Featured Article / Invited Peer-reviewed Article
What You Eat Reveals Who You Are:
Food Identity and Its Value Implications in Consumer Behavior
2025 Volume 45 Issue 1 Pages 5-13
Details
2025 Volume 45 Issue 1 Pages 5-13
We propose that food choice may serve as a categorical basis for an individual’s self-identity, meaning “what you eat reveals who you are.” This type of social identity, namely, food identity (FI), can be a significant determinant of consumers’ attitudes and behavior. Whereas previous studies on FI have focused mainly on the choice of eating meat, we maintain that concerns about health and the moral implications of eating meat may entail FI involving a broader range of food choices. A food identity scale was constructed to measure the strength of FI among Japanese individuals. A factor analysis revealed two components of FI: health-oriented identity and amusement-oriented identity. The former component of FI was found to connote liberal-leaning values, whereas the latter component was interpreted as a reactionary tendency against the “do-gooder” character of the former. The results, including the value-laden implications, were consistent with evidence from a preceding study analyzing social media data. The roles played by social identity and ethical values with ideological connotations in consumer behavior are discussed.
本研究は，食の嗜好が自己概念の一部である社会的アイデンティティーとしての役割を果たすという発想に基づいて，「フード・アイデンティティー」（以下，FI）の概念を提唱し，これが消費者の態度と行動に与える影響を吟味した。欧米を中心に行われてきた先行研究では，肉食とそれ以外との対比が主な論点であったが，本研究ではFI がより広範な健康意識および道徳意識と関連する可能性について検証した。新たに作成したFI尺度を用いて日本人を対象に行った調査の結果について因子分析を行った結果，「健康志向」および「娯楽志向」という2つの下位次元が特定されたた。このうち健康志向FIは，「リベラル」と特徴づけられるイデオロギー的価値意識との関連が見られたのに対し，娯楽志向FIは前者に対する反感を主な根拠とするものであることを示唆する結果が得られた。これは，計算社会科学的な手法を用いてソーシャル・メディア上での言説を分析した先行研究の結果と一貫するものであった。これらの結果をもとに，社会的アイデンティティーと倫理的価値観が消費者行動に与える示唆について論考した。
In recent years, it has become common to see people frequently taking photographs of their meals at restaurants and cafes and uploading them to social media. Clearly, the purpose of sharing what they eat with an indefinite number of people is not solely to make their food diaries public or to advertise these establishments. From a psychological perspective, a core part of such an act likely lies in the motivation to present aspects of their character implied by their food preferences (Bisogni et al., 2002; Rosenfeld & Burrow, 2018). In other words, displaying their food choices to the public is a form of self-presentation of their social identity (i.e., a part of the self-concept defined by membership in a social group or category, with evaluative significance attached to it; Tajfel & Turner, 1979). This type of social identity, what we call food identity (FI for short), draws on a consensual assumption that what people eat reflects who they are. Vegetarians and vegans are among the most widely recognized food-based identity categories, at least in industrialized Western societies. On the other hand, declaring oneself a meat eater would also emphasize one’s identity with an intended contrast to vegetarians.
From a marketing perspective, elucidating the structure and correlates of FI has important implications for segmentation strategies. Knowing the psychological bases, particularly those associated with consumers’ self-views and relevant beliefs under their dietary choice, would likely illuminate what to consider in marketing plans and decisions about which aspects of a food product to emphasize and which segments are the primary target.
In the present study, we aimed to examine the structure and implications of FI with a particular interest in its implications concerning social and ethical values. The analyses were not limited to the contrast involving meat eating but extended to a broader range of value-laden dietary choices.
1. Distinction: To eat or not to eat animal meatPsychological studies on FI, particularly in Western cultures, have centered on distinguishing between vegetarians with a wide range of variations (e.g., vegans) and omnivores. Past studies have revealed the contents of associations between each dietary orientation and specific traits, frequently with blatant evaluative connotations. These associations are often stereotypic and yet appear to constitute the core part of the social identity of each group. Being vegetarian tends to be associated with traits such as femininity, being White, self-restraint, and having high moral standards, whereas eating meat (and unhealthy food) tends to be associated with masculinity, being Black, relatively less concerned about being virtuous, and having higher sexual drive (Rosenfeld et al., 2023; Rozin et al., 2012; Ruby, 2012)1). Such stereotypic views are enhanced when seen from an outsider’s (i.e., the out-group’s) perspective, as typically found in research on intergroup cognition; however, similar tendencies are also prevalent among in-group members (i.e., self-stereotyping). In addition to the “from behavior to trait” inference process (i.e., from dietary choice to character evaluations), a reversal inference is also observed. For example, an individual who is described as generally desirable tends to be assumed to choose healthy rather than unhealthy food (König et al., 2017). In short, the evidence indicates that one’s food preference serves as a basis for their identity, either in the eyes of others or from their own perspective.
Among the trait associations of each camp of dietary choices, moral connotations have a strong effect on each type of FI. Avoiding animal meat tends to be regarded as desirable from the standpoint of health. Whereas the industrialization of food production in modern society has dramatically facilitated the accessibility of meat products, excessive intake of animal flesh has become a major health issue because of its unwanted consequences, namely, a high risk of fatal diseases such as cardiovascular illness. Doing something good or bad for one’s health inevitably becomes viewed in a moral sense (Ruby, 2012). This is the case across different cultural backgrounds: Christianity emphasizes the sacredness of the human body as a shrine, whereas Confucian ethics promote gratitude toward one’s parents for gifting them a healthy body at birth.
A different line of moral concern pertains to an aversion to killing animals. Ample empirical evidence shows that moral motives and the feeling of disgust associated with meat eating are major reasons people give for becoming vegetarians (Ruby, 2012). Notably, ethics of this kind are reflected in various religious creeds and practices, including Hinduism and Buddhism, at least in part. The strength of will, inferred from the act of self-restraint against temptations and indulgence in undesirable food products, may generalize to a comprehensive moral judgment about the overall character of vegetarians. The inclination toward associating vegetarianism with a higher moral standard, often observed in the views of both in- and out-group members, appears to draw on these different moral inferences. Additionally, from the perspective of environmental protection, meat production has come under strong criticism in recent years due to its hazardous consequences, such as the resources required to produce animal feed crops and the generation of massive methane gas emissions. Because eating animal flesh poses various ethical concerns, the association between vegetarianism and high moral standards is broadly facilitated.
Trait attributions to each food-based category serve as a basis not only for stereotyping by outsiders but also for the social identity of people belonging to each category. In their attempt to develop a Dietarian Identity Questionnaire (DIQ), Rosenfeld and Burrow (2018) revealed a variety of FI components. These components included motivations for dietary choice with prosocial, personal, or moral implications. Other components were closely related to the definition of social identity: centrality, positive self-regard, and negative out-group regard. Using this scale, Rosenfeld and Burrow demonstrated that vegetarians had higher ratings on their motivations, self-regard, and centrality of FI compared to omnivores. Hence, the overall configuration of the vegetarian identity was in accordance with stereotypical views.
Another noteworthy finding from Rosenfeld and Burrow’s (2018) study was that, despite their generally positive self-regard, vegetarians gave lower ratings for their public self-regard than omnivores. In other words, vegetarians acknowledged their potentially lower popularity among outsiders. This self-deprecating perception may, in fact, reflect the reality of the views of the general public. Although the moral worth of vegetarians may be accepted, ironically, this morally laden character makes vegetarian advocacy a challenge to traditional majority values, that is, showing little or no hesitation about eating meat. As such, vegetarianism may become perceived as a threat to the self-image of the nonvegetarian majority. Because people with strong moral convictions prefer distancing themselves from those with opposite views (Skitka et al., 2005), vegetarians may be seen as dogmatically separating themselves from the mainstream to the extent that their beliefs carry moral implications. As a consequence, omnivores may regard vegetarianism as threatening, and therefore, reactively depreciate it. The results from a study by Minson and Monin (2012) were consistent with this reasoning. The participants who identified themselves as meat eaters rated vegetarians as having more moral traits relative to their self-ratings, and yet, they estimated that vegetarians would show greater differences in morality ratings between the two groups. Furthermore, the larger the gap between their actual self-ratings and those estimated to be made by vegetarians, the lower the ratings the omnivore participants gave to words associated with vegetarians. Together, the results demonstrated that the meta-stereotype concerning vegetarians’ view of meat eaters threatened the latter by generating an anticipation of reproach, resulting in denigrating vegetarians as “do-gooders.” An open-ended response from one participant who stated, “I’m the antithesis of vegetarian,” illustrates the self-awareness of the defensive derogation process (Minson & Monin, 2012, p. 204).
The potency of this reactive tendency against “do-gooders” suggests that threat-based derogation may not be limited to the issue of meat eating. The apparent divergence in this specific dietary choice may be a mere reflection of a more comprehensive and fundamental contention about what the “right” foods are. In this sense, the underlying disparity between the two camps’ consumption behavior may pertain to their worldviews.
2. Beyond “meat or no meat”Vegetarians are a small minority, typically with a population ratio lower than 10%, even in countries that are assumed to be more “advanced,” such as Western countries (Buchholtz, 2022). Although the number may vary depending on the definition and across surveys, this number is notably low compared with the salience and impact of vegetarian advocacy. The essence of vegetarian values and reactions from the majority can include something other than the act of eating meat per se, reflecting a broader range of value connotations behind food choice.
A Japanese sociologist, Hayamizu (2013), proposed a distinction between different consumer orientations that he calls “food left-wing” versus “food right-wing,” with the former defined as “the people who try to bring food that has become an industrial product back to the side of agriculture to make it safe and secure again” and as “originating in counter-culture and having been through the process of commercialization” (p. 83). He used the word “left” to characterize this group because of their similar lifestyles to the politically liberal class, particularly American urban dwellers, with their stronger concerns about ethical consumption. He referred to Alice Waters’ Chez Panisse restaurant in California and a commune-like farm in suburban Japan to exemplify this ideological orientation. The opposing faction, the “food right-wing” individuals using his terminology, was characterized by their tendency to value low price and quantity (typified as “junk-oriented”) and to be less concerned about the potentially low quality of mass-produced food commodities. He depicted this group using the analogy of the conservative backlash against the “irritating liberals,” who appear as self-righteous and excessively idealistic (p. 74).
The significance of Hayamizu’s (2013) ethnological research lies in his insight that the contrast between the two camps of food preferences reflects a more deep-rooted structural conflict between different sets of values. He interpreted the “left-wing” value as a counterforce against industrialization and mass production, which was seen as threatening the fundamental value of human life, that is, food in this specific case. According to Hayamizu, when this orientation was viewed as overly politicized and moralized, a reactive movement, the “food right,” arose and dared to emphasize the enjoyment of consuming allegedly “wrong” food.
The contention by Hayamizu (2013) that each “wing” of food preferences presents ideological values has received support from an analysis of the discourse on the Twitter (currently “X”) social media platform conducted by Sasahara (2019). Using a machine learning technique and drawing on keywords derived from Hayamizu (2013), he identified 1,233 Japanese users who frequently used words associated with so-called left-wing (e.g., “vegetarian,” “organic,” and “slow food”) and 5,010 right-wing users (i.e., using keywords such as “fast food,” “junk food,” and “supersized portion”). Additionally, he developed another database by collecting the most recent posts (n=3,200 at maximum) from each “wing” to examine how these two groups were expressed in various value-laden domains. These domains were either directly related to food (e.g., “meat”) or suggestive of other ideological orientations, such as “environmental protection” and “liberal.” The degree of co-occurrence of these target words and the keywords that defined each “food wing” were analyzed as an index of “collective interest” among these Twitter users. The results revealed that those who were classified as food-left showed a greater interest not only in food (e.g., “genetically modified” and “trans fatty acids”) but also in other issues, such as health (“jogging,” “low fat”), and what was named socio-environmental (“eco,” “greenhouse gases,” “animal experiment,” “fair trade”). Their interest was directed toward health- and ethically conscious consumer behavior, resonating with their characterization as “left-wing.” In contrast, those in the other cluster presented opposing views, with higher frequencies of co-occurrences in their use of words such as “instant noodle,” “high calorie,” and, interestingly, “conservative” and “online right-wingers” representing the political domain. The overall pattern of each wing’s collective interest was consistent with Hayamizu’s characterization.
Moreover, an analysis of the pattern of users’ retweets (currently reports in X) revealed that users in the same camp, regardless of whether they had a left-wing or right-wing position, retweeted the comments of other users in the same wing more often than those from people on the other side. In other words, an “echo chamber” phenomenon was observed regarding what they communicated with others on the platform. Overall, the results of Sasahara (2019) demonstrated the collective interest among those who shared the same orientation in their food-related vocabulary both in their word usage across a broad range of ideologically laden domains and in their selectiveness in networking behavior on social media.
Findings from these ethnographic observations and social media analyses suggest that dietary choices reflect not only a personal preference regarding how to satisfy their hunger but also more abstract social values across broader domains, including those with moral and ideological connotations. It is plausible that such value-laden choices serve, at least in part, as a basis for defining self-identity. However, little is known about how FI is psychologically constructed and related to other beliefs and social values. Two major issues should be addressed for a thorough analysis.
First, the degree to which consumers internalize their food choices and other related events as a basis of their identity is not entirely clear in the previous findings. Sasahara (2019) discovered distinct clusters of social media users according to their word use, but direct measurement of their self-identification with each cluster of concepts remains necessary to substantiate the assumed psychological process of identification. As an additional note, the word count of specific words (e.g., “organic”) does not indicate their valence, such as “like” versus “dislike,” as is common in studies of social media in general. A direct measurement of identity should also solve this problem by revealing the strength of attachment versus detachment.
Second, empirical evidence concerning the validity of characterizing the assumed FI in ideological terms is limited. The terms “food-left” and “right” were used by Hayamizu (2013) as heuristic labels to represent each orientation analogically, yet there is no established empirical basis. As described earlier, Sasahara (2019) identified associations with words such as “conservative” among “right-wing” individuals, whereas the result was equivocal for the “left.” A direct measure reflecting an individual’s ideological position would clarify this issue, complementing the results of social media analyses. Importantly, establishing the operational definitions of FI, on the one hand, and attitudes toward consumer items that carry implications of value judgments, on the other hand, will allow us to test the hypothesized correlation between the two.
3. Constructing the food identity scaleThere are two possible approaches for assessing an individual’s self-reported FI. First, what we call a targeted approach might ask participants to categorically define their dietary patterns and then examine various qualities that may define their identity in the designated category. The study by Rosenfeld and Burrow (2018) using the DIQ, which we mentioned earlier, is a good example of this approach. These researchers presented the participants with five food types, such as “red meat,” “fish,” and “dairy,” to ask them which kinds of food they generally avoided eating, with another option allowing them to indicate, “I eat all of these.” Under the premise of this self-definition as to gradient levels of meat eating, participants responded to as many as 52 questions designed to assess different aspects of their identity derived from selected food categories. These aspects of identity detected by the DIQ included centrality, self-regard and regard for the other, morality, and so forth. This approach is useful for extensively examining the contents of social identity associated with a specifically targeted dietary pattern.
Alternatively, one could take a latent approach. We did so in the present study because our primary interest was in revealing an underlying value orientation that may manifest in the interest in different domains of food products and relevant activities. Specifically, we presented our participants with various concepts related to dietary choices, such as “vegetarians” and “fast food.” We then assessed the extent to which the participants construed each food item as a basis for their social identity. Because the presented concepts were selected mainly from the pool of keywords used by Sasahara (2019), this Food Identity Scale was expected to reveal whether separate orientations in food choice (characterized as “wings” in previous studies) can be found in FI along with individual differences in the strength of identification with each orientation. Additionally, to examine the ideological value connotation of each dietary orientation, we measured participants’ attitudes toward consumer products and activities that represent such values and analyzed the relationship between FI and these attitudes.
To summarize, whereas a previous analysis of word usage among social media users (Sasahara, 2019) revealed different orientations in food choice and the ideology-laden directions of interest among users, the present study was expected to provide psychological evidence of identification with such orientations and the attitudinal directions associated with each orientation along with their ideological connotations. Specifically, in the present study, we aimed (1) to develop a scale to measure the type and level of consumers’ social identity based on their dietary choices (i.e., FI) and (2) to test the hypothesis that different types of FI would predict attitudes toward consumer items with ideological value implications.
We recruited participants through the sampling service provided by Qualtrics LLC. Responses from 725 participants were obtained, but 16 were excluded from our analyses because they failed our attention-check tests. We included the remaining 709 (356 females, M age=47.98, SD age=10.74) participants in the final analyses.
2. Materials (1) Food identity scaleTo construct our FI Scale, we prepared a list of 12 food-related target concepts. To maximize comparability with previous studies, we selected as many concepts as possible from the keywords used by Sasahara (2019). Drawing on Hayamizu’s (2013) characterization of the so-called “food left,” we chose five target words that pertained to concerns with health and liberally minded ethicality, such as vegetarian and organic food, to represent this orientation (for a complete list, see Table 1). Because the other orientation, “food right,” was characterized as a counterforce against the vocal “left,” we chose seven target concepts that often appeared on TV shows and other media to depict unhealthy food, including fast food and B-class (meaning “second class”) gourmet. The target words representing the first group were outnumbered because a preliminary discussion with a pilot panel revealed that the words referring to that group tended to be highly technical and unrecognized, hence making it more difficult to develop a pool of easily comprehensible terms.
Mean ratings, factor loadings, and communalities of items on the food identity scale
a The reliability coefficient across the three questions for each item
b The reliability coefficient across the items for each subscale
Participants were presented with the target concepts and asked to rate each concerning: (1) the extent to which the word accurately described their self-identity; (2) the extent to which the word accurately represented their orientations and preferences concerning food (both with a range from “1=not at all” to “7=extremely accurate”); and (3) how often they had purchased or used the described item (“1=never, 7=extremely often”)2).
(2) Measures of consumer attitudes with ideological implicationsTo test the hypothesis that FI is associated with ideological values, we measured participants’ attitudes toward various products known to have ethical and, arguably, ideological connotations. The target items for the attitudinal ratings were chosen from the word pool Sasahara (2019) used to maintain comparability. Among the selected targets, items such as plastic bags and genetically modified food were used because they were unpopular, particularly among “progressive” or liberally minded Japanese consumers, due to their potentially unethical nature (for a complete list, see Table 2). The participants were asked to rate the extent to which they felt resistance to each item (1=not at all, 7=a great deal). In contrast, other items included in the attitudinal judgments were relatively popular and hence expected to receive consumers’ interest or desire to purchase rather than dislike or active exclusion (e.g., recycled goods and electric cars expected to be popular on the liberal side, as well as sports cars and luxury tobacco targeted toward people who enjoy consumption as a self-entertaining activity). They rated the extent to which they would experience favorable feelings toward each item (1=not at all, 7=a great deal). We aimed to avoid response bias by including the former items to measure resistance and the latter to measure favorableness.
Mean ratings, factor loadings, and communalities of items for attitudinal ratings
a Higher scores for the Ethical Opposition items indicate less favorable attitudes whereas those for the remaining items indicate more favorable attitudes.
b The value after excluding “whale meat” because of its low loading
The participants first completed the FI Scale. Next, they made a series of attitudinal judgments concerning the 13 products. They were then asked to indicate their ideological position on a more explicit measure of the conservative versus liberal stance in the social and economic domains. They read the definition of being socially conservative versus liberal (i.e., preferring “the conservation of traditional values over drastic changes” or “advancing the status of the disadvantaged rather than maintaining the status quo”) concerning issues such as the environment and human rights. They then indicated their position on a scale ranging from 1=extremely liberal to 7=extremely conservative. Likewise, they read the definition of economic conservative versus liberal (i.e., preferring “the market principle of free competition” or “equal distribution of wealth based on government interventions”) and then rated their self-position. Finally, they answered questions concerning demographic variables, including gender, age, and household income. After completing these tasks, they were debriefed and thanked for their participation.
We conducted an exploratory factor analysis on the scores from the 12 items of the FI Scale using the maximum likelihood method. Two factors with eigenvalues greater than 1.0 (4.29 and 3.21, respectively) emerged, and the scree plot also corroborated that a 2-factor solution was the most valid interpretation. Table 1 shows the factor loadings of the items based on a promax rotation, along with the mean rating of each item. Because examining whether each factor represented an ideological value was the main task in the present analysis, we attempted to exclude any ideological connotation in our interpretation of the factors at this stage. Without an ideological presupposition, Factor 1 was interpreted as representing a tendency to find amusement in self-identification with potentially undesirable food items. Hence, Factor 1 was defined as the amusement-oriented component of FI (AMUSE). Factor 2, on the other hand, was interpreted as a health-oriented component of FI (hereinafter HEALTH) because the items with high loadings mainly represented health concerns when viewed with ideological implications suppressed to a minimum. On the basis of these interpretations of these FI components, we calculated each participant’s subscale score by averaging their ratings of the seven items representing AMUSE and the five items representing HEALTH, respectively.
2. Attitudes toward consumer productsTo ascertain the target items used for attitudinal judgments assessed favorable versus unfavorable attitudes among consumers, we conducted an exploratory factor analysis using a maximum likelihood method. Three factors with eigenvalues greater than 1.0 were identified, and a scree plot analysis also suggested a three-factor solution. Table 2 shows the factor loadings after a promax rotation. Factor 1 was heavily loaded with unfavorable ratings for the intended cluster of unpopular products that typically appear in daily discourse, along with concerns about their potentially hazardous and/or unethical characteristics. Hence, we interpreted this as the “ethical opposition” (hereinafter OPPOSITION) to purportedly controversial consumer goods. As intended, Factors 2 and 3 pertained to favorable ratings of the products, with the former factor representing “pro-green” (GREEN) and the latter representing “pro-luxury” (LUXURY) attitudes. For each subcomponent of these attitudinal judgments, we averaged the ratings of the items (see Table 2 for the subscales)3).
We conducted a series of hierarchical multiple regression analyses to examine the extent to which FI contributed to each consumer attitude. For each attitudinal component, we first regressed the combined attitude rating on the two types of FI. We next included self-identified social and economic ideology (with higher scores representing a more conservative position) in the regression model. Finally, in the third step, we included demographic variables (see Table 3).
Regression coefficients (βs) for food identity, ideology, and demographic variables in predicting consumer attitudes
- “Amusement”=Amusement-oriented food identity; “Health”=Health-oriented food identity: “Social”=Socially conservative vs. liberal;
“Economic”=Economically conservative vs. liberal; “Income”=Household income.
a Positive coefficients indicate contributions to negative attitudes toward these items
b Positive coefficients indicate contributions to positive attitudes toward these items
c Ideology scores are coded so that higher numbers represent more conservative position
d Male=0, Female=1
* p<.05, ** p<.01, *** p<.001
Both the AMUSE and HEALTH components of FI consistently showed significant predictive power regarding attitudes toward consumer products across the different domains. Specifically, the HEALTH component was positively associated with OPPOSITION to ethically questionable goods, whereas AMUSE showed a negative association. The latter result indicates that those with stronger self-identity linked to unhealthy but seductive food exhibited a greater willingness to purchase those unpopular items. This result is consistent with our contention that the AMUSE component is not merely indifferent to ethical concerns but demonstrates an “anti-do-gooder” orientation. Overall, the difference between the HEALTH and AMUSE orientations appeared to be roughly parallel to the contrast between the “food left” and “food right” groups proposed by previous studies (i.e., Hayamizu, 2013; Sasahara, 2019), with the former implying progressive values and the latter reacting against them. Caution is necessary, however, because the explicit and direct measure of conservative versus liberal ideology resulted in a relatively small effect, if anything. Particularly notable was that the participants were informed of the definition of the conservative versus liberal distinction in the social domain, which specifically referred to environmental issues. Nevertheless, their self-defined social ideology had a null effect on their scores for OPPOSITION to supposedly hazardous items such as synthetic detergent and genetically modified food, which often appear in social activists’ critical claims. Instead, people with a liberal ideology concerning economic issues tended to oppose these items. This may reflect liberals’ stronger tendency to tolerate governmental intervention and regulations.
The results concerning attitudes toward the GREEN items revealed a less obvious pattern. The HEALTH component of FI showed a positive association with pro-environment consumer attitudes, again suggesting a relatively liberal tendency. However, those who scored high on the AMUSE subscale also showed more favorable attitudes toward these items. Here, too, the present data suggests that caution is necessary when viewing the difference between AMUSE and HEALTH in FI merely along the “right versus left” or “conservative versus liberal” dimensions. Indeed, self-defined ideology, either in the social or the economic domain, showed no unique relevance to this GREEN part of consumer attitudes.
The above results suggest ethical implications of HEALTH. Yet, this FI component is not exclusively based on self-restraint because it also showed a positive association with LUXURY items. Interestingly, the extent of positive associations with LUXURY did not differ between the HEALTH and AMUSE orientations. The association for HEALTH was not as self-evident as that for AMUSE, defined as a self-entertaining tendency in food consumption. These results suggest that, regardless of an orientation toward HEALTH or AMUSE, individuals who identify strongly with a style of food consumption may generally be inclined to spend more on expensive merchandise.
The present study provides evidence that dietary choice may serve as a basis for consumers’ social identity, the core part of their self-concept associated with the social group or category to which they perceive themselves to belong. We developed a concise questionnaire scale to assess FI and revealed its multi-dimensional structure. Using this scale, consistent with contentions and findings from previous studies, we discovered two distinct subcomponents of FI, namely, health-oriented and amusement-oriented identities. The unique finding from the present analysis is that the distinction may go beyond the dichotomous characterizations prevalent in previous investigations, such as “eat or do not eat meat” and “left-versus right-wing.” Instead, the relatively low inter-factor correlation observed in the present data suggests that the two components of FI likely represent two distinct orientations that may operate on one’s social identity in independent manners. Furthermore, FI is not confined to a limited contrast involving meat eating, which has attracted much previous research and attention from the general public but can encompass broader areas of value-laden choices.
Another important finding is that dietary choices may entail moral connotations. In particular, the HEALTH component of FI showed predictive power for consumer attitudes, representing their ethical oppositions to items allegedly hazardous to health and those in broader domains, such as environmental concerns. The results are consistent with the contention that the health-oriented FI can be characterized as leaning toward a relatively liberal ideology (Hayamizu, 2013; Sasahara, 2019).
In contrast, the ideological implications of the AMUSE orientation were less straightforward. Self-identification with supposedly unhealthy food by no means implies that those with this orientation lean toward suicidal ideation. Admittedly, those who scored high in AMUSE expressed reactive attitudes with lower scores in OPPOSE concerning ethically questionable products. Nevertheless, they showed more favorable attitudes toward green items such as solar panels and recycled goods. Notably, the self-defined conservative ideology in the social or economic domain showed little relevance to this attitudinal domain. It seems more appropriate to regard the AMUSE orientation in FI as a cynical or sarcastic counterforce against vocal ethics-minded liberals but not based on a conservative ideology per se. The peculiar but noteworthy characteristic of AMUSE appears to resonate with the recent political climate worldwide, particularly involving populist trends. One might suspect that it is likely not a coincidence that the reactionary surge of anti-establishment attitudes and anti-intellectualism is observed in a domain such as FI. Together, these results suggest the importance of not viewing different FI orientations as a simplistic dichotomy.
Because this study is only a preliminary investigation of FI, numerous limitations and issues for clarification remain. To point out a few, the relationship between each FI orientation and moral judgments should be pursued more vigorously. One possibility is to correlate them with various morality-related individual differences. Another possibility is to incorporate FI into experimental studies using moral scenarios and clarify the psychological processes that are assumed to underlie FI and moral values.
Additionally, because food is strongly tied to culture, thorough consideration should be given to the cultural background and implications of FI. The main issues addressed in the present study included the different dietary choices, either concentrated on meat eating or extended to more general domains, that originated in the Western cultural context. These issues are becoming common across cultures because health-related concerns and controversies regarding dietary choices are spreading globally due to global food production and cultural shifts. However, it should also be that different cultures have different concerns about food. Taking cultural context into consideration should provide better insight into FI, which is fine-tuned to potential differences in the structure and consequences of FI.
A similar argument may be addressed regarding the likelihood of cultural variation in values associated with food choices. Specifically, moral judgments and ideological values are clearly subject to individuals’ cultural backgrounds. For instance, concepts such as liberalism and conservatism, which are defined in the context of Western countries, have notably different meanings from those used in the Japanese context. We designed the present study as a starting point to maintain the common issues and research questions derived from previous research, but further investigation will inevitably need to incorporate culture-specific aspects of the concepts into their theoretical framework and empirical analyses. The present findings should serve as an initial yet significant step toward such research development.
It is a daily observation that social categorization profoundly impacts social life. Moreover, identities involving factors such as race, gender, and social class exert enormous influence. The domain of consumer attitudes and behavior is not an exception (Oyserman, 2009). Compared with those impactful categorizations, differences in issues such as food preference may appear to play a less significant role. However, since food is essential to human life, its importance cannot be negligible, especially when it comes to consumer choices and decisions. Furthermore, as discussed throughout this paper, food choice goes beyond a simple matter of nutrition. Dietary choices serve as a basis for social connections such as conversations over the dinner table and, moreover, may invite discussions of related issues such as morality and religion. The importance of food identity is expected to increase rather than decrease in the foreseeable future.
The raw data and the materials in Japanese are available from the first author upon request.
1) In recent Japanese discourse, the neologism “herbivore boys”（草食男子）appears to refer to young males who are relatively low in activity level and sex drive. In contrast, the word “carnivorous”（肉食系）typically relates to women with high levels of energy and sexuality. The fact that both connote counter-stereotypic cases of their gender roles illustrates that the opposite associations represent traditional views.
2) They also rated two filler items, steamed white rice and Chinese food, which were both assumed to be irrelevant to either camp of food orientation.
3) We deleted “whale meat” from the OPPOSITION subscale because of its low factor loading. An analysis including this item did not change the conclusion reported here.
Minoru Karasawa
Dr. Karasawa is a Research Professor and an Emeritus at Nagoya University. His primary area of research is social psychology, and he earned his Ph.D. from UCLA in 1991. He is currently the President of the Asian Association of Social Psychology.
Kazutoshi Sasahara
Dr. Sasahara is a professor at the Tokyo Institute of Science. His primary area of research is computational social science, and he completed his Ph.D. at the University of Tokyo. He is a board member of the Society for Computational Social Science of Japan and an Associate Member of the Science Council of Japan.
Hakche Ryu
Mr. Ryu is a Manager at Le-Techs Inc. He received his Master’s of Psychology from the Graduate School of Environmental Studies, Nagoya University, majoring in social psychology. He studied political attitudes, ideology, and moral judgments.
https://ror.org/00hhkn466 
