,900  2     $30,000–44,000  14    $25,000–37,400  3     $45,000–59,000  13    $37,500–49,900  7     $60,000–74,000  12    $50,000–62,400  10     $75,000–89,000  11    $62,500–74,900  12     $90,000–119,000  10    $75,000–87,400  30     $120,000–49,000  6    $87,500–99,900  17     $150,000 or more  5    $100,000 or more  18    Rural area (%)b  18  19    25  19  Sources: The data of the Norwegian population were extracted from Statistics Norway (2017) and the data of the US population were extracted from the United Census Bureau (United Stated Census Bureau, 2017). aExchange rate during the survey (15 October 2015) was USD 1 = NOK 8.00, which was used to convert the Norwegian income figures to USD. bThe standard definition of rural area according to ‘Norway Statistics’ is ‘a hub of buildings that is inhabited by less than 200 persons’, while the definition of rural area in the US Census Bureau is an area which is inhabited by less than 2,500 individuals. In our survey, we defined rural area as a settlement with a population lower than 1,000 individuals. Gender distribution was fairly similar in both samples with about 50 per cent and 51 per cent female respondents in the USA and Norway. The average age of the respondents was substantially higher in Norway (53 years) than in the USA (40 years). The average age of the Norwegian sample is higher than the average age of the Norwegian population. Regarding the education level, both samples are more educated than their respective country populations. Norwegian respondents, on average, had a somewhat higher education level than the US sample, and the Norwegian sample was also characterised by a higher percentage of married people (54 per cent) and cohabitants (15 per cent) than the US sample (48 per cent and 7 per cent, respectively). However, the percentage of married individuals in the Norwegian sample was higher than the Norwegian population, while the US sample was characterised by a slightly lower percentage of married people in comparison to the US population. On the other hand, respondents in the USA tended to have more children in the household compared with Norwegian respondents; however, most respondents in both countries indicated having no children in their household (70 per cent for Norway and 55 per cent for the USA), which closely resemble the statistics of the populations in the two countries. Notably, the majority of the respondents in the USA had an annual income equal or below $59,000 (56 per cent), while only 23 per cent of the Norwegian sample had an annual income equal or below $62,400. This is consistent with the median income of the populations in both countries, indicating that the annual median income is higher than in the USA. Importantly though, the income differences are calculated at market exchange rates that vary considerably over time and are quite different from the exchange rates calculated at rates that reflect the purchasing power. Finally, Table 1 shows that Norwegian and US populations have the same percentage of people residing in rural areas (19 per cent). However, the Norwegian sample included a higher percentage of people living in rural areas (28 per cent) than in the US sample (18 per cent). 2.2. Experimental design 2.2.1. Food values As previously mentioned, we followed the work of LB who specified 11 food values (naturalness, safety, environmental impact, origin, fairness, nutrition, taste, appearance, convenience, price and tradition). LB selected these attributes in an attempt to resemble the 10 values identified by Schwartz (1994). LB noted that some values considered by Schwartz, such as achievement and power, might not have a direct relation with food. However, one of the values identified by Schwartz is ‘stimulation’ that could be related to the excitement that ‘novelty’ could present. With the improvement in food technologies and growing globalisation, consumers are continuously offered new food products (Siro et al. 2008; Lee et al., 2015). In addition, a large body of literature shows that variety seeking plays an important role in consumers’ food choices and eating behaviour (Van Trijp and Steenkamp, 1992; Adamowicz and Swait, 2012; Frewer, Risvik and Schifferstein, 2013). Hence, we included ‘novelty’ in our set of food values. Recent literature also shows that consumers are increasingly interested in animal welfare (Carlsson, Frykblom and Lagerkvist, 2007; Napolitano, et al., 2008; Barber and Gertler, 2009). Animal welfare could also be associated with the Schwartz value of ‘universalism’ which resembles individuals’ ‘understanding, appreciation, tolerance, and protection for the welfare of all people and for nature’ (Schwartz, 1994: 22). Hence, we also included ‘animal welfare’ in our set of food values. However, we excluded ‘tradition’ which LB defined as ‘preserving traditional consumption patterns’ due to the growing globalisation of food markets. Indeed, due to increasing ethnic diversity in US and Norwegian populations, tradition is likely to be interpreted differently across respondents. Moreover, studies investigating the meaning of food tradition in six European countries (including Norway) showed that respondents tended to give different interpretations of food tradition depending on the country they belonged to and they especially tended to associate food tradition with different aspects of food consumption such as origin, locality, processing transformation, habits, naturalness, sensory property and familiarity (Guerrero et al., 2009; Pieniak et al., 2009; Almli, et al., 2011; Verbeke et al., 2016). Thus, the inclusion of ‘tradition’ in our set of food values would have been a confounder or would have been overlapping with other food values in our study. The 12 food values incorporated into our study, the food values in LB and the definitions used in the surveys are exhibited in Table 2. Table 2. Food values with descriptions in parentheses Lusk and Briggeman (2009)  This study  Naturalness (extent to which food is produced without modern technologies)  Naturalness (made without modern food technologies like genetic engineering, hormone treatment and food irradiation)  Safety (extent to which consumption of food will not cause illness)  Safety (eating the food will not make you sick)  Environmental impact (effect of food production on the environment)  Environmental impact (effects of food production on the environment)  Origin (where the agricultural commodities were grown)  Origin (whether the food is produced locally, in USA/Norway or abroad)  Fairness (the extent to which all parties involved in the production of the food equally benefit)  Fairness (farmers, processors and retailers get a fair share of the price)  Nutrition (amount and type of fat, protein, vitamins, etc.)  Nutrition (amount and type of fat, protein, etc.)  Taste (extent to which consumption of the food is appealing to the senses)  Taste (the flavour of the food in your mouth)  Appearance (extent to which food looks appealing)  Appearance (the food looks appealing and appetising)  Convenience (ease with which food is cooked and/or consumed)  Convenience (how easy and fast the food is to cook and eat)  Price (the price that is paid for the food)  Price (price you pay for the food)  Tradition (preserving traditional consumption patterns)      Animal welfare (well-being of farm animals)    Novelty (the food is something new that you have not tried before)  Lusk and Briggeman (2009)  This study  Naturalness (extent to which food is produced without modern technologies)  Naturalness (made without modern food technologies like genetic engineering, hormone treatment and food irradiation)  Safety (extent to which consumption of food will not cause illness)  Safety (eating the food will not make you sick)  Environmental impact (effect of food production on the environment)  Environmental impact (effects of food production on the environment)  Origin (where the agricultural commodities were grown)  Origin (whether the food is produced locally, in USA/Norway or abroad)  Fairness (the extent to which all parties involved in the production of the food equally benefit)  Fairness (farmers, processors and retailers get a fair share of the price)  Nutrition (amount and type of fat, protein, vitamins, etc.)  Nutrition (amount and type of fat, protein, etc.)  Taste (extent to which consumption of the food is appealing to the senses)  Taste (the flavour of the food in your mouth)  Appearance (extent to which food looks appealing)  Appearance (the food looks appealing and appetising)  Convenience (ease with which food is cooked and/or consumed)  Convenience (how easy and fast the food is to cook and eat)  Price (the price that is paid for the food)  Price (price you pay for the food)  Tradition (preserving traditional consumption patterns)      Animal welfare (well-being of farm animals)    Novelty (the food is something new that you have not tried before)  The food values include credence, experience and price attributes. Naturalness and safety are considered credence attributes since they are product characteristics that consumers cannot decipher just by looking at the product without any label information. In addition to naturalness and food safety, credence attributes were included that are related to sustainability and ethical issues such as environmental impact, origin, animal welfare and fairness. Finally, nutrition is a credence attribute related to the nutritional content of the food products. On the other hand, taste and appearance are experience attributes. Convenience and novelty can also be considered experience attributes; consumers can personally experience whether a food product is easy or fast to eat, or whether they have never tried a product before. Finally, price is the search attribute that identifies the money individuals pay to buy a food product. LB’s definitions were slightly modified in our study in order to make them more understandable to respondents in Norway and the USA. To illustrate, for naturalness, we indicated that this is food produced without the use of modern food technologies such as genetic engineering, hormone treatment and food irradiation. 2.2.2. Best-worst scaling The BWS approach was developed by Louviere and Woodworth (1990) and first published by Finn and Louviere (1992). It consists of a series of choice sets where respondents are asked to indicate among a (sub)set of attributes or statements which one they prefer the most (or consider the most important) and which one they prefer the least (or consider the least important). This approach has been defined by researchers as an extension of Thurstone’s (1927) paired comparison method in which respondents are asked to choose the best between paired items. Nowadays, BWS is a popular methodology that has been implemented in several research fields such as psychology, marketing and social and environmental sciences (Auger, Devinney and Louviere, 2007; Cohen, 2009; Scarpa et al., 2011; Lancsar et al., 2013). In food consumption literature, BWS has been mainly used for the estimation of consumers’ valuations for product attributes, as well as consumers’ food attitudes (Jaeger et al. 2008; Cohen, 2009; Lusk and Briggeman, 2009; Lagerkvist, Okello and Karanja, 2012; de-Magistris, Gracia and Albisu, 2014). The growing popularity of the BWS method is due to the fact that it provides several advantages over other common rating-based methods such as the Likert scale. In BWS, individuals can respond to the question only in one way, indicating which value is the most important and which one is the least important. This method forces individuals to make choices among values of the scale and does not allow the possibility to give the same value to all the issues in question. Comparatively, in rating scales, individuals might have their own evaluation for the scale values; for example, a three for one person could represent a four for another person, so they might use the scale differently. Finally, using a BWS approach, researchers can construct individual-level scales of preference/importance for each issue under consideration and accurately compare these scales (Cohen, 2009; Hein et al., 2008; Lusk and Briggeman, 2009). In BWS surveys, researchers have the option to use one of the three response mechanisms, which are generally described as BWS cases (Flynn and Marley, 2014; Rose, 2014; Beck, Rose and Greaves, 2017). In Case 1, the respondents are asked to choose the most preferred (most important) and the least preferred (least important) item among a list of items. In Case 2, items are not presented as a whole; rather for each choice set, respondents are asked to make a selection among a list of associated attributes and attribute levels. In Case 3, for each choice set, respondents are asked to select the best and worst from the alternatives which are described by a number of attributes and attribute levels of the items under investigation. In this study, we chose to use the Case 1 mechanism since this is the most appropriate approach for our research goal, i.e. investigating relative preferences for a list of food values (Flynn and Marley, 2014). When designing BWS experiments, researchers have to take into consideration both the potential number of choice sets and the potential number of the items per choice set. A large number of choice sets might induce fatigue to respondents, while a large number of items per choice set might decrease individuals’ attendance to the different attributes (Louviere et al., 2008; Scarpa et al. 2011; de-Magistris, Gracia and Albisu, 2014). For the allocation of the different items across the choice sets, we used a nearly balanced incomplete block design (NBIBD). The balanced incomplete block design (BIBD) is in general one of the most implemented experimental designs in Case 1 BWS surveys (Lee et al., 2007; Auger, Devinney and Louviere, 2007; Cohen, 2009; Flynn and Marley, 2014). This balance is due to each choice set being characterised by an equal number of items, and each item being repeated the same number of times across the choice sets. In addition, the items are orthogonally allocated, meaning that each item is paired with other items an equal number of times across the choice sets. However, researchers might find difficulties in generating a BIBD with a restricted number of choice sets and attributes per choice set. For this reason, different studies have implemented experimental designs where the orthogonality requirement is relaxed, i.e. partially balanced incomplete designs or nearly balanced designs (Erdem et al., 2012; Hamada, 1973; Street and Street, 1996; Orme, 2005; Jaeger et al. 2008; Thomson, Crocker and Marketo, 2010; Lagerkvist, Okello and Karanja, 2012). Our nearly BIBD consists of 12 choice sets, with each of the choice sets containing a subset of four food values. Each food value was repeated four times across the 12 choice sets and each food value was compared with each other 1.09 times, maximising the D-efficiency score (98.71 per cent) to satisfy the orthogonality property (Kuhfeld, 2005). Another important aspect of the nearly balanced incomplete design is that it also helps to minimise the possibility that preferences for values can be unintentionally inferred by features of the design. This way, violations of transitivity and dominance that may be related to the use of BWS can be reduced (Lagerkvist, 2013; Flynn and Marley, 2014). In Figure 1, we report an example of a choice set. Fig. 1. View largeDownload slide Example of a choice set. Fig. 1. View largeDownload slide Example of a choice set. In every choice set, respondents were asked to indicate which one among the four food values they considered the most important and which one they considered the least important when buying food products. If a respondent tried to choose more than one food value as the most or least important, they were told to choose only one value before they could continue to the next choice set in the online survey. 2.3. Econometric analysis Marley and Louviere (2005) describe the different probabilistic models for the best, worst and best-worst choices, by explaining theoretically the processes that respondents might follow in providing best and worst observations in BWS. These models are distinguished in three overlapping classes: random ranking and random utility models, joint and sequential and ratio scale models. Finally, in their paper, Marley and Louviere (2005) make a larger distinction between sequential and maximum difference (maxdiff) models (Flynn and Marley, 2014). A sequential model assumes that respondents make best and worst choices in a particular order (e.g. best first and then worst), while the maxdiff model, which is a well-established probabilistic model that was introduced by the pioneering work of Finn and Louviere (1992), assumes that respondents simultaneously choose the pair of items that maximises the difference between the best and the worst choices. In this study, we apply the maxdiff model for two main reasons: (i) the maxdiff is the most appropriate probabilistic model for the Case 1 BWS approach and (ii) estimating best and worst values separately can be a source of bias due to potential error variance differences between the best and worst choice observations (Louviere et al., 2015; Flynn and Marley, 2014; Scarpa et al. 2011; Rose, 2014). Data were analysed using a discrete-choice framework. Notably, discrete-choice models are consistent with random utility (McFadden, 1974) and Lancaster consumer theories (Lancaster, 1966). According to random utility theory, the utility for respondent n in choosing alternative j in choice set t, is   Unjt=Vnjt+εnjt (1)where Vnjt is a systematic component that can be observed by the researcher, while ɛnjt is the unobserved error term, which is assumed to be independent of Vnjt. Generally, when respondents are presented with a choice set, they make choices on the basis of the maximisation of the utility they can derive from each alternative of the presented choice set. As such, in making a choice between alternative j and alternative k, respondent n will pick alternative j over alternative k when:   Unjt>Unkt for⁢all⁢ ⁢j≠k. (2) However, in BWS experiments, respondents make choices depending on which pair of alternatives (most important and least important) maximises their utility. Specifically, in each choice set, respondent n chooses the pair of alternatives j and k as the best and worst, respectively, when   Unjt−Unkt>Unlt−Unmt for all⁢j≠l ⁢an⁢d⁢ k≠m. (3) Given that each choice set has J food values (4 in our case), the pair of items chosen by the respondent as best and worst represents a choice from all J(J − 1) possible pairs (12 in this study), which maximises the difference in importance. Following LB, λj, is defined as the observable location of the value j on the scale of importance. Given taste homogeneity this parameter will be constant across respondents. The unobserved level of importance of food value j for respondent n, Inj, is given by   Inj=λj+εnj (4)where ɛnj is the random error term; hence, the probability that the respondent chooses food value j as the best and food value k as the worst will be equal to the probability that the difference among Inj and Ink is larger than any of the J(J − 1) − 1 possible differences among the other food values in the choice set. Using a multinomial logit model (MNL), the probability of respondent n choosing j as the best and k as the worst among pairs of alternative J(J −1) is specified as follows:   Pnjk=eλj−λk∑l=1J∑m=1Jeλl−λm (5)where the choice of respondent n takes the value of 1 for the pair of values chosen by the respondent as best and worst, and the value of 0 for the non-chosen J(J − 1) − 1 pairs of food values. Specifically, λj represents the relative importance of food value j over one of the values, which is normalised to 0. In this way, the dummy variable trap can be avoided. Effects coding was applied: the food value takes the value 1 when the value is described as the best alternative, −1 when the value is described as the worst alternative, and 0 otherwise. The MNL assumes that the error terms are independently and identically distributed (IID) with a Gumbel (Extreme value type I) distribution and implies independence within the alternatives and taste homogeneity across respondents. Heterogeneity in respondents’ food values is likely. When heterogeneity valuations are expected, discrete-choice models such as the random parameters logit (RPL) model should be used. The RPL model allows for random taste variations and accounts for the panel structure of the data (Train, 2003). As such, in contrast to the MNL model, the importance parameter of value j in the RPL model is assumed to be different for each respondent n and was specified as follows:   λ̃nj=λ̅j+σjμnj (6)where λ̅j and σj are the mean and the standard deviation of λj, respectively, and μnj is a random error term that is assumed to be normally distributed with mean zero and unit standard deviation. Substituting equation (6) into equation (5), the RPL can be estimated by maximising a simulated log-likelihood function for μnj (Train, 2003).4 In the standard RPL model, independency across taste parameters is assumed; however, food values are expected to be interdependent. In order to take this interdependency into account,5 the correlation structure of the attribute parameters was assumed to follow a multivariate normal distribution. The estimates from the RPL model might be difficult to interpret because the random error term might vary across respondents, and therefore the mean of the parameter estimates of λj may be confounded with differences in scale. Hence, following LB, we calculated the share of preference, Sj, for each value, which explains how important respondents rate one value j over the other J values:   Sj=eλˆj∑k=1Jeλˆk. (7) Each share can be interpreted as the forecasted probability that the corresponding value is chosen as the most important. If value j has a twice as big preference share as another value, this indicates that the value j is twice as important as the other value. The share of preferences of all the J values must sum to one. 3. Results and discussion In this section, we describe the results obtained from the econometric analysis. Specifically, in our study, the standard MNL model and the RPL model were estimated. From the RPL model estimates, the respondents’ specific preferences for the different food values were calculated using the estimated parameters as priors conditional on actual choices made by each respondent.6 From these posterior estimates, the mean and individual shares of preferences for the 12 values were calculated. 3.1. Model estimates In the BWS approach, the importance of a set of attributes is estimated relative to one of these attributes (Marley and Louviere, 2005; Lusk and Briggeman, 2009; de-Magistris, Gracia and Albisu, 2014). Following LB, we used as the baseline the least important food value, based on the calculation of the percent of times each item was selected best or worst, which in our case is novelty. Estimates from the MNL and RPL models are reported in Table 3. Table 3. Estimates from MNL and RPL models Food value    MNL  RPL      USA  Norway  USA  Norway  Naturalness  Mean  1.598***  2.502***  2.974***  4.491***  (0.036)  (0.042)  (0.091)  (0.107)    SD      2.719***  3.676***      (0.085)  (0.122)  Safety  Mean  2.746***  3.381***  5.139***  6.158***  (0.040)  (0.045)  (0.070)  (0.111)    SD      3.478***  3.177***      (0.101)  (0.117)  Environmental impact  Mean  1.360***  2.131***  2.406***  3.801***  (0.036)  (0.041)  (0.084)  (0.097)    SD      2.421***  3.207***  (0.086)  (0.106)  Origin  Mean  0.918***  1.578***  1.732***  2.738***  (0.037)  (0.040)  (0.079)  (0.102)    SD      2.010***  3.432***  (0.081)  (0.117)  Fairness  Mean  1.228***  2.186***  2.185***  3.939***  (0.036)  (0.041)  (0.091)  (0.098)    SD      2.146***  3.170***  (0.094)  (0.010)  Nutrition  Mean  1.922***  2.404***  3.612***  4.466***  (0.037)  (0.042)  (0.092)  (0.097)    SD      2.606***  3.170***  (0.076)  (0.010)  Taste  Mean  2.113***  2.714***  3.912***  5.133***  (0.038)  (0.043)  (0.095)  (0.086)    SD      2.648***  1.487***  (0.081)  (0.098)  Convenience  Mean  0.748***  0.850***  1.331***  1.496***  (0.033)  (0.036)  (0.069)  (0.084)    SD      1.826***  2.464***  (0.080)  (0.084)  Appearance  Mean  1.114***  1.469***  2.112***  2.670***  (0.036)  (0.040)  (0.076)  (0.077)    SD      2.015***  1.609***  (0.102)  (0.073)  Price  Mean  1.741***  1.780***  3.219***  3.337***  (0.037)  (0.041)  (0.097)  (0.094)    SD      2.855***  2.388***  (0.083)  (0.108)  Animal welfare  Mean  1.544***  2.470***  2.750***  4.452***  (0.036)  (0.042)  (0.091)  (0.100)    SD      2.738***  3.124***  (0.102)  (0.104)  Novelty  Mean  0.000  0.000  0.000  0.000  SD  0.000  0.000  Number of choices    12,300  12,444  12,300  12,444  Log-likelihood    −26,384  −25,057  −22,161  −19,951  BIC    52,897  50,217  45,048  40,628  AIC    52,790  50,135  44,477  40,056  AIC/N    4.292  4.029  3.616  3.219  Food value    MNL  RPL      USA  Norway  USA  Norway  Naturalness  Mean  1.598***  2.502***  2.974***  4.491***  (0.036)  (0.042)  (0.091)  (0.107)    SD      2.719***  3.676***      (0.085)  (0.122)  Safety  Mean  2.746***  3.381***  5.139***  6.158***  (0.040)  (0.045)  (0.070)  (0.111)    SD      3.478***  3.177***      (0.101)  (0.117)  Environmental impact  Mean  1.360***  2.131***  2.406***  3.801***  (0.036)  (0.041)  (0.084)  (0.097)    SD      2.421***  3.207***  (0.086)  (0.106)  Origin  Mean  0.918***  1.578***  1.732***  2.738***  (0.037)  (0.040)  (0.079)  (0.102)    SD      2.010***  3.432***  (0.081)  (0.117)  Fairness  Mean  1.228***  2.186***  2.185***  3.939***  (0.036)  (0.041)  (0.091)  (0.098)    SD      2.146***  3.170***  (0.094)  (0.010)  Nutrition  Mean  1.922***  2.404***  3.612***  4.466***  (0.037)  (0.042)  (0.092)  (0.097)    SD      2.606***  3.170***  (0.076)  (0.010)  Taste  Mean  2.113***  2.714***  3.912***  5.133***  (0.038)  (0.043)  (0.095)  (0.086)    SD      2.648***  1.487***  (0.081)  (0.098)  Convenience  Mean  0.748***  0.850***  1.331***  1.496***  (0.033)  (0.036)  (0.069)  (0.084)    SD      1.826***  2.464***  (0.080)  (0.084)  Appearance  Mean  1.114***  1.469***  2.112***  2.670***  (0.036)  (0.040)  (0.076)  (0.077)    SD      2.015***  1.609***  (0.102)  (0.073)  Price  Mean  1.741***  1.780***  3.219***  3.337***  (0.037)  (0.041)  (0.097)  (0.094)    SD      2.855***  2.388***  (0.083)  (0.108)  Animal welfare  Mean  1.544***  2.470***  2.750***  4.452***  (0.036)  (0.042)  (0.091)  (0.100)    SD      2.738***  3.124***  (0.102)  (0.104)  Novelty  Mean  0.000  0.000  0.000  0.000  SD  0.000  0.000  Number of choices    12,300  12,444  12,300  12,444  Log-likelihood    −26,384  −25,057  −22,161  −19,951  BIC    52,897  50,217  45,048  40,628  AIC    52,790  50,135  44,477  40,056  AIC/N    4.292  4.029  3.616  3.219  Note: ***Indicate significance at the 1 per cent level. Numbers in parentheses are standard errors. Table 3 indicates that we obtained a better fit with the RPL than the MNL model in both samples as shown by the increases in the log-likelihood values and the reductions in the AIC, BIC and AIC/N statistics. In addition, Table 3 shows that the derived standard deviations of the attributes’ parameters are statistically different from zero, and our assumption of heterogeneity in preferences for the 12 values across individuals cannot be rejected. 3.2. Shares of preferences for the 12 food values On the basis of the RPL estimates, we assessed the preferences for the 12 food values by calculating their shares of preference. In Table 4, we report the shares of preference for the different values, from the most to the least important in each country. Table 4. Preference shares and rankings of importance of food values in USA and Norway Rank  USA  Norway  Food value  Share  Food value  Share  1  Safety  0.380*  Safety  0.313*  2  Price  0.115*  Naturalness  0.125*  3  Taste  0.112  Taste  0.112  4  Nutrition  0.088*  Animal welfare  0.098*  5  Naturalness  0.078*  Nutrition  0.094*  6  Animal welfare  0.077*  Price  0.074*  7  Environmental impact  0.039*  Fairness  0.060*  8  Fairness  0.028*  Origin  0.047*  9  Appearance  0.027*  Environmental impact  0.046*  10  Origin  0.026*  Appearance  0.018*  11  Convenience  0.020*  Convenience  0.011*  12  Novelty  0.012*  Novelty  0.002*  Rank  USA  Norway  Food value  Share  Food value  Share  1  Safety  0.380*  Safety  0.313*  2  Price  0.115*  Naturalness  0.125*  3  Taste  0.112  Taste  0.112  4  Nutrition  0.088*  Animal welfare  0.098*  5  Naturalness  0.078*  Nutrition  0.094*  6  Animal welfare  0.077*  Price  0.074*  7  Environmental impact  0.039*  Fairness  0.060*  8  Fairness  0.028*  Origin  0.047*  9  Appearance  0.027*  Environmental impact  0.046*  10  Origin  0.026*  Appearance  0.018*  11  Convenience  0.020*  Convenience  0.011*  12  Novelty  0.012*  Novelty  0.002*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the two samples is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 4 shows that the mean preference shares are statistically different in the two countries, except in the case of taste attribute. However, if we consider the differences in the ranking of the importance of food values across the two countries, respondents’ preferences are quite similar in many aspects. In both countries, safety is clearly the most important value with a share of 38.0 per cent in the USA and 31.3 per cent in Norway. The high importance of safety is in line with the results of LB, who also found that safety was the most important food value in the USA. After safety, there is a group of five values that are fairly close in importance with shares ranging between 11.5 per cent and 7.7 per cent in the USA, and 12.5 per cent and 7.4 per cent in Norway (price, taste, nutrition, naturalness and animal welfare). The remaining values have preference shares ranging between 3.9 per cent and 1.2 per cent in the USA, and 6.0 per cent and 0.2 per cent in Norway. Convenience and novelty are the least important values in both countries. These similarities in values may reflect a convergence in food values between Europe and the USA. Within these broad similarities in the rankings of food values, there are also some notable differences. Price was the second-most important value among the US respondents, which is consistent with the LB study. In contrast, Norwegian respondents considered price as the sixth-most important value. The relatively lower importance of price in Norway may be a reflection of the more equal income distribution. Furthermore, taste was rated as the third-most important value both in the USA and Norway, which again is consistent with the results in LB. Nutrition was predicted as most important for about 9 percent of the respondents in each country. This result is somewhat at odds with past studies that showed consumers in the USA tend to pay more attention to the nutritional content of food products, as compared to European consumers (Rozin, Levine and Stoess, 1991; Bech-Larsen and Grunert, 2003); however, the result may reflect a convergence between the two countries. Additionally, naturalness was the second-most important value of Norwegian respondents, while it was the fifth-most important value of US respondents, which is all consistent with the current literature and not surprising given the differences in food environment. Indeed, several studies have shown that European consumers are generally less willing to consume food that has been produced with technologies such as genetic modification, or with cattle fed with growth hormones (Chern et al. 2002; Alfnes and Rickertsen, 2003; Lusk, Roosen and Fox, 2003). In addition, this result is also consistent with LB who found that naturalness was rated as the fifth-most important food value. Food values concerning ethical aspects of food production such as fairness and animal welfare were ranked as more important by the Norwegian than the US respondents. The higher importance of fairness in Norway is as expected given that the social and economic welfare of farmers are crucial aspects in the Norwegian food system, and the result is also consistent with the high equality in income distribution. The higher importance of animal welfare in Norway may also reflect that animal welfare labelling regulations tend to be more developed in Europe than in the USA (Mitchell, 2001; Napolitano et al., 2010; Vandemoortele and Deconinck, 2014). Environmental impact was ranked as the seventh-most important attribute by the US sample and the ninth-most important by the Norwegian sample, however, the actual preference share was slightly higher in Norway. This result is not unexpected given the higher presence of regulated eco-food labels in the European food system than in the USA (Czarnezki, 2011). Not surprising, origin was rated as somewhat more important by the Norwegian respondents than the US ones. Although existing literature reports that consumers both in the USA and Europe are generally willing to pay a price premium for local or designated origin of food products (Darby et al., 2008; Aprile, Caputo and Nayga, 2012; de-Magistris and Gracia, 2014; Meas et al., 2015), origin is ranked relatively low in both countries. This result is consistent with LB, who found that origin was considered as the least important value in their US study. 3.3. Socio-demographic information and shares of preferences for the 12 food values Overall, the results suggest that US and Norwegian respondents differ mostly in terms of the ranking of price and naturalness. However, Table 1 shows that the US and Norwegian samples differ in terms of some socio-demographic variables, which might explain some of the similarities/differences in preferences for food values across the countries. Specifically, we observed that the two samples differ in terms of age, education, having children or not, income and belonging to rural/urban areas. As such, in order to test whether individuals’ preferences for food values may differ in terms of socio-demographic characteristics, we divided the US and Norwegian samples into different subgroups based on age (young/old), education (low/high), the presence of children in the household (with/without), income level (low/high), whether residing in urban/rural area (urban/rural) and whether the respondent had purchased organic food during the previous 12 months (purchased/not purchased). In case of age, education and income, we determined the grouping based on the median values in the samples, and then divided each sample into two groups. We estimated the RPL model for each subgroup and calculated the respondents’ shares of preferences for the subgroups. In addition, we also report results from t-tests to test whether the preferences for the food values differed among the subgroups within each country (indicated with asterisks in Tables 5–10). Table 5. Shares of preferences and rankings by country and age Rank  USA  Norway  Old (n = 526)  Young (n = 499)  Old (n = 490)  Young (n = 547)  1  Safety  0.411*  Safety  0.352*  Safety  0.287*  Safety  0.362*  2  Price  0.129*  Taste  0.107*  Naturalness  0.147*  Taste  0.131*  3  Taste  0.124*  Price  0.103*  Taste  0.124*  Animal welfare  0.098*  4  Naturalness  0.083  Nutrition  0.094*  Nutrition  0.083  Naturalness  0.087*  5  Animal welfare  0.073  Animal welfare  0.078  Animal welfare  0.073*  Nutrition  0.087  6  Nutrition  0.072*  Naturalness  0.078  Price  0.072  Price  0.082  7  Environmental impact  0.027*  Environmental impact  0.051*  Fairness  0.027*  Fairness  0.052*  8  Appearance  0.022*  Fairness  0.033*  Origin  0.022*  Environmental impact  0.039  9  Fairness  0.022*  Origin  0.031*  Environmental impact  0.022  Origin  0.037*  10  Origin  0.021*  Appearance  0.029*  Appearance  0.021  Appearance  0.015  11  Convenience  0.011*  Convenience  0.026*  Convenience  0.011  Convenience  0.010  12  Novelty  0.005*  Novelty  0.018*  Novelty  0.005  Novelty  0.001  Rank  USA  Norway  Old (n = 526)  Young (n = 499)  Old (n = 490)  Young (n = 547)  1  Safety  0.411*  Safety  0.352*  Safety  0.287*  Safety  0.362*  2  Price  0.129*  Taste  0.107*  Naturalness  0.147*  Taste  0.131*  3  Taste  0.124*  Price  0.103*  Taste  0.124*  Animal welfare  0.098*  4  Naturalness  0.083  Nutrition  0.094*  Nutrition  0.083  Naturalness  0.087*  5  Animal welfare  0.073  Animal welfare  0.078  Animal welfare  0.073*  Nutrition  0.087  6  Nutrition  0.072*  Naturalness  0.078  Price  0.072  Price  0.082  7  Environmental impact  0.027*  Environmental impact  0.051*  Fairness  0.027*  Fairness  0.052*  8  Appearance  0.022*  Fairness  0.033*  Origin  0.022*  Environmental impact  0.039  9  Fairness  0.022*  Origin  0.031*  Environmental impact  0.022  Origin  0.037*  10  Origin  0.021*  Appearance  0.029*  Appearance  0.021  Appearance  0.015  11  Convenience  0.011*  Convenience  0.026*  Convenience  0.011  Convenience  0.010  12  Novelty  0.005*  Novelty  0.018*  Novelty  0.005  Novelty  0.001  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 6. Shares of preferences and rankings by country and education level Rank  USA  Norway  High (n = 531)  Low (n = 494)  High (n = 653)  Low (n = 384)  1  Safety  0.384  Safety  0.363  Safety  0.293  Safety  0.319  2  Nutrition  0.113*  Price  0.139*  Naturalness  0.148*  Animal welfare  0.156*  3  Taste  0.112  Taste  0.120  Taste  0.128*  Taste  0.097*  4  Price  0.096*  Animal welfare  0.088*  Nutrition  0.108*  Naturalness  0.084*  5  Naturalness  0.093*  Naturalness  0.072*  Animal welfare  0.072*  Price  0.077  6  Animal welfare  0.063*  Nutrition  0.055*  Price  0.072  Nutrition  0.074*  7  Environmental impact  0.037*  Environmental impact  0.042*  Fairness  0.056  Fairness  0.061  8  Fairness  0.026  Fairness  0.031  Environmental impact  0.050*  Origin  0.059*  9  Origin  0.025  Origin  0.029  Origin  0.040*  Environmental impact  0.039*  10  Appearance  0.024*  Appearance  0.029*  Appearance  0.017*  Appearance  0.020*  11  Convenience  0.018  Convenience  0.021  Convenience  0.014  Convenience  0.012  12  Novelty  0.005*  Novelty  0.011*  Novelty  0.001*  Novelty  0.003*  Rank  USA  Norway  High (n = 531)  Low (n = 494)  High (n = 653)  Low (n = 384)  1  Safety  0.384  Safety  0.363  Safety  0.293  Safety  0.319  2  Nutrition  0.113*  Price  0.139*  Naturalness  0.148*  Animal welfare  0.156*  3  Taste  0.112  Taste  0.120  Taste  0.128*  Taste  0.097*  4  Price  0.096*  Animal welfare  0.088*  Nutrition  0.108*  Naturalness  0.084*  5  Naturalness  0.093*  Naturalness  0.072*  Animal welfare  0.072*  Price  0.077  6  Animal welfare  0.063*  Nutrition  0.055*  Price  0.072  Nutrition  0.074*  7  Environmental impact  0.037*  Environmental impact  0.042*  Fairness  0.056  Fairness  0.061  8  Fairness  0.026  Fairness  0.031  Environmental impact  0.050*  Origin  0.059*  9  Origin  0.025  Origin  0.029  Origin  0.040*  Environmental impact  0.039*  10  Appearance  0.024*  Appearance  0.029*  Appearance  0.017*  Appearance  0.020*  11  Convenience  0.018  Convenience  0.021  Convenience  0.014  Convenience  0.012  12  Novelty  0.005*  Novelty  0.011*  Novelty  0.001*  Novelty  0.003*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 7. Shares of preferences and rankings by country and the presence of children in the household Rank  USA  Norway  With (n = 457)  Without (n = 568)  With (n = 307)  Without (n = 730)  1  Safety  0.414*  Safety  0.354*  Safety  0.385*  Safety  0.279*  2  Taste  0.100*  Price  0.140*  Naturalness  0.119  Taste  0.123*  3  Nutrition  0.094*  Taste  0.131*  Taste  0.105*  Naturalness  0.117  4  Price  0.089*  Animal welfare  0.082  Price  0.093*  Animal welfare  0.116*  5  Naturalness  0.087*  Nutrition  0.076*  Nutrition  0.088  Nutrition  0.101  6  Animal welfare  0.071  Naturalness  0.072*  Animal welfare  0.062*  Price  0.065*  7  Environmental impact  0.041  Environmental impact  0.035  Fairness  0.046*  Fairness  0.061*  8  Origin  0.027  Fairness  0.029*  Environmental impact  0.042  Origin  0.054*  9  Fairness  0.024*  Appearance  0.028  Origin  0.033*  Environmental impact  0.048  10  Appearance  0.023  Origin  0.023  Appearance  0.020  Appearance  0.017  11  Convenience  0.019  Convenience  0.019  Convenience  0.006*  Convenience  0.015*  12  Novelty  0.011  Novelty  0.011  Novelty  0.001*  Novelty  0.002*  Rank  USA  Norway  With (n = 457)  Without (n = 568)  With (n = 307)  Without (n = 730)  1  Safety  0.414*  Safety  0.354*  Safety  0.385*  Safety  0.279*  2  Taste  0.100*  Price  0.140*  Naturalness  0.119  Taste  0.123*  3  Nutrition  0.094*  Taste  0.131*  Taste  0.105*  Naturalness  0.117  4  Price  0.089*  Animal welfare  0.082  Price  0.093*  Animal welfare  0.116*  5  Naturalness  0.087*  Nutrition  0.076*  Nutrition  0.088  Nutrition  0.101  6  Animal welfare  0.071  Naturalness  0.072*  Animal welfare  0.062*  Price  0.065*  7  Environmental impact  0.041  Environmental impact  0.035  Fairness  0.046*  Fairness  0.061*  8  Origin  0.027  Fairness  0.029*  Environmental impact  0.042  Origin  0.054*  9  Fairness  0.024*  Appearance  0.028  Origin  0.033*  Environmental impact  0.048  10  Appearance  0.023  Origin  0.023  Appearance  0.020  Appearance  0.017  11  Convenience  0.019  Convenience  0.019  Convenience  0.006*  Convenience  0.015*  12  Novelty  0.011  Novelty  0.011  Novelty  0.001*  Novelty  0.002*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 8. Shares of preferences and rankings by country and income level Rank  USA  Norway  Low (n = 441)  High (n = 584)  Low (n = 358)  High (n = 679)  1  Safety  0.373*  Safety  0.382*  Safety  0.258*  Safety  0.339*  2  Price  0.160*  Taste  0.123*  Animal welfare  0.128*  Taste  0.123  3  Taste  0.096*  Nutrition  0.101*  Taste  0.122  Naturalness  0.113  4  Animal welfare  0.080*  Price  0.089*  Naturalness  0.119  Nutrition  0.087*  5  Naturalness  0.076  Naturalness  0.084  Nutrition  0.112*  Price  0.083*  6  Nutrition  0.068*  Animal welfare  0.073*  Price  0.070*  Animal welfare  0.083*  7  Environmental impact  0.038  Environmental impact  0.039  Fairness  0.070*  Fairness  0.061*  8  Fairness  0.028  Appearance  0.026  Environmental impact  0.052*  Origin  0.059*  9  Appearance  0.026  Origin  0.026  Origin  0.051*  Environmental impact  0.039*  10  Origin  0.025  Fairness  0.026  Appearance  0.017  Appearance  0.020  11  Convenience  0.021  Convenience  0.019  Convenience  0.014  Convenience  0.012  12  Novelty  0.011  Novelty  0.012  Novelty  0.001*  Novelty  0.003*  Rank  USA  Norway  Low (n = 441)  High (n = 584)  Low (n = 358)  High (n = 679)  1  Safety  0.373*  Safety  0.382*  Safety  0.258*  Safety  0.339*  2  Price  0.160*  Taste  0.123*  Animal welfare  0.128*  Taste  0.123  3  Taste  0.096*  Nutrition  0.101*  Taste  0.122  Naturalness  0.113  4  Animal welfare  0.080*  Price  0.089*  Naturalness  0.119  Nutrition  0.087*  5  Naturalness  0.076  Naturalness  0.084  Nutrition  0.112*  Price  0.083*  6  Nutrition  0.068*  Animal welfare  0.073*  Price  0.070*  Animal welfare  0.083*  7  Environmental impact  0.038  Environmental impact  0.039  Fairness  0.070*  Fairness  0.061*  8  Fairness  0.028  Appearance  0.026  Environmental impact  0.052*  Origin  0.059*  9  Appearance  0.026  Origin  0.026  Origin  0.051*  Environmental impact  0.039*  10  Origin  0.025  Fairness  0.026  Appearance  0.017  Appearance  0.020  11  Convenience  0.021  Convenience  0.019  Convenience  0.014  Convenience  0.012  12  Novelty  0.011  Novelty  0.012  Novelty  0.001*  Novelty  0.003*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 9. Shares of preferences and rankings by country and living in urban/rural area Rank  USA  Norway  Rural (n = 188)  Urban (n = 837)  Rural (n = 257)  Urban (n = 780)  1  Safety  0.369  Safety  0.376  Safety  0.317  Safety  0.315  2  Price  0.132  Taste  0.117  Naturalness  0.159*  Taste  0.118*  3  Taste  0.124  Price  0.110  Animal welfare  0.122*  Naturalness  0.109*  4  Naturalness  0.099*  Nutrition  0.088*  Taste  0.084*  Nutrition  0.106*  5  Animal welfare  0.087  Naturalness  0.085*  Origin  0.077*  Animal welfare  0.090*  6  Nutrition  0.065*  Animal welfare  0.072  Fairness  0.065  Price  0.083*  7  Environmental impact  0.033  Environmental impact  0.040  Nutrition  0.061*  Fairness  0.057  8  Fairness  0.026  Fairness  0.028  Price  0.049*  Environmental impact  0.051*  9  Appearance  0.023  Appearance  0.027  Environmental impact  0.032*  Origin  0.038*  10  Origin  0.020  Origin  0.026  Appearance  0.021  Appearance  0.019  11  Convenience  0.015  Convenience  0.021  Convenience  0.011  Convenience  0.013  12  Novelty  0.006*  Novelty  0.012*  Novelty  0.001*  Novelty  0.002*  Rank  USA  Norway  Rural (n = 188)  Urban (n = 837)  Rural (n = 257)  Urban (n = 780)  1  Safety  0.369  Safety  0.376  Safety  0.317  Safety  0.315  2  Price  0.132  Taste  0.117  Naturalness  0.159*  Taste  0.118*  3  Taste  0.124  Price  0.110  Animal welfare  0.122*  Naturalness  0.109*  4  Naturalness  0.099*  Nutrition  0.088*  Taste  0.084*  Nutrition  0.106*  5  Animal welfare  0.087  Naturalness  0.085*  Origin  0.077*  Animal welfare  0.090*  6  Nutrition  0.065*  Animal welfare  0.072  Fairness  0.065  Price  0.083*  7  Environmental impact  0.033  Environmental impact  0.040  Nutrition  0.061*  Fairness  0.057  8  Fairness  0.026  Fairness  0.028  Price  0.049*  Environmental impact  0.051*  9  Appearance  0.023  Appearance  0.027  Environmental impact  0.032*  Origin  0.038*  10  Origin  0.020  Origin  0.026  Appearance  0.021  Appearance  0.019  11  Convenience  0.015  Convenience  0.021  Convenience  0.011  Convenience  0.013  12  Novelty  0.006*  Novelty  0.012*  Novelty  0.001*  Novelty  0.002*  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the two subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. Table 10. Shares of preferences and rankings by country and organic food purchases Rank  USA  Norway  Not purchased (n = 393)  Purchased (n = 577)  Not purchased (n = 491)  Purchased (n = 458)  1  Safety  0.373†  Safety  0.367†  Safety  0.311*†  Safety  0.294*†  2  Price  0.160*†  Naturalness  0.131*†  Taste  0.154*†  Naturalness  0.167*†  3  Taste  0.096*†  Nutrition  0.111*  Price  0.116*†  Animal welfare  0.115*†  4  Animal welfare  0.062†  Taste  0.083*  Nutrition  0.090†  Nutrition  0.105  5  Nutrition  0.053*†  Animal welfare  0.076†  Naturalness  0.087*†  Taste  0.084*  6  Naturalness  0.031*†  Price  0.070*†  Animal welfare  0.087*†  Fairness  0.075*†  7  Appearance  0.030*  Environmental impact  0.045*  Origin  0.046†  Environmental impact  0.069*  8  Environmental impact  0.027*  Fairness  0.030*†  Fairness  0.043*†  Origin  0.046†  9  Fairness  0.023*†  Origin  0.030*†  Appearance  0.024*  Price  0.032*†  10  Origin  0.018*†  Appearance  0.022*†  Environmental impact  0.023*  Appearance  0.007*†  11  Convenience  0.014*  Convenience  0.021*†  Convenience  0.015*  Convenience  0.006*†  12  Novelty  0.006*  Novelty  0.014*†  Novelty  0.003*  Novelty  0.001*†  Rank  USA  Norway  Not purchased (n = 393)  Purchased (n = 577)  Not purchased (n = 491)  Purchased (n = 458)  1  Safety  0.373†  Safety  0.367†  Safety  0.311*†  Safety  0.294*†  2  Price  0.160*†  Naturalness  0.131*†  Taste  0.154*†  Naturalness  0.167*†  3  Taste  0.096*†  Nutrition  0.111*  Price  0.116*†  Animal welfare  0.115*†  4  Animal welfare  0.062†  Taste  0.083*  Nutrition  0.090†  Nutrition  0.105  5  Nutrition  0.053*†  Animal welfare  0.076†  Naturalness  0.087*†  Taste  0.084*  6  Naturalness  0.031*†  Price  0.070*†  Animal welfare  0.087*†  Fairness  0.075*†  7  Appearance  0.030*  Environmental impact  0.045*  Origin  0.046†  Environmental impact  0.069*  8  Environmental impact  0.027*  Fairness  0.030*†  Fairness  0.043*†  Origin  0.046†  9  Fairness  0.023*†  Origin  0.030*†  Appearance  0.024*  Price  0.032*†  10  Origin  0.018*†  Appearance  0.022*†  Environmental impact  0.023*  Appearance  0.007*†  11  Convenience  0.014*  Convenience  0.021*†  Convenience  0.015*  Convenience  0.006*†  12  Novelty  0.006*  Novelty  0.014*†  Novelty  0.003*  Novelty  0.001*†  An asterisk implies that the hypothesis that the mean of the corresponding values are the same across the subgroups within each country is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. †The hypothesis that the mean of the corresponding values are the same across the subgroups within the organic purchasers and non-purchaser is rejected at the 0.05 level of significance according to a two-tailed unpaired t-test. In Tables 5 and 6, we report the shares of preferences, respectively, for young and old respondents, and for respondents with a high and low education level in each country. From Tables 5 and 6, we observe that the shares of preferences for the different food values tend to be similar across the age and education groups in both countries. Only in the USA did we observe a difference in the rank of the attribute nutrition across higher and lower educated groups. Nutrition is rated as the second-most important value by higher educated people, while it is rated as the sixth-most important value by lower educated people. We would expect that the presence of children in the household would also conspicuously influence respondents’ preferences for the attribute nutrition (Drichoutis, Lazaridis and Nayga, 2006). However, Table 7 shows that this is not the case in either sample. Indeed, nutrition is rated in the USA as the third-most important value by respondents with children in the household and the fifth-most important by respondents without children. In the Norwegian sample, nutrition is equally rated by the two subgroups. Table 7 actually shows that there are no substantial differences in the rating of the importance of the food values across respondents with and without children in the household. However, an interesting result is that price is ranked fourth by respondents living with children both in the USA and Norway. In regards to the price attribute, the difference in the importance of price may be explained by a higher income level and a more equal income distribution in Norway. In Table 8, we report the preference shares of low- and high-income respondents in both countries. In the USA, lower income respondents considered price to be slightly more important than higher income respondents, and price is the second-most important food value for lower income respondents and the fourth-most important for higher income respondents. On the other hand, price is rated as the fifth-most important food value by higher income Norwegian respondents and the sixth-most important value by lower income respondents. These results suggest that price preferences between the income subgroups within each country tend to be similar. Finally, we tested whether residing in rural/urban area had a significant impact on respondents’ preferences for the food values. Table 9 shows that in Norway, the ranking of origin notably changes depending on whether the respondent resides in rural or urban area: origin is on average the fifth-most important attribute for individuals living in rural areas and the ninth-most important for individuals living in urban areas. However, in the case of the USA, we do not observe this difference in the ranking between the two subgroups. According to these results, we might conclude that socio-demographic variables scarcely explain the differences/similarities in preferences for food values between the two countries. However, LB found that the preferences for the different food values particularly differed among organic food purchasers and non-purchasers. Specifically, LB observed that price and naturalness (the two attributes which Norwegian and US respondents valued most differently in our survey) were the most differently rated food values between consumers who purchased organic foods and consumers who did not purchase organic foods in the USA. In Table 10, we report the mean shares of preference of US and Norwegian respondents who have purchased and not purchased organic food during the 12 months before the survey.7 Consistent with LB, the importance of price was rated differently across the organic food purchaser/non-purchaser subgroups. In the case of the USA, organic non-purchasers rated price as the second-most important value, while organic purchasers rated price as the sixth-most important food value. Notably, Norwegian respondents who did not buy organic food, rated price as the third-most important food value, while organic purchasers only rated price as the ninth-most important value. In Table 10, we also report results from t-tests to test whether the preferences for the food values differed between organic purchasers in the USA and Norway and organic non-purchasers in the USA and in Norway (indicated with ‘†’ in the table). Although the mean shares of preferences were statistically different for most of the values both in the case of Norwegian and American organic purchasers and organic non-purchasers, the food values become more similar across the samples when mean shares of preferences between organic purchasers and non-purchasers in the two countries were compared. Generally, organic purchasers gave more importance to naturalness and food values related to sustainability issues, while organic non-purchasers gave more importance to attributes such as appearance and especially price. Specifically, price was rated as one of the least important attributes by organic purchasers, while one of the most important attributes by the organic non-purchasers both in the USA and in Norway. 4. Conclusions To the best of our knowledge, our study is the first attempt to estimate consumers’ preferences for food values in a multi-country setting. We used a BWS approach in order to compare consumers’ preferences for food values in the USA and Norway. We included 12 food values: naturalness, taste, price, safety, convenience, nutrition, novelty, origin, fairness, appearance, environmental impact and animal welfare. Our results show that there were large similarities in preferences for food values among US and Norwegian respondents. For instance, safety was clearly the most important value in both countries. Similarly, taste was rated as the third-most important value and convenience and novelty were rated as the two least important values in both countries. There were also notable differences in the evaluation of the importance of price. Price was considered to be the second-most important value by US respondents but only the sixth-most important value by the Norwegian respondents. In addition, naturalness was rated as the second-most important value by Norwegian respondents but rated only as the fifth-most important value by US respondents. This difference is in line with the existing literature that shows European consumers are generally willing to pay a higher price for non-GM or hormone free foods, as compared with US consumers (Chern et al. 2002; Lusk, Roosen and Fox, 2003). Specifically, Norwegian respondents gave more importance to food values related to ethical aspects of food production such as fairness and animal welfare; however, it is important to note that animal welfare was rated among the six-most important food values in both countries. This finding confirms that the addition of animal welfare to LB’s set of attributes is important in assessing food values. This result also suggests that animal welfare labelling could be a potential strategy for the marketing of food products both in the USA and in Europe. For example, given that origin is rated less important than animal welfare, this might suggest that imported meat and dairy products might have a better market share if produced respecting the welfare of the animals. On the other hand, novelty, i.e. the other new food value that we have introduced besides animal welfare, was the least important value for both samples. We included novelty in order to capture the value related to enthusiasm, excitement for new products, but this outcome might suggest that novelty in food products may be associated with food neophobia, which could make individuals reluctant towards food products they do not know (Lähteenmäki and Arvola, 2001; Camarena, Sanjuán and Philippidis, 2011; Mielby et al., 2012). Nutrition was ranked as one of the most important food values in both countries, which should encourage policy makers in Europe to use regulated nutritional labelling on food products. Furthermore, origin was rated as one of the least important values in both countries. This result is at odds with the existing literature that shows that consumers both in the USA and Europe are usually willing to pay a price premium for locally or nationally produced food products, even over other attributes such as organic production, fair trade or low carbon emission (Basu and Hicks, 2008; Darby et al. 2008; Hu, Woods and Bastin, 2009; Onozaka and Mcfadden, 2011; Campbell, Mhlanga and Lesschaeve, 2013; de-Magistris and Gracia, 2014; Gracia, Barreiro-Hurlé and Galán, 2014). As LB suggest, the low importance of origin may be explained by the fact that in preference elicitation methods such as conjoint analyses or choice experiments, differences in consumers’ preferences are estimated in a range of specific attributes levels, which might not be the levels that endogenously come to mind for the consumer. As such, the use of specific attribute levels could be a source of bias in revealing individuals’ preferences. In addition, the use of a specific food product might also influence consumers’ perceptions for different food attributes. For example, Scarpa, Philippidis and Spalatro (2005) observed that consumers’ evaluations for locally produced and organic food claims varied according to the product under consideration. In this study, we did not specify any food product or food attribute level. To test the robustness of our findings, more research on comparison of consumers’ preferences for food values across countries and also across different product groups is warranted. On the other hand, regarding the naturalness, consumers who purchase organic food gave more importance to this value, as compared with non-purchasers in both countries. Moreover, our results show that organic purchasers and organic non-purchasers in both countries rated naturalness similarly; the same applied for price. Hence, these findings suggest that purchasing attitudes might also be an important factor when assessing consumers’ preferences for food values in different countries. Specifically, we observed that while an attitudinal variable, such as buying organic food, substantially affected how individuals rated the 12 food attributes, socio-demographic information, including the belonging to a certain country, did not have such a relevant impact on respondents’ preferences for the food values. These results might then confirm the intuition of LB in defining these more abstract food attributes as more stable, intrinsic meta-preferences which drive consumers’ food choices. As such, our results suggest that food values are an important factor in explaining behavioural reasoning in food consumption. Finally, our findings have important implications for food marketing and policy. For example, since safety is the top food value in both the USA and Norway, food producers need to constantly be cognisant of potential food safety incidents in their products. The recent insecticide contamination in eggs in Europe is a prime example of how incidents like these can significantly affect the livelihood of food producers and erode the confidence of consumers. The implications for food safety policy are equally compelling obviously given the immense importance that consumers are putting on the safety of the products they consume. In terms of the food marketing benefits, the relative rankings of the food values in our study can be useful and informative in terms of the new products that could be developed by the food industry and also in terms of product differentiation using food labels, particularly for the credence food values that are valued more by consumers (e.g. nutrition and animal welfare). Our results can also potentially be used for ongoing trade negotiations between Europe and the USA. The similarities in food values among consumers in both countries are large, with a clear emphasis on safety, suggesting that consumers in both countries highly value the safety of food products, both domestically produced and imported products. Furthermore, price, taste, nutrition, naturalness and animal welfare are considered to be the five-most important food values after safety in both countries. Consumers in both the USA and Europe would benefit from increased trade in agricultural products that could potentially lower prices and increase the variation in the attributes of food, as long as the products are safe and adequately labelled for the key attributes related to nutrition, naturalness and animal welfare. Supplementary data Supplementary data are available at European Review of Agricultural Economics online. Funding This work was supported by the Norwegian Research Council (grant # 233800/E50) and partially supported by the National Research Foundation of Korea (Grant #NRF-2014S1A3A2044459). Acknowledgements A special thank goes to Jayson Lusk for his support in the design of the Best Worst Scaling and in the models estimation. The authors also want to thank Tiziana de-Magistris, Diana Danforth and Vincenzina Caputo for the suggestions and support in the data analysis. Finally, the authors want to thank the anonymous reviewers for their insightful comments. Any remaining errors are the responsibility of the authors. Footnotes 1 We use the term ‘food values’ in order to be consistent with the terminology used by LB. However, it is important to point out as suggested to us by a reviewer that many of these food values are actually ‘food quality attributes’ and not higher constructs such as ‘values’, which are cognitive representation of concepts of beliefs (Schwartz and Bilsky, 1987). 2 It is difficult to find comparable statistics for households’ income distribution in USA and Norway. However, according to the OECD database (OECD, 2016a), the GDP per capita calculated at 2011 purchasing power parities was 172 and 138 in Norway and in the USA, respectively, as compared with an OECD average of 100. For household consumption expenditures, the numbers were 121 for Norway and 146 for the USA. The Gini index in 2011 was 0.25 for Norway and 0.39 for the USA (OECD, 2016b). 3 For more information: http://ipsos-mmi.no/ 4 We used 1000 Halton draws for the simulation. 5 In both samples, most of the estimates in the Cholesky matrix were statistically significant, indicating that correlation across the parameters exists across both models (results are available upon request). 6 These posterior estimates are precisely the means of the parameter distributions, which are conditional on the actual choices of each respondent. These estimates might not be the same as the actual respondent’s coefficients (Train, 2003); however, when respondents face several choice scenarios, the difference between the values can be very small. 7 In our survey, we asked respondents: ‘Have you purposely purchased organic foods during the last 12 months?’ We also gave them the opportunity to answer: ‘I do not know.’ In defining the ‘organic food purchasers’ and the ‘organic food non-purchasers’ subgroups, we did not include the observations of respondents who replied: ‘I do not know.’ References Adamowicz, W. L. and Swait, J. D. ( 2012). Are food choices really habitual? Integrating habits, variety-seeking, and compensatory choice in a utility-maximizing framework. American Journal of Agricultural Economics  95( 1): 17– 41. Google Scholar CrossRef Search ADS   Alfnes, F. ( 2004). Stated preferences for imported and hormone-treated beef: application of a mixed logit model. European Review of Agriculture Economics  31( 1): 19– 37. Google Scholar CrossRef Search ADS   Alfnes, F. and Rickertsen, K. ( 2003). European consumers’ willingness to pay for U.S. beef in experimental auction markets. American Journal of Agricultural Economics  85( 2): 396– 405. Google Scholar CrossRef Search ADS   Almli, V. L., Verbeke, W., Vanhonacker, F., Næs, T. and Hersleth, M. ( 2011). General image and attribute perceptions of traditional food in six European countries. Food Quality and Preference  22( 1): 129– 138. Google Scholar CrossRef Search ADS   Aprile, M. C., Caputo, V. and Nayga, R. M.Jr. ( 2012). Consumers’ valuation of food quality labels: the case of the European geographic indication and organic farming labels. International Journal of Consumer Studies  36( 2): 158– 165. Google Scholar CrossRef Search ADS   Auger, P., Devinney, T. M. and Louviere, J. J. ( 2007). Using best-worst scaling methodology to investigate consumer ethical beliefs across countries. Journal of Business Ethics  70( 3): 299– 326. Google Scholar CrossRef Search ADS   Barber, S. L. and Gertler, P. J. ( 2009). Empowering women to obtain high quality care: evidence from an evaluation of Mexico’s conditional cash transfer programme. Health Policy and Planning  24( 1): 18– 25. Google Scholar CrossRef Search ADS PubMed  Basu, A. K. and Hicks, R. L. ( 2008). Label performance and the willingness to pay for Fair Trade coffee: a cross-national perspective. International Journal of Consumer Studies  32( 5): 470– 478. Google Scholar CrossRef Search ADS   Baumgartner, H. and Steenkamp, J. B. E. ( 2001). Response styles in marketing research: a cross-national investigation. Journal of Marketing Research  38( 2): 143– 156. Google Scholar CrossRef Search ADS   Bazzani, C. and Canavari, M. ( 2013). Alternative agri-food networks and short food supply chains: a review of the literature. Economia Agro-Alimentare  2: 11– 34. Google Scholar CrossRef Search ADS   Bech-Larsen, T. and Grunert, K. G. ( 2003). The perceived healthiness of functional foods. Appetite  40( 1): 9– 14. Google Scholar CrossRef Search ADS PubMed  Beck, M. J., Rose, J. M. and Greaves, S. P. ( 2017). I can’t believe your attitude: a joint estimation of best worst attitudes and electric vehicle choice. Transportation  44( 4): 753– 772. Google Scholar CrossRef Search ADS   Bonsmann, S. S. G. and Wills, J. M. ( 2012). Nutrition labeling to prevent obesity: reviewing the evidence from Europe. Current Obesity Reports  1( 3): 134– 140. Google Scholar CrossRef Search ADS PubMed  Budsjettnemda for jordbruket. ( 2015). Resultatkontroll for gjennomføringen av landbrukspolitikken (Result control for the agricultural policy). Report from the Agricultural Budget Commision, Oslo. Camarena, D. M., Sanjuán, A. I. and Philippidis, G. ( 2011). Influence of ethnocentrism and neo-phobia on ethnic food consumption in Spain. Appetite  57( 1): 121– 130. Google Scholar CrossRef Search ADS PubMed  Campbell, B. L., Mhlanga, S. and Lesschaeve, I. ( 2013). Perception versus reality: Canadian consumer views of local and organic. Canadian Journal of Agricultural Economics  61( 4): 531– 558. Google Scholar CrossRef Search ADS   Carlsson, F., Frykblom, P. and Lagerkvist, C. J. ( 2007). Consumer willingness to pay for farm animal welfare: mobile abattoirs versus transportation to slaughter. European Review of Agricultural Economics  34( 3): 321– 344. Google Scholar CrossRef Search ADS   Chern, W. S., Rickertsen, K., Tsuboi, N. and Fu, T. T. ( 2002). Consumer acceptance and willingness to pay for genetically modified vegetable oil and salmon: a multiple-country assessment. AgBioForum  5( 3): 105– 112. Cohen, E. ( 2009). Applying best‐worst scaling to wine marketing. International Journal of Wine Business Research  21( 1): 8– 23. Google Scholar CrossRef Search ADS   Czarnezki, J. J. ( 2011). Future of food eco-labelling: organic, carbon footprint, and environmental life-cycle analysis. Stanford Environmental Law Journal  30( 3): 3– 50. Darby, K., Batte, M. T., Ernst, S. and Roe, B. ( 2008). Decomposing local: a conjoint analysis of locally produced foods. American Journal of Agricultural Economics  90( 2): 476– 486. Google Scholar CrossRef Search ADS   de-Magistris, T. and Gracia, A. ( 2014). Do consumers care about organic and distance labels? An empirical analysis in Spain. International Journal of Consumer Studies  38( 6): 660– 669. Google Scholar CrossRef Search ADS   de-Magistris, T., Gracia, A. and Albisu, L. M. ( 2014). Wine consumers’ preferences in Spain: an analysis using the best-worst scaling approach. Spanish Journal of Agricultural Research  12( 3): 529– 541. Google Scholar CrossRef Search ADS   Dekhili, S., Sirieix, L. and Cohen, E. ( 2011). How consumers choose olive oil: the importance of origin cues. Food Quality and Preference  22( 8): 757– 762. Google Scholar CrossRef Search ADS   Delwaide, A. C., Nalley, L. L., Dixon, B. L., Danforth, D. M., Nayga, R. M., Van Loo, E. J. and Verbeke, W. ( 2015). Revisiting GMOs: are there differences in European consumers’ acceptance and valuation for cisgenically vs transgenically bred rice? PLoS ONE  10( 5): 1– 16. Google Scholar CrossRef Search ADS   Drichoutis, A. C., Lazaridis, P. and Nayga, R. M.Jr. ( 2006). Consumers’ use of nutritional labels: a review of research studies and issues. Academy of Marketing Science Review  9: 1– 25. Erdem, S., Rigby, D. and Wossink, A. ( 2012). Using best–worst scaling to explore perceptions of relative responsibility for ensuring food safety. Food Policy  37( 6): 661– 670. Google Scholar CrossRef Search ADS   Finn, A. and Louviere, J. J. ( 1992). Determining the appropriate response to evidence of public concern: the case of food safety. Journal of Public Policy and Marketing  11( 2): 12– 25. Flynn, T. and Marley, T. ( 2014). Best-worst scaling: theory and methods. In: S. Hess and A. Daly (eds), Handbook of Choice Modelling . Cheltenham, UK: Edward Elgar, 178– 201. Frewer, L. J., Risvik, E. and Schifferstein, H. (eds) ( 2013). Food, People and Society: A European Perspective of Consumers’ Food Choices . Berlin: Springer. Getz, C. and Shreck, A. ( 2006). What organic and fair trade labels do not tell us: towards a place-based understanding of certification. International Journal of Consumer Studies  30: 490– 501. Google Scholar CrossRef Search ADS   Golden, J. S., Vermeer, D., Clemen, B., Noyes, C. and Akella, A. ( 2010). An overview of ecolabels and sustainability certifications in the global marketplace. Corporate Sustainability Initiative – Nicholas Institute for Environmental Policy Solutions at Duke University  1( 10): 1– 99. Gracia, A., Barreiro-Hurlé, J. and Galán, B. L. ( 2014). Are local and organic claims complements or substitutes? A consumer preferences study for eggs. Journal of Agricultural Economics  65( 1): 49– 67. Google Scholar CrossRef Search ADS   Grunert, K. G., Hieke, S. and Wills, J. ( 2014). Sustainability labels on food products: consumer motivation, understanding and use. Food Policy  44: 177– 189. Google Scholar CrossRef Search ADS   Guerrero, L., Guàrdia, M. D., Xicola, J., Verbeke, W., Vanhonacker, F., Zakowska-Biemans, S. and Scalvedi, M. L. ( 2009). Consumer-driven definition of traditional food products and innovation in traditional foods. A qualitative cross-cultural study. Appetite  52( 2): 345– 354. Google Scholar CrossRef Search ADS PubMed  Hamada, N. ( 1973). On the p-rank of the incidence matrix of a balanced or partially balanced incomplete block design and its applications to error correcting codes. Hiroshima Mathematical Journal  3( 1): 153– 226. Hein, K. A., Jaeger, S. R., Carr, B. T. and Delahunty, C. M. ( 2008). Comparison of five common acceptance and preference methods. Food Quality and Preferences  19: 651– 661. Google Scholar CrossRef Search ADS   Hu, W., Woods, T. and Bastin, S. ( 2009). Consumer acceptance and willingness to pay for blueberry products with nonconventional attributes. Journal of Agricultural and Applied Economics  41( 1): 47– 60. Google Scholar CrossRef Search ADS   Ilbery, B., Morris, C., Buller, H., Maye, D. and Kneafsey, M. ( 2005). Product, process and place: an examination of food marketing and labelling schemes in Europe and North America. European Urban and Regional Studies  12( 2): 116– 132. Google Scholar CrossRef Search ADS   Jaeger, S. R., Jørgensen, A. S., Aaslyng, M. D. and Bredie, W. L. P. ( 2008). Best–worst scaling: an introduction and initial comparison with monadic rating for preference elicitation with food products. Food Quality and Preference  19( 6): 579– 588. Google Scholar CrossRef Search ADS   Kuhfeld, W. F. ( 2005). Experimental Design, Efficiency, Coding, and Choice Designs. Marketing Research Methods in SAS: Experimental Design, Choice, Conjoint, and Graphical Techniques . Iowa State University, IA: SAS Institute Inc., 53– 241. Lagerkvist, C. J. ( 2013). Consumer preferences for food labelling attributes: Comparing direct ranking and best–worst scaling for measurement of attribute importance, preference intensity and attribute dominance. Food Quality and Preference  29( 2): 77– 88. Google Scholar CrossRef Search ADS   Lagerkvist, C. J., Okello, J. and Karanja, N. ( 2012). Anchored vs. relative best–worst scaling and latent class vs. hierarchical Bayesian analysis of best–worst choice data: investigating the importance of food quality attributes in a developing country. Food Quality and Preference  25( 1): 29– 40. Google Scholar CrossRef Search ADS   Lancsar, E., Louviere, J., Donaldson, C., Currie, G. and Burgess, L. ( 2013). Best–worst discrete choice experiments in health: methods and an application. Social Science & Medicine  76( 1): 74– 82. Google Scholar CrossRef Search ADS   Lancaster, K. J. ( 1966). A new approach to consumer theory. The Journal of Political Economy  74( 2): 132– 157. Google Scholar CrossRef Search ADS   Lee, S. H., Han, D. B., Caputo, V. and Nayga, R. M. ( 2015). Consumers’ valuation for a reduced salt product: a nonhypothetical choice experiment. Canadian Journal of Agricultural Economics  63( 4): 563– 582. Google Scholar CrossRef Search ADS   Lee, J. A., Soutar, G. N. and Louviere, J. ( 2007). Measuring values using best–worst scaling: the LOV example. Psychology & Marketing  24( 12): 1043– 1058. Google Scholar CrossRef Search ADS   Lim, K. H., Hu, W., Maynard, L. J. and Goddard, E. ( 2013). US consumers’ preference and willingness to pay for country‐of‐origin‐labeled beef steak and food safety enhancements. Canadian Journal of Agricultural Economics  61( 1): 93– 118. Google Scholar CrossRef Search ADS   Loose, S. M. and Lockshin, L. ( 2013). Testing the robustness of best worst scaling for cross-national segmentation with different numbers of choice sets. Food Quality and Preference  27( 2): 230– 242. Google Scholar CrossRef Search ADS   Loureiro, M. L. and Umberger, W. J. ( 2007). A choice experiment model for beef: what US consumer responses tell us about relative preferences for food safety, country-of-origin labeling and traceability. Food Policy  32( 4): 496– 514. Google Scholar CrossRef Search ADS   Louviere, J. J., Street, D., Burgess, L., Wasi, N., Islam, T. and Marley, A. A. J. ( 2008). Modeling the choices of individual decision-makers by combining efficient choice experiment designs with extra preference information. Journal of Choice Modelling  1( 1): 128– 164. Google Scholar CrossRef Search ADS   Louviere, J. J., Flynn, T. N. and Marley, A. A. J. ( 2015). Best Worst Scaling: Theory, Methods and Applications . Cambridge, England: Cambridge University Press. Google Scholar CrossRef Search ADS   Louviere, J. J. and Woodworth, G. G. ( 1990). Best–worst scaling: a model for largest difference judgments. Working Paper, Faculty of Business, University of Alberta, Edmonton. Luckstead, J. and Devadoss, S. ( 2016). Impacts of the transatlantic trade and investment partnership on processed food trade under monopolistic competition and firm heterogeneity. American Journal of Agricultural Economics  98( 5): 1389– 1402. Google Scholar CrossRef Search ADS   Lusk, J. L. and Briggeman, B. C. ( 2009). Food values. American Journal of Agricultural Economics  91( 1): 184– 196. Google Scholar CrossRef Search ADS   Lusk, J. L., House, L. O., Jaeger, S. R., Moore, M., Morrow, J. L. and Traill, W. B. ( 2004). Effect of information about benefits of biotechnology on consumer acceptance of genetically modified food: evidence from experimental auctions in the United States, England, and France. European Review of Agricultural Economics  31( 2): 179– 204. Google Scholar CrossRef Search ADS   Lusk, J. L., Roosen, J. and Fox, J. A. ( 2003). Demand for beef from cattle administered growth hormones or fed genetically modified corn: a comparison of consumers in France, Germany, the United Kingdom, and the United States. American Journal of Agricultural Economics  85( 1): 1– 58. Google Scholar CrossRef Search ADS   Lähteenmäki, L. and Arvola, A. ( 2001). Food neophobia and variety seeking – consumer ear or demand for new food products. In: L. J. Frewer, E. Risvik and H. Schifferstein (eds), Food, People and Society . Berlin: Springer, 161– 175. Google Scholar CrossRef Search ADS   Marley, A. A. J. and Louviere, J. J. ( 2005). Some probabilistic models of best, worst, and best-worst choices. Journal of Mathematical Psychology  49( 6): 464– 480. Google Scholar CrossRef Search ADS   Martinez, S., Hand, M., Da Pra, M., Pollack, S., Ralston, K., Smith, T., Vogel, S., Clark, S., Lohr, L., Low, S. and Newman, C. ( 2010). Local Food Systems: Concepts, Impacts, and Issues . Washington, DC: US Department of Agriculture, Economic Research Service. McFadden, D. ( 1974). Conditional logit analysis of qualitative choice behavior. In: P. Zarembka (ed.), Frontiers in Econometrics . New York: Academic Press, 105– 142. Meas, T., Hu, W., Batte, M. T., Woods, T. A. and Ernst, S. ( 2015). Substitutes or complements? Consumer preference for local and organic food attributes. American Journal of Agricultural Economics  97( 4): 1044– 1071. Google Scholar CrossRef Search ADS   Mielby, L. H., Nørgaard, M. K., Edelenbos, M. and Thybo, A. K. ( 2012). Affective response of adolescents toward fruit‐ and vegetable-based snacks and the role of neophobia, gender and age. Journal of Sensory Studies  27( 6): 425– 438. Google Scholar CrossRef Search ADS   Mitchell, L. ( 2001). Impact of consumer demand for animal welfare on global trade. Changing Structure of Global Food Consumption and Trade. Department of Agriculture, Economic Research Service, Market and Trade Economics Division, Agriculture and Trade Report WRS-01-1. Washington, DC: US Department of Agriculture, Economic Research Service. Mitchell, L. ( 2004). US and EU consumption comparisons. In Normile and Leetmaa (eds), US-EU Food and Agriculture Comparisons. ERS, USDA, WRS-04-04, 49–65. Napolitano, F., Braghieri, A., Piasentier, E., Favotto, S., Naspetti, S. and Zanoli, R. ( 2010). Effect of information about organic production on beef liking and consumer willingness to pay. Food Quality and Preference  21( 2): 207– 212. Google Scholar CrossRef Search ADS   Napolitano, F., Pacelli, C., Girolami, A. and Braghieri, A. ( 2008). Effect of information about animal welfare on consumer willingness to pay for yogurt. Journal of Dairy Science  91( 3): 910– 917. Google Scholar CrossRef Search ADS PubMed  Nayga, R. M., Lipinski, D. and Savur, N. ( 1998). Consumers’ use of nutritional labels while food shopping and at home. Journal of Consumer Affairs  32( 1): 106– 120. Google Scholar CrossRef Search ADS   NIBIO. ( 2016). Utsyn over norsk landbruk, tilstand og utviklingstrekk 2016 (Norwegian agriculture: state and trends). NIBIO Bok 2(3), Oslo. Norsk landbruk. ( 2014). http://www.norsklandbruk.no/norsk-landbruk/ni-av-ti-vil-opprettholde-det-norske-jordbruket/. Accessed 22 March 2017. OECD. ( 2010). Agricultural policies in OECD countries: at a glance. http://www.oecd.org/tad/agricultural-policies/agriculturalpoliciesinoecdcountriesataglance.htm. Accessed 18 May 2016. OECD. ( 2016a). OECD.Stat. Expenditures at national prices in national currencies (millions). http://stats.oecd.org/viewhtml.aspx?datasetcode=PPP2011&lang=en. Accessed 26 August 2016. OECD. ( 2016b). OECD.Stat, Income distribution and poverty. http://stats.oecd.org/Index.aspx?DataSetCode=IDD. Accessed 26 August 2016. OECD. ( 2017). Inequality. http://www.oecd.org/social/inequality.htm. Accessed 22 March 2017. Onozaka, Y. and Mcfadden, D. T. ( 2011). Does local labeling complement or compete with other sustainable labels? A conjoint analysis of direct and joint values for fresh produce claim. American Journal of Agricultural Economics  93( 3): 693– 706. Google Scholar CrossRef Search ADS   Orme, B. K. ( 2005). Accuracy of HB estimation in MaxDiff experiments. Sequim, WA: Sawtooth Software. 98382(360). www.sawtoothsoftware.com. Accessed 13 April 2016. Pieniak, Z., Verbeke, W., Vanhonacker, F., Guerrero, L. and Hersleth, M. ( 2009). Association between traditional food consumption and motives for food choice in six European countries. Appetite  53( 1): 101– 108. Google Scholar CrossRef Search ADS PubMed  Progressive Dairyman. ( 2016). 2016 U.S. dairy statistics. http://www.progressivedairy.com/site/stats/us-dairy-stats. Accessed 13 May 2017. Rickertsen, K., Gustavsen, G. W. and Nayga, R. M.Jr. ( 2017). Consumer willingness to pay for genetically modified vegetable oil and salmon in the United States and Norway. AgBioForum  20( 2): 1– 11. Roininen, K., Lähteenmäki, L. and Tuorila, H. ( 1999). Quantification of consumer attitudes to health and hedonic characteristics of foods. Appetite  33( 1): 71– 88. Google Scholar CrossRef Search ADS PubMed  Rose, J. M. ( 2014). Interpreting discrete choice models based on best–worst data: a matter of framing. Transportation Research Board, Annual Meeting, 12–16 January, Washington DC. Rozin, P., Levine, E. and Stoess, C. ( 1991). Chocolate craving and liking. Appetite  17( 3): 199– 212. Google Scholar CrossRef Search ADS PubMed  Scarpa, R., Notaro, S., Louviere, J. and Raffaelli, R. ( 2011). Exploring scale effects of best/worst rank ordered choice data to estimate benefits of tourism in alpine grazing commons. American Journal of Agricultural Economics  93( 3): 809– 824. Google Scholar CrossRef Search ADS   Scarpa, R., Philippidis, G. and Spalatro, F. ( 2005). Product-country images and preference heterogeneity for Mediterranean food products: a discrete choice framework. Agribusiness  21( 3): 329– 349. Google Scholar CrossRef Search ADS   Schwartz, S. H. and Bilsky, W. ( 1987). Toward a universal psychological structure of human values. Journal of Personality and Social Psychology  53: 550– 562. Google Scholar CrossRef Search ADS   Schwartz, S. H. ( 1994). Are there universal aspects in the structure and contents of human values? Journal of Social Issues  50( 4): 19– 45. Google Scholar CrossRef Search ADS   Siro, I., Kapolna, E., Kapolna, B. T. A. and Lugasi, A. ( 2008). Functional food. Product development, marketing and consumer acceptance – a review. Appetite  51( 3): 456– 467. Google Scholar CrossRef Search ADS PubMed  Soederberg Miller, L. M. and Cassady, D. L. ( 2015). The effects of nutrition knowledge on food label use. A review of the literature. Appetite  92: 207– 216. Google Scholar CrossRef Search ADS PubMed  Statistica. ( 2017). ‘Average size of farms’. https://www.statista.com/statistics/196106/average-size-of-farms-in-the-us-since-2000/. Accessed 6 May 2017. Statistics Norway. ( 2017). ‘Population’. https://www.ssb.no/en/befolkning. Accessed 25 April 2017. Street, D. J. and Street, A. P. ( 1996). Partially Balanced Incomplete Block Design. In: C. J. Colbourn and J. H. Dinitz (eds), Encyclopedia of Biostatistics . Boca Raton, FL: CRC Press. Ter Hofstede, F., Steenkamp, J.-B. E. M. and Wedel, M. ( 1999). International market segmentation based on consumer-product relations. Journal of Marketing Research  36( 1): 1– 17. Google Scholar CrossRef Search ADS   Thomson, D. M. H., Crocker, C. and Marketo, C. G. ( 2010). Linking sensory characteristics to emotions: an example using dark chocolate. Food Quality and Preference  21( 8): 1117– 1125. Google Scholar CrossRef Search ADS   Thurstone, L. L. ( 1927). A law of comparative judgment. Psychological Review  34( 4): 273– 286. Google Scholar CrossRef Search ADS   Train, K. ( 2003). Discrete Choice Methods with Simulation . Cambridge, UK: Cambridge University Press. Google Scholar CrossRef Search ADS   United Stated Census Bureau ( 2017). ‘Population’. www.census.gov. Accessed 26 April 2017. Van Trijp, H. C. and Steenkamp, J. B. E. ( 1992). Consumers’ variety seeking tendency with respect to foods: measurement and managerial implications. European Review of Agricultural Economics  19( 2): 181– 195. Google Scholar CrossRef Search ADS   Vandemoortele, T. and Deconinck, K. ( 2014). When are private standards more stringent than public standards? American Journal of Agricultural Economics  96( 1): 154– 171. Google Scholar CrossRef Search ADS   Verbeke, W., Guerrero, L., Almli, V. L., Vanhonacker, F. and Hersleth, M. ( 2016). European consumer definition and perception of traditional foods. In: K. Kristbergsson and J. Oliveira (eds), Traditional Foods: General and Consumer Aspects . Boston, MA: Springer, 3– 16. Google Scholar CrossRef Search ADS   Author notes Review coordinated by Ada Wossink © Oxford University Press and Foundation for the European Review of Agricultural Economics 2018; all rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

European Review of Agricultural EconomicsOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off