Abstract
This study examines the role of ethnic cuisine as a potential mechanism for fostering outgroup tolerance. Data were collected via a web survey administered to a stratified, nationally representative sample of White adults from the United Kingdom (N = 1,025). Results showed that greater enjoyment and frequency of ethnic food consumption were significantly associated with lower perceptions of immigrants as cultural and economic threats and with decreased likelihood of voting for anti-immigrant politicians. Mediation analyses further indicated that these associations were largely indirect, operating through increased positive interactions with foreigners. These effects remained robust after controlling for pertinent factors, including right-wing political orientation, selective attention to conservative newspapers, region, settlement type, and minority group friendships. Thus, overall, the findings suggest that exposure to ethnic cuisine may serve as a subtle yet meaningful catalyst for direct intergroup contact, capable of reducing hostility towards immigrants. As such, this study contributes novel theoretical and contextual insights into the social-psychological literature on intergroup contact and conflict.
Introduction
Recent sharp increases in immigration to most Western countries have often been accompanied by rises in perceptions of outgroup threat, leading to greater voter support for punitive immigration policies (McLaren, 2003; Rios et al., 2018) and for authoritarian populist politicians among majority-ingroup members (Ramos et al., 2022). In the United Kingdom (UK), for example, anti-immigrant sentiments are widely believed to be a primary driver of the 2016 Brexit outcome (Golec de Zavala et al., 2017; Goodwin & Milazzo, 2017; Hutchings & Sullivan, 2019). More recently, in August 2024, said animosity manifested in nationwide violent race riots directed at asylum seekers and minority communities at large (Duncan et al., 2014).
This context of heightened anti-immigrant sentiment provides a salient backdrop for examining factors that may foster more inclusive attitudes in majority-ingroup members. Congruently, an extensive body of research shows that intergroup interactions can decrease anti-immigrant sentiments (Bond & Tejeiro, 2019; Green et al., 2020; Jolly & DiGiusto, 2014; White et al., 2021). However, although the contact hypothesis is well supported at a general level, there is limited research on how to generate intergroup contact in the first place (Kauff et al., 2021). Indeed, as Pettigrew (2021, p. 264) points out, understanding the “process of encouraging people to have intergroup contact is the next big practical task for social psychologists in this area.”
Accordingly, this article introduces and empirically examines a previously unexplored mechanism for catalysing positive and sustained intergroup contact: ethnic food engagement. 1 We begin with an overview of intergroup contact theory and then connect this to our proposed mechanism. Next, we describe our UK survey design and findings, and conclude with a summary of this study’s theoretical contributions, policy implications, and limitations.
Intergroup Contact Theory
Allport’s (1954) original contact hypothesis posited that interactions between different social groups are likely to decrease prejudice and improve intergroup relations if four conditions are met: (1) individuals in the contact situation are of equal status, (2) they are required to cooperate, (3) they share common goals, and (4) the contact is legitimized through institutional support. While contact situations meeting these four conditions appear to reduce prejudice more effectively than those that do not, contemporary contact theorists argue that these conditions should be viewed as facilitators rather than strict requirements. This perspective reflects the understanding that several complex factors (e.g., contact setting, individual openness, sociopolitical context, intergroup history) moderate contact effects (Kauff et al., 2021).
Correspondingly, although weakened, significant contact effects have been observed in multiple studies even when the four outlined conditions are not fulfilled (Kokkonen et al., 2016). For example, a meta-analysis of 515 studies by Pettigrew and Tropp (2006) found that relationship between direct contact and the reduction of various types of prejudice yielded an effect size of (r = −.21). A subsequent meta-analysis by Davies et al. (2011) further showed that contact that involved friendships with minority members, had a stronger effect in diminishing prejudice. Moreover, McLaren’s (2003) multivariate analysis of 17 Western European countries with varying levels of immigration, showed that intimate contact with outgroup members in the form of friendships reduced the level of willingness to expel legal immigrants. This study also showed that such contact helped to lower the perceived threat of immigrants in high immigration contexts. Similarly, a representative survey study of German participants found that the initiation of positive contact decreased anti-Muslim attitudes even amongst groups who avoid or limit their contact with foreigners for example, the elderly and authoritarians (Pettigrew, 2008).
Despite these consistent findings, less theorizing has been done to explain and specify their underlying processes and contexts. However, recent scholarship from contact theory and parallel research fields has begun to shed light on what some of these processes and contexts might entail. To wit, several studies consistently show that mere exposure to outgroup members alone enhances liking and reduces feelings of uncertainty or threat (Flores et al., 2018; Lee et al., 2004; Zebrowitz et al., 2008). Research by MacInnis and Page-Gould (2015) further shows that whilst intergroup contact is often awkward and stressful at inception, as contact continues, a threshold is reached where initial stress diminishes, and favourable outcomes emerge. Other related studies indicate that successful intergroup contact situations are those that foster repeated, meaningful interactions and lead to subsequent friendships (Davies et al., 2011). Together, these findings suggest that initial exposure to outgroup members can encourage ongoing contact, thereby reducing intergroup threat and anxiety, which, in turn, may lower prejudicial views and practices.
Ethnic Food as a Catalyst for Positive Intergroup Contact
Furthermore, intergroup contact, when expressed as an intentional and active choice by the individual, has stronger effects than incidental contact that occurs under circumstantial conditions. However, as Bagci et al. (2021) argue, additional research is also needed to identify factors that induce contact-seeking behaviour in general and volitional contact more specifically. To this end, we propose that ethnic food engagement has the potential to function as such a mechanism, for the following reasons.
First, ethnic foods can be broadly defined as non-native traditional cuisines introduced to a host country by migrant communities. Zanger (1996) notes that immigrant families tend to maintain the cuisines from their culture and country of origin because these provide a source of stable identity. This is particularly true for recent immigrants, who are surrounded by the unfamiliar customs of their host country. Ethnic foods are thus considered a cultural bridge, connecting people from different ethnicities, races, and cultures (Barbas, 2003). Second, people are often interested in foods from other countries, especially diners seeking unique dining experiences. Moreover, according to Roseman (2006), restaurants are regularly the first avenue through which consumers are introduced to foods from different cultures. Tsai and Lu (2012) additionally note that people who wish to learn about different cultures tend to do so by patronizing ethnic restaurants and are likelier to return if they have a good initial impression and experience. It hence follows that ethnic restaurants play an important role in cultural transmission (Zelinsky, 1985). This likely holds true for other ethnic food outlets like markets, festivals, and cultural events, since these settings may involve more personal interaction and social learning.
Third, studies from the UK and the United States show that ethnic cuisines are increasingly popular, appealing to a diverse demographic and thriving in both metropolitan and non-metropolitan areas (Jang & Ha, 2015; Roseman, 2006). In other words, it is not just cosmopolitan, educated urban elites who frequent such establishments. There are several reasons for this, including a desire to eat something different, the convenience of dining out, and increased disposable income. Fourth, according to a British study by Warde et al. (1999), the more a person ate out, the likelier they were to try ethnic foods. This study also found that a person’s consumption of one type of ethnic food increased their likelihood of trying other types, thereby expanding their interactions with foreign cultures.
Fifth, ethnic food vendors facilitate boundary-crossing in a way that other outgroup cultural institutions and mediums do not, providing a non-threatening means for ingroup members to engage amicably, regularly, and voluntarily with foreigners (Barbas, 2003). Arguably, these commercial settings also enable ingroup members to directly observe and experience the friendliness and strong work ethic of outgroups, possibly making it easier to recognize and relate to their shared humanity. Parenthetically, the intergroup contact that ethnic food vendors can foster is also likely to meet Hewstone et al. (2014, p. 41) stipulation that “if the contact is arranged so that it takes place between ingroup and outgroup members who can be regarded as sufficiently typical or representative of their groups, then the positive changes that occur should generalize to the groups as a whole.”
Finally, the consumption of ethnic food often entails a pleasurable sensory experience that can become firmly embedded in memory and lead to the formation of a positive association with the respective outgroup culture. That is, the proposed ethnic food–culture–cognitive mechanism operates through affective and sensory pathways that underpin intergroup perception. Ethnic food consumption is not only a sensory pleasure but also a symbolic entry point into a foreign culture, often experienced in informal, low-threat environments. Such sensory experiences can evoke positive emotions that become cognitively associated with the outgroup culture, subtly shifting attitudes. Furthermore, repeated exposure to pleasurable ethnic food experiences may foster a sense of familiarity and reduce the psychological distance between ingroup and outgroup members—a process supported by the mere exposure effect and heuristic-based processing in intergroup cognition (Flores et al., 2018; Isbell et al., 2016). In this way, food serves not just as a cultural artefact but as a psychologically and potentially potent medium for shaping intergroup attitudes. Therefore, based on everything discussed, we posit the following hypotheses:
Method and Materials
Data were collected in the United Kingdom during October 2022 through an online questionnaire. The opening information page/consent form stated that the survey aimed to gain a broad overview of leisure activities, cultural tastes, group dynamics, and public opinion in the UK. Additionally, several distractor questions related to UK pop culture were interspersed between the measures for the predictor and outcome variables. Given that immigration is a highly salient and controversial topic, said design features were implemented to mask the study’s purpose and minimize priming and demand effects. 2
To further ensure data quality and generalizability, we implemented the following protocols. (1) A stratified random sampling quota was used to obtain representative figures in terms of gender, region/country, education, and age. (2) Participants were informed that their responses would be completely anonymous and confidential. (3) All attitudinal and behavioural ordinal measures used 4-point forced-choice response scales. (4) Three discrete attention filters/validity checks were evenly distributed across the survey, and participants who failed any of these had their sessions automatically terminated and data deleted. (5) Only participants with a 100% completion rate were included in the analysis. (6) Of the 1,040 participants who successfully completed the survey, seven participants selected the N/A option for all the food engagement measures described in the Predictor Variables section below, and eight participants consistently evidenced satisficing (e.g., by straight-lining item batteries and/or being in the top 5% of the fastest response times). These 15 participants were therefore removed from the final analysed sample.
Sample
The surveying agency Qualtrics was commissioned to recruit a nationally representative sample of White British adults aged 18 and over. Notably, the contact-prejudice link is significantly stronger for majority ingroup members than for members of disadvantaged groups (Pettigrew & Tropp, 2006). Said ethnic/racial demographic was thus targeted because it is the dominant native ingroup in the UK, with around 76% of the population identifying as “White British.” Following Bond and Tejeiro’s (2019) sampling rationale for a related study, White Britons were also selected since we focussed on measuring racial prejudice towards Asian, Afro-Caribbean and non-UK EU individuals, and because intergroup contact for non-White UK residents, would encompass contact with other racial groups including White British natives.
In total, 1,025 participants were included in the analysis. Of these, 53.9% identified as female, 45.7% as male, and 0.4% as other. Additionally, 24.3% hailed from Northern England, 26.1% from the Midlands, 23.2% from Southern England, 10.9% from Greater London, 5.0% from Wales, 8.4% from Scotland, and 2.1% from Northern Ireland. For their highest educational qualification, 20.4% of participants reported having a General Certificate of Secondary Education or equivalent, 25.6% an A-level qualification or equivalent, 14.6% a Higher National Certificate or Diploma, 28.4% an undergraduate degree, and 10% a postgraduate degree. The average age was 46.7 years (SD = 16.1). These figures roughly reflect national demographic averages.
Outcome Variables
Perceived Cultural & Economic Threat of Immigrants
This was measured with reformulated statements of five items from McLaren’s (2003) immigrant threat scale, which are themselves variations of items used in several xenophobia/anti-immigrant attitude scales (see e.g., Hjerm, 1998; Jolly & DiGiusto, 2014). These reformulations, in part, included: “In schools where there are many children from immigrants, the quality of education suffers”; “The religious practices of immigrants threaten our way of life”; and “Immigrants bring down wages for native British workers?.” Participants indicated the extent they agreed (1 = Strongly Disagree to 4 = Strongly Agree) with each statement. The scores were averaged to create a five-item composite variable (M = 2.18, SD = 0.69, a = 0.88).
Furthermore, to test whether the hypothesized predictor’s effects on general attitudes towards immigrants differ from attitudes towards specific migrant groups as well as to ensure consistent responses, we employed and positively reframed three items from Theorin’s (2022) different immigrant threat measures. Participants were asked to indicate the extent to which they agreed (1 = Strongly Disagree to 4 = Strongly Agree) with statements about immigrants from Africa or the Caribbean: (1) “are good for the UK economy,” (2) “enrich cultural life in the UK,” and (3) “DO NOT increase crime.” This same three-item set was used twice more with reference to European and Asian immigrants, respectively. The scores for each three-item set were averaged to create the following composite variables: Pro-Afro-Caribbean Immigrant Attitudes (M = 2.79, SD = 0.66, a = 0.87), Pro-European Immigrant Attitudes (M = 2.77, SD = 0.68, a = 0.84), and Pro-Asian Immigrant Attitudes (M = 2.84, SD = 0.66, a = 0.87).
Voter Support For Anti-Immigrant Politicians
This was measured with four items—two were self-developed and two were borrowed from Leyva’s (2023) intention to vote for an authoritarian populist measurements. Specifically, participants were asked to indicate their likelihood (1 = Very Unlikely to 4 = Very Likely) of voting for a politician who will genuinely work to: (1) “Cut immigration rates”; (2) “Increase government efforts and resources to deport illegal immigrants”; (3) “Remove immigrants’ access to benefits (i.e., welfare),” and (4) “Enact stricter rules and requirements for immigrants to settle in the UK.” The scores were averaged to create a four-item composite variable (M = 2.45, SD = 0.87, a = 0.93).
Predictor Variables
The list of ethnic food categories itemized below is not exhaustive nor unavoidably, entirely free of subjective interpretation. However, it was selected based on probabilistic criteria for enhancing positive intergroup contact in the British context. Specifically, the list reflects ethnic foods that are relatively popular and sold primarily in affordable restaurants and market kiosks. Additionally, to develop more representative measures, we selected foods from various global regions, cooked by some of the largest non-native ethnic or racial groups residing in the UK. Lastly, to better gauge participants’ likelihood of exposure to a particular foreign culture, we chose cuisines that are typically authentic and prepared and served by members of an associated non-native outgroup. For example, Italian cuisine was excluded because of the strong presence of British chains of Italian restaurants that are rarely staffed by Italians. Similarly, while Japanese and Mexican cuisines are popular and generally authentic, they were excluded because the associated foreign groups are comparatively smaller in the UK. We, therefore, developed the following two measures of ethnic food engagement based on exposure to six cuisine types: Indian, Turkish, Chinese, Thai, Caribbean, and Spanish. To assist with recognition, these broad categories were listed alongside two examplars of well-known meals served in the UK for example, “Indian (e.g., Tikka Masala, Madras).”
Ethnic Food Enjoyment
This measure was developed to gauge sensory pleasure, which we propose as a key psychological mechanism through which ethnic food experiences can transform social attitudes. Participants indicated their level of enjoyment for each of the six selected cuisines (1 = Strongly Dislike to 4 = Strongly Enjoy). Since some participants may have never tried certain cuisines, they could also select an “N/A (I’ve Never Tried It)” option, coded at −99. All these measures minus the N/A responses were summed to create a six-item additive index that demonstrated acceptable reliability (M = 16.97, SD = 4.6, α = .79). To assess the internal consistency and dimensional structure of this construct more rigorously, we conducted an Exploratory Factor Analysis (EFA) with a maximum likelihood extraction method. All six items loaded onto a single factor (eigenvalue = 2.9) that explained 48.7% of the variance, and all factor loadings were greater than 0.40. These results indicate that our additive scale for ethnic food enjoyment is psychometrically sound and internally consistent.
Ethnic Food Consumption Frequency
Although enjoyment of an ethnic food is key to fostering ongoing consumption and, consequently, intergroup contact, enjoyment does not necessarily imply frequent exposure. This is because various factors (e.g., financial constraints, restaurant availability) may limit the regularity to which individuals consume certain ethnic cuisines. Therefore, to gain a more comprehensive measure of ethnic food engagement, participants were also asked to indicate their frequency of eating each of the six aforementioned cuisines (1 = Never to 4 = Very Often, or N/A coded at −99). These measures sans the N/A responses were summed to create a six-item additive index, which demonstrated acceptable reliability (M = 12.38, SD = 3.8, α = .76). A follow up EFA, showed that this construct also has strong unidimensionality, as all items loaded onto a single factor (eigenvalue = 2.6) that explained 44.7% of the variance, and all factor loadings were between 0.42 and 0.67.
Mediator Variable
Quality of Intergroup Contact
This was measured with a self-developed item that asked participants to indicate their level of agreement (1 = Strongly Disagree to 4 = Strongly Agree) with the statement: “Most of my interactions with foreign people in the UK are generally positive” (M = 3.2, SD = 0.58).
Individual-Level Controls
The following antecedents have been shown to modulate beliefs and attitudes about, and/or direct contact with, immigrants. They were thus, in addition to the sample demographic variables, controlled for in the relevant analyses.
Political Orientation
About 21.5% of participants identified as right-wing/conservative (coded at 3), 49.4% as centrist/moderate (coded at 2), and 29.1% as left-wing/liberal (coded at 1).
Party Affiliation
About 29% of participants self-affiliated with the Tories, 37% with Labour, 9.8% with the Liberal Democrats, 3.7% with the Scottish National Party, 1% with the British National Party, 6.8% selected Other, and 12.8% did not identify with any political party. A dummy variable was created from these measures to control for Right-Wing Party Support (with Tory and British National Party participants coded at 1 and all others at 0).
Right-wing Newspaper Consumption
Participants noted their frequency (1 = Never to 4 = Very Often) of receiving news from The Sun, The Daily Mail, and The Daily Express. These measures were averaged to create a three-item composite variable (M = 1.78, SD = 0.72, a = 0.70).
Employment Status
About 46.4% of participants reported as being full-time employed, 20% as part-time employed, 13.6%, as not employed, 20% as retired.
Perceived Neighbourhood Diversity
Participants indicated the extent (1 = Not Diverse to 4 = Very Diverse) of their living area’s cultural diversity (M = 2.52, SD = 0.86).
Settlement Type
About 21.3% of participants lived in a city, 54.2% in a suburb, and 24.5% in a rural area.
Number of Outgroup Friendships
Participants indicated (0, 1 = 1–2, 2 = 3–4, 3 = 5–6, 4 = more than 7) their approximate number of friends from a foreign or non-White British background (M = 1.4, SD = 1.3).
Results
We began the analysis by running simple one-tailed Spearman’s Rho correlations between all predictor and outcome variables. As shown in Table 1, Ethnic Food Enjoyment and Ethnic Food Consumption Frequency were each significantly correlated with all five outcome variables in the predicted directions. Specifically, both ethnic food engagement variables were significantly negatively correlated with the Perceived Cultural and Economic Threat of Immigrants and Voter Support for an Anti-Immigrant Politician. They were also positively and significantly correlated with Pro-Afro-Caribbean Immigrant, Pro-European Immigrant, and Pro-Asian Immigrant Attitudes.
Zero-Order Spearman’s Correlations For All Outcome & Predictor Variables.
Correlation is significant at the .01 level (one-tailed).
To further test our first hypothesis (H1), we employed multiple regression to determine whether these associations held when controlling for all the demographic and covariate measures listed earlier. This analysis produced 10 models (five per predictor variable; see Tables 2 and 3 for full results). In Table 2, Model 1 (M1) explained 20% of the variance (Adjusted R2 = .20, F = 22.77, p < .001) and showed that, over and above the control variables, Ethnic Food Enjoyment (β = −.10, p = .001) predicted a significant decrease in the Perceived Cultural and Economic Threat of Immigrants. Model 2 (M2) explained 15% of the variance (Adjusted R2 = .15, F = 16.32, p < .001) and showed that Ethnic Food Enjoyment (β = .12, p < .001) predicted a significant increase in Pro-Afro-Caribbean Immigrant Attitudes. Model 3 (M3) explained 18% of the variance (Adjusted R2 = .18, F = 19.92, p < .001) and showed that Ethnic Food Enjoyment (β = .11, p < .001) predicted a significant increase in Pro-European Immigrant Attitudes. Model 4 (M4) explained 14% of the variance (Adjusted R2 = .14, F = 14.64, p < .001) and showed that Ethnic Food Enjoyment (β = .14, p < .001) predicted a significant increase in Pro-Asian Immigrant Attitudes. Model 5 (M5) explained 29% of the variance (Adjusted R2 = .29, F = 35.97, p < .001) and showed that Ethnic Food Enjoyment (β = −.07, p = .007) predicted a significant decrease in Voter Support for Anti-Immigrant Politicians. Interestingly, in Models 1, 2, and 3, only education yielded a stronger positive effect on greater tolerance towards immigrants than ethnic food enjoyment, and in Models 4 and 5, both education and ethnic food enjoyment were tied in being the strongest predictors of outgroup tolerance.
Effects of Ethnic Food Enjoyment on Anti-Immigrant Attitudes & Policy Support.
Note. All estimates are standardized coefficients, with standard errors in parenthesis.
p < .05. **p < .01. ***p < .001.
Effects of Ethnic Food Consumption Frequency on Anti-Immigrant Attitudes & Policy Support.
Note. All estimates are standardized coefficients, with standard errors in parenthesis.
p < .05. **p < .01. ***p < .001.
The first model in Table 3 (Model 6, M6) explained 19% of the variance (Adjusted R2 = 0.19, F = 21.85, p < .001), and showed that Ethnic Food Consumption Frequency (β = −.06, p = .04) predicted a significant decrease in the Perceived Cultural and Economic Threat of Immigrants. Model 7 (M7) explained 14% of the variance (Adjusted R2 = .14, F = 15.22, p < .001) and showed that Ethnic Food Consumption Frequency (β = .07, p = .01) predicted a significant increase in Pro-Afro-Caribbean Immigrant Attitudes. Model 8 (M8) explained 18% of the variance (Adjusted R2 = .18, F = 19.95, p < .001) and showed that Ethnic Food Consumption Frequency (β = .12, p < .001) predicted a significant increase in Pro-European Immigrant Attitudes. Model 9 (M9) explained 13% of the variance (Adjusted R2 = .13, F = 14.14, p < .001), and showed that Ethnic Food Consumption Frequency (β = .13, p < .001) predicted a significant increase in Pro-Asian Immigrant Attitudes. Lastly, Model 10 (M10) explained 29% of the variance (Adjusted R2 = .29, F = 35.81, p < .001), and showed that Ethnic Food Consumption Frequency (β = −.07, p = .01) predicted a significant decrease in Voter Support for Anti-Immigrant Politicians. Notably, ethnic food consumption frequency was the second strongest predictor of pro-European and pro-Asian immigrant attitudes after education and was tied with education and number of non-white friends in having the strongest negative effect in the likelihood of voting for anti-immigrant politicians. Overall, these correlations and regression models support H1.
To test H2, we employed a bootstrapping approach for estimating indirect effects using Hayes (2013) PROCESS macro (model 4) with 10,000 bootstrap estimates for 95% confidence intervals (CIs). First, the Perceived Culture and Economic Threat of Immigrants (Y), Quality of Intergroup Contact (M), and Ethnic Food Enjoyment variables (X) were entered as the outcome, mediator, and predictor variables, respectively, with demographic covariates controlled, continuous variables mean-centred, and all coefficients fully standardized. 3 We then repeated this procedure for each of the other four outcome variables to produce five mediation models. The results presented in Table 4, show that Ethnic Food Enjoyment had a significant negative total effect on viewing immigrants as cultural and economic threats, which was fully mediated by Quality of Intergroup Contact. It also had a significant positive total effect on Pro-Afro-Caribbean Immigrant Attitudes partially mediated by positive encounters with foreigners; and a significant positive total effect on Pro-European Immigrant Attitudes that was fully mediated by positive encounters with foreigners. For Pro-Asian Immigrant Attitudes, Ethnic Food Enjoyment had a significant positive total effect partially and significantly mediated by positive encounters with foreigners. Lastly, Ethnic Food Enjoyment had a significant negative total effect on Voter Support for Anti-Immigrant Politicians fully mediated by positive encounters with foreigners.
Indirect Effects of Ethnic Food Enjoyment on Sentiments Towards Immigrants Via Quality of Intergroup Contract.
p < .05. **p < .01. ***p < .001.
The analysis was repeated with Ethnic Food Consumption Frequency entered as the predictor and Quality of Intergroup Contact as the mediator. The results of these five medation presented in Table 5, show that Ethnic Food Consumption Frequency had a significant negative total effect on the Perceived Cultural and Economic Threat of Immigrants, with the effect fully mediated by the Quality of Intergroup Contact. It also had a significant positive total effect on Pro-Afro-Caribbean Immigrant Attitudes fully mediated by the Quality of Intergroup Contact; a significant positive total effect on Pro-European Immigrant Attitudes partially mediated by the Quality of Intergroup Contact; and a significant positive total effect on Pro-Asian Immigrant Attitudes partially mediated by the Quality of Intergroup Contact. Lastly, Ethnic Food Consumption Frequency had a significant negative total effect on Voter Support for Anti-Immigrant Politicians that was fully mediated by the Quality of Intergroup Contact.4
Indirect Effects of Ethnic Food Consumption Frequency on Sentiments Towards Immigrants Via Quality of Intergroup Contract.
p < .05. **p < .01. ***p < .001.
Furthermore, there is a possibility that having positive encounters with foreigners causes greater ethnic food engagement more so than the converse. We, therefore, tested for this alternative causal pathway by repeating the procedure for each of the same five outcome variables as above, but with Quality of Intergroup Contact entered as the predictor (X) and Ethnic Food Enjoyment as the mediator (M). Results showed that Ethnic Food Enjoyment partially mediated (β = .01, Bootstrap SE = 0.00; [CI] = [0.0004, 0.0219]) the total effect of Quality Intergroup Contact (β = .44, SE = 0.03, p < .001) on Pro- Afro-Caribbean Immigrant Attitudes. It also partially mediated (β = .01, Bootstrap SE = 0.00; [CI] = [0.0031, 0.0266]) the total effect of Quality Intergroup Contact (β = .45, SE = 0.03, p < .001) on Pro-Asian Immigrant Attitudes. However, these mediation effects were much weaker than those from the original models shown in Table 4. Moreover, none of the other three reverse mediation models yielded significant indirect effects, and in all five of these models the direct effect of X on M was also weaker than in said originals.
Next, we tested for reverse mediation again using Quality of Intergroup Contact as the predictor and Ethnic Food Consumption Frequency as the mediator. Ethnic Food Consumption Frequency partially mediated (β = .01, Bootstrap SE = 0.00; [CI] = [0.0012, 0.0249]) the total effect of Quality Intergroup Contact (β = .45, SE = 0.03, p < .001) on Pro-Asian Immigrant Attitudes. However, it did not produce significant mediation effects in the models with any of the other four outcome variables. Additionally, in all five of these models, the effect of X on M was weaker than it was in the previous corresponding models where the predictor and mediation variables are reversed. The results from all 10 reverse mediation tests suggest that both ethnic food enjoyment and consumption frequency more strongly promote positive intergroup contact than the other way around, and this then leads to more favourable attitudes and policy positions towards immigrants. Thus, the overall findings lend strong support for H2.
Discussion
This study aimed to advance intergroup contact theory by identifying a novel, low-barrier mechanism—ethnic food engagement—that facilitates more positive intergroup attitudes and behaviours. Drawing on a nationally representative UK sample, our findings demonstrate that both the enjoyment and frequent consumption of ethnic cuisines are significantly associated with lower anti-immigrant sentiment and reduced support for restrictive immigration policies.
Whilst the standardized beta coefficients reported in Tables 2 and 3 are modest in magnitude, they reflect practically meaningful shifts in public attitudes. For example, the effect of ethnic food enjoyment on perceived immigrant threat (β = −.10, p = .001) implies that a one standard deviation increase in enjoyment is associated with a one-tenth standard deviation reduction in perceived cultural and economic threat—an effect comparable to that of having a personal friendship with an immigrant (Davies et al., 2011; McLaren, 2003). Similarly, the impact of ethnic food consumption frequency on support for anti-immigrant politicians (β = −.07, p = .007) is notable, especially given the typically entrenched influence of partisan and ideological dispositions. These findings suggest that even modest increases in ethnic food engagement may serve as a scalable and low-effort intervention capable of incrementally shifting public opinion on immigration at the population level. 5
Importantly, the mediation analyses indicate that these effects are largely driven by increased positive intergroup contact, reinforcing the core proposition of intergroup contact theory. While the cross-sectional nature of the data prevents firm causal inference, reverse mediation was tested and found to be less plausible, with mediation effects consistently stronger when food engagement was modelled as the predictor rather than the mediator. It is also possible that individuals predisposed to openness or cultural tolerance are more likely to engage with ethnic cuisines. However, the associations remained robust after controlling for three distinct measures of right-wing political preferences, suggesting that food engagement exerts an effect even among individuals likely to be lower in openness and higher in uncertainty avoidance (Kubin & Brandt, 2020; Leyva, 2023).
Finally, the results held across a range of demographic variables—including age, gender, socioeconomic status, region, and exposure to local diversity—indicating that the underlying mechanisms of ethnic food engagement (e.g., sensory pleasure, accessibility, and curiosity) may be broadly effective across the population. These findings mitigate concerns about sampling or ideological bias and underscore the potential of ethnic food related experiences to function as inclusive, low-stakes entry points for engendering more positive intergroup relations in everyday life. Nevertheless, since this is the first study to examine ethnic food engagement as a prejudice-reducing mechanism, further empirical corroboration is needed.
Conclusion
Intergroup contact theory—one of the most empirically substantiated theories in social psychology—proposes that congenial interactions under favourable conditions between different social groups reduce intergroup hostility (Green et al., 2020; Kokkonen et al., 2016). However, its original formulation primarily explains when contact reduces prejudice, not how or why those changes occur (Pettigrew, 1998). As such, further research is needed to identify low-barrier, real-world mechanisms that encourage voluntary contact and unpack the psychological processes through which such interactions foster more tolerant attitudes.
This study contributes to that goal by introducing and testing a novel mechanism: ethnic food engagement. We argue that food consumption offers a uniquely accessible avenue for everyday intergroup contact. Direct outgroup interactions can often provoke anxiety, feel threatening, or be physically impractical—particularly in socially or spatially segregated contexts (White et al., 2021). In contrast, ethnic food settings—such as restaurants, takeaways, or market stalls—typically provide safe, welcoming, and informal environments for casual intergroup encounters. These venues are often more appealing and cognitively accessible than other cultural mediums like music or art, which may require specific historical, linguistic, or aesthetic familiarity. Food, by contrast, meets a daily biological need and engages the senses in ways that transcend language or ideology.
Additionally, when ethnic foods are visible, affordable, and perceived as authentic, their consumption can elicit pleasurable sensory experiences; leading to recurring intake and fostering interactions with the foreign people they represent. Such repeated exposure and positive affect may, therefore, generate initial and lasting favourable associations of those outgroups among ingroup members. Based on this reasoning, we operationalized ethnic food engagement as the enjoyment and frequency of consuming six culturally distinct cuisines—Indian, Chinese, Turkish, Thai, Caribbean, and Spanish—that are widely available, affordable, and typically served by members of non-native outgroups in the UK context.
Our first hypothesis—that ethnic food engagement would be associated with lower anti-immigrant sentiment and voter support for anti-immigrant politicians—was confirmed across all five outcome measures. Regression analyses showed significant and meaningful associations. Our second hypothesis—was that these relationships would be mediated by positive intergroup contact experiences. Results from a battery of mediation tests supported this hypothesis, and provided indirect but consistent evidence against the reverse-causality explanation that positive encounters with foreigners cause greater ethnic food engagement It is also worth noting that during the data collection period for this study (October 2022), UK food price inflation was 16.2%-nearly a 15-fold increase from the same time in 2021 (1.2%) and one of the highest rates in decades. This likely suppressed some participants’ frequency of dining out. Yet, despite this unexpected and pertinent macroeconomic factor, the predictive effects were robust across models.
Moreover, despite growing scientific understanding of intergroup contact, discrimination, populist nationalism, and support for exclusionary immigration policies are intensifying across many liberal democracies—as exemplified most recently by the 2024 re-election of Donald Trump in the US. This goes to O’Donnell et al. (2021, p. 172) point that there remain “significant shortcomings in our understanding and application of contact as a tool for social change.” Our findings help address these gaps by offering evidence that ethnic food engagement may serve as a practical, scalable gateway to more direct and sustained intergroup contact. Its underlying mechanisms—curiosity, sensory enjoyment, and social accessibility—make it an especially inclusive form of engagement, capable of reaching across ideological, generational, and geographic divides. If corroborated, these findings could inform integration policies and strategies such as incorporating ethnic food tasting in multicultural education curricula, offering local government grants or tax incentives for migrant-run food businesses, or promoting ethnic food festivals as community cohesion initiatives.
Finally, although this study focussed on anti-immigrant attitudes among the White British majority, its findings may have broader implications for social integration processes across ethnically diverse populations more generally. Ethnic food engagement could act as a culturally accessible medium through which both majority and minority group members build familiarity, mutual appreciation, and shared social spaces. For example, food festivals, markets, and restaurants often function as communal hubs that attract people from varied backgrounds, including minority-to-minority interactions. In this way, ethnic food may not only reduce prejudice but also cultivate cross-cultural cohesion and solidarity in multiethnic societies. Future research might explore these dynamics in more granular ways, including among ethnic minority groups, migrants of different generations, or in neighbourhoods marked by superdiversity.
Limitations
This study has strengths in its survey design, large representative sample, and extensive covariate controls. However, although the latter controls provide valuable discernments, future research could explore how these mechanisms vary across intersectional identities, such as the interplay between socioeconomic status and regional context, to deepen our understanding of their broader applicability. Furthermore, our operationalization of ethnic food engagement was confined to consumption behaviours and gauged using broad ordinal measures for only six types of cuisine, 6 whilst our positive intergroup contact variable was measured with a single question. Also, due to the political items in the questionnaire, this study’s results may be susceptible to social desirability response bias. Lastly, this study's cross-sectional design does not allow us to firmly determine whether regualar ethnic food engagement leads to more substantive support for outgroup rights over time. Therefore, research employing a longitudinal or experimental design, refined and expanded ethnic food and intergroup engagement scales, as well as implicit attitudinal measures would serve to comprehensively verify and build on the findings from this study.
Footnotes
Acknowledgements
Not applicable.
Ethical Considerations
This research was approved by the Humanities and Social Sciences Ethics Committee from the University of Birmingham (ERN_2023-0688).
Consent to Participate
Data were collected through an online questionnaire, and consent to participate was agreed to on the opening information page/consent form which provided the study’s relevant details and contacts.
Consent for Publication
Not applicable.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.