AUTHENTICITY IN ETHNIC RESTAURANTS: INVESTIGATING THE ROLES OF ETHNOCENTRISM AND XENOCENTRISM

INTRODUCTION

It is clear from industry reports that U.S. consumers" inclination toward ethnic restaurants has been growing over the past few decades. A recent survey showed that 67% of U.S. consumers eat ethnic cuisine at least once a month (Mintel 2019). Surveys of professional chefs (e.g., National Restaurant Association 2019) also show corresponding results, reporting ethnic food items as a leading restaurant industry trend in the U.S.

Given the above, it is not surprising that many studies have attempted to identify factors that impact the consumers’ intention to eat at ethnic restaurants. In general, these studies investigated either 1) the influential characteristics of ethnic restaurants or 2) the personality traits that predict consumers’ intention to eat at ethnic restaurants. The former stream of research commonly proposed perceived authenticity of ethnic restaurants as one of the most significant predictors of positive behavioural intention (e.g., Jang et al. 2011;Kim et al. 2020). As for the latter stream of research, it was commonly found that food neophobia (lack of familiarity) negatively impacts behavioural intention (e.g., Camarena et al. 2011;Choe and Cho 2011); and that cosmopolitanism positively impacts behavioural intention (e.g., Zhang and Hanks 2018). While it is obvious that these findings explain important aspects of ethnic restaurant consumer behaviour, a significant research gap is that the interaction effect is relatively unknown (for exception, see Wang and Mattila 2015). Is perceived authenticity crucial to the success of an ethnic restaurant regardless of the consumers’ personality traits? Do consumers with certain personality traits prefer inauthentic ethnic restaurants over authentic ethnic restaurants and vice versa? This study attempted to address the lack of clarity by investigating how the perceived authenticity of an ethnic restaurant interacts with personality traits to predict behavioural intention.

In doing so, we had to determine first what personality traits to test as potential moderators. One may suggest testing the moderating effects of food neophobia and cosmopolitanism – traits that have already been identified to impact ethnic restaurant intention. However, such an approach was considered limited due to the following reasons. First, regarding food neophobia,Pliner and Salvy (2006) mentioned that the trait refers to the tendency to avoid unfamiliar foods. Accordingly, the trait has been utilized not only in ethnic restaurant contexts but also to study the acceptance of a variety of novel foods such as edible insects or cultured meat (Sogari et al. 2019;Siegrist and Hartmann 2020). This suggests that the negative impact of food neophobia found in ethnic restaurant studies is not necessarily related to the restaurant’s ethnic-ness but rather due to the unfamiliarity toward dishes served in the restaurant. Therefore, in order to hypothesize that food neophobia moderates the impact of perceived authenticity, one must assume that perceived authenticity corresponds to unfamiliar foods. Such an assumption, however, would be inaccurate since authentic ethnic restaurants are likely to offer a mix of familiar and unfamiliar foods (rather than only unfamiliar foods). Second, according toCleveland et al. (2009), cosmopolitanism refers to the willingness to engage with ethnic cultures other than their own. An interesting aspect of cosmopolitanism is that the trait is not characterized by loyalty toward ethnic products, but rather is indicative of openness towardany products regardless of whether it is ethnic or non-ethnic (Riefler et al. 2012). InWang and Mattila (2015), despite empirical evidence suggesting the positive impact of cosmopolitanism on intention toward ethnic restaurants, the researchers were not able to demonstrate an interaction effect between perceived authenticity and cosmopolitanism. We assume that the insignificance was due to cosmopolitan consumers being open toward any ethnic restaurant regardless of its perceived authenticity. Therefore, while it is clear that cosmopolitanism is a significant predictor of positive intention toward ethnic restaurants, the usefulness of the trait as a moderator is debatable since a positive impact is likely to be found regardless of whether the restaurant is ethnic or non-ethnic.

In all, the above indicates that food neophobia and cosmopolitanism are limited as potential moderators since behaviours stemming from those traits are not necessarily relevant to the perceived authenticity of ethnic restaurants. Ethnocentrism and Xenocentrism, on the other hand, are personality traits that may be indicative of preference or aversion toward ethnic commodities that are authentic or inauthentic. Accordingly, we argue that investigating the moderating effects of these two traits is more likely to provide results that are useful in an ethnic restaurant context (see figure 1). Next section provides the rationale for our argument, followed by relevant hypotheses.

1. ETHNOCENTRISM, XENOCENTRISM, AND PERCEIVED AUTHENTICITY

The concept of ethnocentrism, which was originally developed and utilized by researchers within the fields of sociology and psychology, first appeared in the consumer behaviour literature whenShimp (1984) coined the term "consumer ethnocentrism". According toShimp (1984), ethnocentric consumers have the tendency to perceive domestic products as superior compared to imported alternatives and believe that purchasing imported products is immoral and unpatriotic. Since the initial introduction, the consumer ethnocentrism construct has been tested via numerous studies leading to important findings within various areas of research. Within the context of restaurants, it has been found that ethnocentric consumers tend to avoid ethnic restaurants (Camarena et al. 2011).

Given the above, our question is whether the perceived authenticity of an ethnic restaurant plays any role in determining the intention of ethnocentric consumers. Various elements can make an ethnic restaurant seem authentic – for instance, the use of ethnic décor, hiring ethnic chefs, and offering dishes that are cooked using authentic ingredients (e.g., Youn and Kim 2017). According to previous research, such elements that signify the authenticity of an ethnic restaurant positively impacts behavioural intention (e.g., Jang et al. 2012). However, for ethnocentric consumers, we expect that their intention would be negatively impacted since those elements are in contradiction to ethnocentric consumers" tendency to favour domestic (culturally similar) products. On the other hand, we expect ethnocentric consumers to respond more favourably to ethnic restaurants that are perceived as inauthentic since elements that signify inauthenticity (e.g., use of common décor, hiring local chefs, cooking with locally available ingredients) are likely to better align with ethnocentric consumers’ beliefs. In all, our expectations discussed thus far can be expressed more formally as follows:

H1: The positive effect of perceived authenticity on behavioural intention toward ethnic restaurant is attenuated by consumers’ level of ethnocentrism.

"Consumer xenocentrism" refers to consumers’ tendency to favour foreign products due to negative stereotypical perception of domestic products (Balabanis and Diamantopoulos 2016). Initially, the concept was considered to be relevant only in emergent markets, however, according toHyun and Fairhurst (2018), the phenomenon is becoming common also in developed countries such as the U.S. In line with the above,Balabanis and Diamantopoulos (2016) recently developed the C-XENSCALE (consumer xenocentrism scale) based on a series of statements designed to capture consumer’s belief of inferiority of domestic products and a corresponding tendency to prefer foreign products.

As mentioned earlier, the positive impact of perceived authenticity on the general consumer’s behavioural intention is well-documented. For xenocentric consumers, we expect perceived authenticity to play an even stronger role in determining positive behavioural intentions, given their significant inclination toward foreign products (Camacho et al. 2020). In support,Diamantopoulos et al. (2019) showed that xenocentric consumers display positive behavioural intention toward genuine foreign brands, but not toward foreign counterfeit brands. In all, our expectations can be expressed more formally as follows:

H2: The positive effect of perceived authenticity on behavioural intention toward ethnic restaurant is amplified by consumers’ level of xenocentrism.

2. METHOD

A self-administered online questionnaire was developed then distributed via Amazon Mechanical Turk. The questionnaire was structured as follows. First, participants were randomly presented with questions that measure either their ethnocentrism or xenocentrism (ethno group n=292, xeno group n=289). Ethnocentrism was measured using the CETSCALE (Shimp and Sharma 1987) and xenocentrism was measured using the C-XENSCALE (Balabanis and Diamantopoulos 2016). Second, participants were randomly assigned to either a high or a low authentic Japanese restaurant scenario. Authenticity was manipulated by utilizing descriptions that signify low or high level of authenticity in terms of the restaurant"s atmospherics (Jang et al. 2011) and servicescape components (Wang and Mattila 2015). After reviewing the scenario, manipulation check was performed using four items fromJang et al. (2011). Lastly, behavioural intention was measured using five items fromZeithaml et al. (1996). All items were measured on a seven-point Likert type scale (1= strongly disagree, 7= strongly agree). Reliability of the scales was confirmed via Cronbach’s alpha values that ranged from 0.944 to 0.970.

3. RESULTS AND DISCUSSION

Manipulation check results supported the effectiveness of our scenarios. Participants assigned to the low authenticity scenario perceived the restaurant as less authentic compared to those that were assigned to the high authenticity scenario. Specifically, a series of t-tests showed significant differences in perceived authenticity between low and high conditions for both the ethno group and the xeno group (ethno group t=-14.46, p<0.001, Mlow=3.00<Mhigh=5.56; xeno group t=-15.16, p<0.001, Mlow=2.91<Mhigh=5.57).

After controlling for potential confounders (age, gender, education, income, marital status, ethnicity, familiarity toward the Japanese cuisine), both hypotheses were tested via a series of regression analyses. For H1, perceived authenticity was shown to significantly interact with ethnocentrism to predict behavioural intentions (β=-0.44, t=-4.73, p=0.001). Specifically, Johnson-Neyman test showed a significance region of 5.84, implying that there was a significant difference in behavioral intentions between participants with high level of ethnocentrism (above 5.84) and low level of ethnocentrism. As shown in figure 2, within the low ethno group, participants assigned to the high authenticity scenario (M=5.31) displayed significantly higher behavioural intentions compared to those assigned to the low authenticity scenario (M=2.96). However, within the high ethno group, the difference in behavioural intentions was much less evident between high (M=5.00) and low (M=4.45) authenticity scenarios. This suggests that the positive impact of perceived authenticity was attenuated by the level of ethnocentrism, thus confirming H1.

For H2, perceived authenticity was also shown to significantly interact with xenocentrism to predict behavioural intentions (β=-0.40, t=-4.26, p<0.001). However, contrary to the expectation, the impact of perceived authenticity was attenuated (rather than amplified) by xenocentrism (Johnson-Neyman significance region=5.55), thus rejecting H2. As shown in figure 3, within the low xeno group, participants assigned to the high authenticity scenario (M=4.97) displayed significantly higher behavioural intentions compared to those assigned to the low authenticity scenario (M=2.62). As for the high xeno group, the difference in behavioural intentions was not as evident between high (M=5.33) and low (M=4.55) authenticity scenarios.

Aside from our testing of the two hypotheses, it is worth noting that low ethno group (M= 5.31) and high xeno group (M= 5.33) displayed higher behavioural intentions compared to their counterparts within the high authenticity condition. This suggests that higher level of authenticity in ethnic restaurants may have varying impacts on intention depending on whether the consumer is ethnocentric or xenocentric. In other words, although both ethnocentrism and xenocentrism were shown to attenuate the impact of perceived authenticity, one should not prematurely assume that the two personality traits lead to identical behaviours.

4. CONCLUSIONS AND FUTURE RESEARCH

The general purpose of this study was to investigate whether the perceived authenticity of an ethnic restaurant interacts with personality traits to predict behavioural intention. While there have been similar attempts in the past (Wang and Mattila 2015;Liu et al. 2018), those studies tested the moderating effects of personality traits (food neophobia, cosmopolitanism) that are not necessarily relevant to the authenticity aspect of ethnic restaurants. This study, on the other hand, tested ethnocentrism and xenocentrism – personality traits that may lead to varying behaviours depending on whether the evaluated object is perceived as authentic or inauthentic. Specific results indicate that perceived authenticity may not be a significant predictor of ethnic restaurant success if the consumers are ethnocentric (H1). Practically, given the fact that ethnocentric consumers tend to be members of low-status groups with lower level of income and education (Shimp and Sharma 1987), this finding would be especially useful for ethnic restaurants making store location or target market decisions (e.g., inauthentic ethnic restaurants may be more likely to succeed in geographical areas with high concentration of lower-status individuals). Contrary to our H2, xenocentrism was also shown to attenuate the impact of perceived authenticity. One potential explanation can be found inDiamantopoulos et al. (2019) which implies that xenocentric consumers seek certain ethnic products only if the associated ethnic culture is perceived to be superior to one"s own. That is, the xenocentric participants in our study may have perceived the Japanese culture as not superior, further leading to lower than expected behavioural intention in the high authenticity scenario. Another possible explanation could be that the restaurant described in the low authenticity scenario was not inauthentic enough to deter the xenocentric participants. That is, xenocentric participants may have perceived the low authenticity ethnic restaurant to be more foreign rather than domestic. In all, although H2 was rejected, the limited influence of xenocentrism found in our study invites future researchers to determine the type of ethnic cultures and the specific level of authentic-ness that causes xenocentric consumers to approach or avoid ethnic restaurants. Lastly, a limitation for this study is that our hypotheses were tested only within the context of a Japanese restaurant. Replication studies should follow to determine the generalizability of our results.