Does contact improve intergroup attitudes? Evidence from Japanese perceptions of Chinese and South Koreans

Abstract

How can in-group members enhance their perceptions of out-groups? Most scholars emphasize intergroup contact as a key factor. Specifically, they often advocate for face-to-face or direct contact. However, the Japanese context presents a contradiction to this suggestion. For Japanese people, the most frequently encountered foreign nationalities are Chinese and South Koreans, yet they exhibit the most negative attitudes towards these groups. I propose that the monocultural, peaceful, and conservative nature of Japanese society may hinder the realization of the assumptions underlying direct contact. Drawing on advanced studies in intergroup contact theory, I adopt a comparative approach to determine which form of contact is more effective. My findings indicate that viewing Chinese films, a form of parasocial contact, can more positively influence Japanese perceptions of Chinese people than direct communication. This conclusion also applies to the Japanese perspective on South Koreans. Consequently, this study offers a broader social strategy for fostering improved bilateral relations in East Asia.

1. Introduction

In the field of International Relations, practitioners increasingly employ survey research to establish micro-level foundations for their contributions. This focus on microfoundational studies has gained significant importance in the discipline, enhancing our understanding of the intricate relationships between countries, including how they perceive each other (Gerring 2008; Kertzer 2017). In analysing Japan’s foreign relations, this holds true. Research should take into account not only political interactions but also the emotional and group dynamics within society. This is especially true when looking at how Japanese people interact with their neighbours, where underlying social factors play a crucial role.

The Intergroup Contact Hypothesis (IC Hypothesis hereafter) posits that contact between members of different groups can, under certain conditions, lead to reduced prejudice and better intergroup relations (Allport 1954). Face-to-face communication is generally seen as more effective than indirect contact through media or cultural products because it allows individuals to build understanding and empathy from personal experiences. Previous studies on improving out-group acceptance in Japan show varied findings. However, Japanese cases testing the IC Hypothesis have produced conflicting conclusions. Although increased personal interactions are expected to enhance attitudes towards people from neighbouring countries, the reality is often disappointing. Numerous studies indicate that xenophobia towards Chinese and South Koreans is markedly more pronounced than towards other nationalities (Higuchi 2015; Tanabe 2019). In addition, analyses that fail to account for causal inferences can significantly undermine the credibility of the hypothesis. For instance, Imamura et al. (2011) find that positive contacts can improve Japanese attitudes towards Americans; however, they do not clearly explain whether this causal relationship holds true in the reverse situation. To date, the validity of the IC Hypothesis in the context of Japan remains controversial.

In this paper, I test the IC Hypothesis within Japanese contexts, and argue that researchers should compare different forms of contact and deeply understand their causal implications. Interactions not only include face-to-face contact but also contact through media, which is often referred to as parasocial contact. It is therefore essential to differentiate the types of interactions, examine their backgrounds and mechanisms, and aim for comprehensive applicability of findings. The proposed framework for testing the IC Hypothesis encompasses diverse temporal and situational contexts, as well as various out-groups. I evaluate both direct and parasocial contacts, leveraging survey data from 2008 to 2018. In addition, to revise classification methods, I mainly employ the Instrumental Variable (IV) method, sometimes nested with the Hierarchical Linear Model (HLM) method. The findings reveal that while direct contact has not significantly improved Japanese attitudes towards Chinese, parasocial contacts hold more promise, a pattern that also applies to attitudes towards South Koreans.

The remainder of the paper proceeds as follows. Section 2 critically examines past studies on the IC Hypothesis and examines the importance of Japanese cases. Section 3 outlines the methods and data sources used to assess Japan’s international perceptions. Section 4 presents the findings, showing limited impact from actual interactions but more positive outcomes from media-based contacts. Finally, Section 5 explores the broader implications of these results for improving mutual attitudes, with a focus on Japanese and broader sociological contexts.

2. Literature review

2.1. Testing the IC hypothesis: direct contact vs. parasocial contact

Recent years have seen a notable increase in foreigners entering Japan, particularly from China and South Korea (Japan National Tourism Organization 2023). Despite ongoing disputes over territorial sovereignty and historical issues, this upward trend in arrivals has persisted. The rise in international movement has led to speculation that increased contact between Japanese people and Chinese or between Japanese people and South Koreans could help mitigate prejudices and misunderstandings, a concept supported by the IC Hypothesis (Fan 2017). This theory, widely recognized in social psychology, suggests that direct, well-managed interactions between different groups can reduce prejudice and improve intergroup relations (Allport 1954). Subsequent studies by researchers like Sherif et al. (1961) and Pettigrew (1998) have validated the connection between real-world contacts and changes in perceptions, making the IC Hypothesis a leading theory for understanding intergroup conflict.

However, individual cases often reveal a less ideal reality than the hypothesis suggests. Direct contact, when negative, can have a more profound effect than positive interactions, reinforcing prejudices and hostilities rather than reducing them (Pettigrew and Tropp 2008). Furthermore, even positive encounters may not lessen hostility if conflicting interests exist (LeVine et al., 1972). In addition, direct contact can be intimidating for those with limited exposure to out-group members, potentially limiting its effectiveness (Brown and Paterson 2016). These factors indicate that fulfilling the essential conditions for successful intergroup contact, including equal status, common goals, and intergroup cooperation, can be challenging in practical scenarios (Pettigrew 1998).

In light of these challenges, research has increasingly focused on the impacts of indirect forms of contact. This involves exposure to out-group members through media, stories, or acquaintances, rather than direct personal interactions. Indirect contact via media or narratives can offer broader access to diverse cultures and perspectives, reduce stereotypes, and alleviate social anxiety, providing a safer environment for understanding out-groups (Crisp and Turner 2009). This approach has been supported by various studies which demonstrated the positive impact of reducing discrimination and enhancing group identification on life satisfaction among ethnic minorities in different cultural contexts. Recently, attention has turned to the Parasocial Contact Hypothesis, which suggests that relationships formed through media, like television and films, can influence perceptions of different groups (Schiappa et al., 2005). Parasocial contact refers to the pseudo-social interaction individuals have with media personas, where meaningful connections and perceptions are formed through sustained exposure to characters in media settings (Horton and Wohl 1956). This phenomenon highlights how virtual interactions with media figures can shape attitudes and understandings of diverse social groups. To date, however, the effectiveness of these interactions above in changing attitudes has been debated (Ortiz and Harwood 2007; Vezzali et al., 2012).¹ This paper aims to investigate whether parasocial contact is more influential than direct interactions with out-group members.

2.2. Significance of the Japanese cases

Reflecting on the preceding analysis, it becomes evident that meeting the criteria for effective direct contact presents significant challenges. This reality underscores the significance of exploring parasocial contact, an aspect that remains less examined in current studies. Moreover, the extensive body of literature on the IC Hypothesis is primarily derived from Western perspectives, as highlighted by Lemmer and Wagner (2015) and Alatas (2006). Such a focus tends to evaluate diverse societies through a Western lens, and most studies emanate from specific countries such as the USA, Israel, or Ireland. This scenario raises questions about the applicability of these findings in different contexts, particularly in East Asia.

In Japan, related research focusing on interactions with foreigners has been conducted. Tanabe (2013a) demonstrates that individuals with experience in interacting with foreigners often exhibit less exclusivism. Nagayoshi (2008), also, highlights that the type of contact with foreigners influences perceptions. While studies such as those by Tanabe (2013b) and Nagayoshi (2008) mainly focus on direct, face-to-face interactions within Japan, they often overlook the potential influence of media-based indirect contact. Such an oversight is particularly significant in Japan’s relatively homogeneous population structure, limited immigration policies, and socially stable environment, underscoring the importance of exploring how parasocial interactions influence intergroup perceptions.

First, in Japan, the majority identify themselves as ethnically Japanese. This contrasts with the United States, where individuals identify with a common nationhood yet hail from various ethnic backgrounds. In Japan, intergroup conflict is influenced more by national interests than by racial differences. In addition, the racial distinctions between Japanese, Chinese, and Koreans are relatively subtle. This nuance in racial dynamics and national interests shapes the Japanese perspective on intergroup conflicts differently from other contexts.

Second, Japan occupies a unique position—removed from direct involvement in war, yet frequently engaged in diplomatic and territorial disputes with its neighbours. Unlike regions like the Middle East, where territorial disputes often lead to armed conflict, Japan’s disagreements with neighbouring countries have not escalated to violent confrontations since the end of World War II. The Japanese attitude towards China and South Korea, characterized by a lack of closeness and trust, does not generally rise to the level of enmity. This context raises the question: Can findings from studies conducted in war-affected regions be seamlessly applied to the East Asian context? In this paper, I seek to explore and expand on the hypothesis within the intricate social dynamics of international relations in East Asia.

The cultural conservatism prevalent in Japan, particularly in attitudes towards foreigners, is the third dimension to consider. Many Japanese exhibit limited interest in engaging with immigrants in their daily lives, let alone accommodating them. According to the World Value Survey, Japan ranks high in reluctance to accept foreign workers and immigrants, a sentiment significantly more pronounced than in other developed countries. When asked ‘Could you please mention any groups you would not like to have as neighbours?’, 36.3 per cent of Japanese respondents in Wave 6 and 29.1 per cent in Wave 7 identified ‘Immigrants/foreign workers’ as undesirable neighbours (WVS 2023). In addition, international travel among Japanese citizens is relatively low, with annual overseas trips accounting for less than 20 per cent of the total population over the past two decades (Immigration Services Agency 2023). This lack of direct contact with foreigners suggests that merely fostering real-person interactions may not be enough to sufficiently improve attitudes. Furthermore, across multiple survey years, Japanese respondents have often exhibited the most unfavourable views towards Chinese and South Koreans compared to other ethnic groups globally (Pew Research Center 2024). Therefore, testing the IC Hypothesis in this context could provide valuable insights into addressing these challenges.

3. Methods

3.1. Data

Data from the Japanese General Social Surveys (JGSS), supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT), are utilized in this analysis (JGSS 2023). The JGSS provides a rich dataset that enables the advancement of a variety of academic studies. Its survey items encompass an extensive array of topics, including individual circumstances, political perceptions, and attitudes towards social issues. Employing a stratified two-stage sampling methodology, the JGSS carefully selects participants aged 20–89 years from all prefectures in Japan, thus ensuring a representative sample and minimizing potential measurement errors. Notably, the JGSS’s International Comparative Survey on Values and Behavioural Patterns, conducted in 2008, 2017, and 2018, offers insights into Japanese attitudes towards Chinese people. This specific component of the survey includes a total adolescent sample size of 3,698, providing a valuable perspective on contemporary intercultural relations. Its detailed descriptions of each variable can be seen in Appendix Table A1 of the Supplementary Material.

3.2. Dependent Variables

The dependent variables are primarily centred on Japanese attitudes towards the Chinese. Because biases towards external groups often manifest in everyday interactions (Brown 1995; Tanabe 2013a), this study focuses on exploring how Japanese individuals perceive Chinese people they may encounter. After examining the JGSS surveys, I identify three relevant questions: ‘How would you feel about working together with people from China?’, ‘How would you feel about having neighbours who are from China?’, and ‘How would you feel about having close relatives who are from China?’. Among the three, the first and third questions effectively represent the realms of work and everyday life. From the perspective of social distance, these two questions capture both distant and close social relations. The second question, however, is more challenging to interpret. Neighbour relations can be greatly influenced by housing conditions, potentially confounding the interpretation of the results. Thus, I decided to select the first and third questions as the dependent variables. Responses indicating a lack of receptiveness are coded as 0, whereas those showing receptivity are assigned a value of 1. This coding strategy allows for a decontextualized understanding of the levels of acceptance or reluctance in these social scenarios.

3.3. Key independent variables

As mentioned previously, my objective is to discern whether parasocial contact is more effective than direct interaction in shaping attitudes. To this end, I identify two primary independent variables. First, to measure direct contact, I employ the question, ‘Do you have acquaintances from China?’ with responses coded as No (0) and Yes (1). This approach is supported by Pettigrew and Tropp (2006), who emphasize the irreplaceable importance of cross-racial ‘acquaintance’ in face-to-face contact. Furthermore, according to Pettigrew and Tropp (2006), the effectiveness of this approach has been empirically verified by Catlin (1977). Second, for parasocial contact, I use the query, ‘How often do you watch Chinese films?’. This choice is predicated on the assumption that Chinese cinema provides Japanese viewers with insights into Chinese culture and history, thereby potentially influencing their perceptions of Chinese individuals (Hall 1997). The original response options are ‘1 = often’, ‘2 = sometimes’, ‘3 = seldom’, and ‘4 = not at all’. However, the proportions for ‘often’ and ‘sometimes’ are very low, at 0.79 per cent and 9.96 per cent, respectively. Retaining the original four categories might lead to statistical instability and errors due to the small sample sizes. Therefore, merging the categories into a binary variable improves the reliability and validity of the statistical analysis (Hosmer, Lemeshow and Sturdivant 2013). Responses indicating an absence of experience are recoded as 0, whereas those denoting experience are assigned a value of 1, allowing this study to quantitatively assess the impact of these interactions.

3.4. Control Variables

The analyses contain an extensive set of additional variables. These include their interest in global issues, assessed with the question ‘How often do you talk about international issues with your families, friends, or other people?’; frequency of social engagement, evaluated by ‘Other than family and relatives, approximately how many people do you contact daily (including phone calls, letters, emails, and direct meetings)?’; as well as factors like age groups, educational level, urbanity of dwelling, income range, and gender. Excluding gender, which is a categorical variable, the rest are continuous, with elevated scores reflecting a higher tier or intensity.

In addition to these individual-level variables, this study also examines macro-level factors. These encompass attributes beyond the control of individual respondents, such as the year of the survey and indices representing Sino-Japanese relations. Because of the absence of a country-level international relations index in the JGSS, this research utilizes the Leaders’ Visit Index from Google’s Global Database of Events, Language, and Tone as a proxy for Sino-Japanese relations (GDELT 2023). Leaders’ visits, often reflecting significant diplomatic interactions and attracting substantial media coverage, provide a valuable tool for analysing the dynamics of the relationship between the two countries. For details on the computation methodology of the index, please refer to the Supplementary Material.

4. Results

The data analysis unfolds in four structured stages. In the first stage, I assess the non-linear relationship between Japanese acceptance and their intergroup contacts with Chinese individuals through binomial logistic regression (Cox 1958; Aldrich and Nelson 1984). Subsequently, to address potential endogenous issues, particularly bilateral causality, I employ the IV method. Third, my analysis extends to attitudes towards South Korean individuals, evaluating if the insights derived from the study on Chinese contacts apply to other neighbouring countries of Japan. Finally, I modify the database and code techniques to ascertain the robustness of the findings.

4.1. Basic regression

The first stage of my analysis examines the bivariate association between acceptance levels of Chinese people and the degree of acquaintance with them while considering social distance as a factor. Results from Model 1 and Model 5 in Table 1 indicate that communication with Chinese individuals, irrespective of domain, enhances the likelihood of their acceptance. Next, the study explores the impact of exposure to Chinese films on acceptance, assessing the role of parasocial contact. Findings from Model 2 and Model 6 suggest that Chinese films positively influence acceptance levels. Crucially, my analysis integrates both direct and parasocial contacts, introducing an interaction term to assess their combined and individual effects on acceptance. My aim is to explore the comprehensive and differential impacts on acceptance levels. Analysis from Model 3 and Model 7 demonstrates that both forms of contact—direct and parasocial—significantly influence acceptance levels. However, subsequent tables will not feature analyses from Model 4 and Model 8 because their effects are not found to be interactive. In conclusion, Table 1 seemingly suggests that both direct and parasocial contacts have a positive influence on Japanese attitudes towards Chinese individuals.

4.2. Endogeneity test

However, similar to the limitations noted in most previous studies, the results from Table 1 cannot rule out the possibility of reverse causation. It is conceivable that Japanese individuals who are inherently more receptive to Chinese people are also more inclined to engage with Chinese cinema or interact with Chinese individuals. To address this issue of endogeneity, I apply the IV method. This involves identifying variables that are correlated with the endogenous explanatory variables but not with the error term, aiming to mitigate the effects of endogeneity (Angrist and Krueger 1991). Consequently, in Table 2, I introduce two instrumental variables into the models. These are designed to be associated with the likelihood of Japanese residents being acquainted with Chinese individuals and their frequency of watching Chinese films. Furthermore, because these instrumental variables are set at the prefecture level, they are less likely to influence directly the individual respondents’ personal preferences regarding their proximity to Chinese individuals, thus fulfilling the criteria for exclusion (Stock and Yogo 2005).

First, in examining the variable of contact with Chinese individuals, this study employs the population density of Chinese residents in each prefecture as an instrumental variable.² The selection criteria are threefold: the instrumental variable increases the chances for Japanese individuals to interact with Chinese people, indirectly influencing their attitudes; with other sociodemographic factors controlled, the proportion of the Chinese population primarily affects attitudes through increased contact opportunities, meeting the exclusion restriction requirement; finally, because there is little evidence that this variable is related to other potential influencing factors on Japanese people’s attitudes towards Chinese people, it fulfils the exogeneity condition (Angrist and Krueger 2001). In addition, according to Raudenbush and Bryk (2002), my models contain both prefecture-level and individual-level variables. To account properly for variability within the prefecture level, the model should simultaneously use the HLM, ensuring that it accurately reflects its estimates. In the first stage of the IV regression for Model 1 and Model 3 from Table 2, the F-values of the instrumental variable regressions are both 23.24, far exceeding the critical value for weak instruments (Staiger and Stock 1997). However, in the second stage, the analysis shows that communication with Chinese individuals does not significantly enhance acceptance among Japanese people (–0.047 in Model 1 and –1.316 in Model 3).

Second, concerning exposure to Chinese films, I utilize the annual box office revenue of Chinese films (in million dollars) as the instrumental variable, drawing on data from the Foreign Films Importer-Distributors Association of Japan and IMDbPro (FFIAJ 2023; IMDbPro 2023). To ensure that this IV variable is not correlated with the equation disturbance, I control for other socioeconomic factors. I also calculate the F-value of the first-stage regression to confirm that it does not belong to a weak instrumental variable (F = 19.48 > 10) (Staiger and Stock 1997; Bollen 2012). After verifying assumptions, the results from Model 2 and Model 4 indicate that Chinese films continue to exert a positive influence on acceptance. In contrast to the results presented in Table 1, after adjusting for endogeneity, the parasocial contact does still appear to reduce in-group members’ prejudices (2.034*** in Model 2 and 1.755*** in Model 4).

4.3. Sensitivity test

As previously emphasized, regarding Japanese perceptions of the Chinese, parasocial contact appears to be more effective than direct contact. In addition, as highlighted by Tanabe (2008), Japanese attitudes often demonstrate less enthusiasm towards proximate neighbours, including China and South Korea. Consequently, it becomes crucial to ascertain whether my findings regarding the two types of contacts are uniquely applicable to the Chinese context or indicative of a more general sentiment towards neighbouring countries. To explore this, my study extends its analysis to include three questions from JGSS that focus on South Korean influences, aiming to uncover broader patterns and subtleties in Japanese perspectives towards foreign nationals.

Mirroring the approach used for Chinese cases, my study selects two pivotal questions as variables for South Korean cases: ‘How would you feel about working together with people from South Korea?’ and ‘How would you feel about having close relatives from South Korea?’. Responses expressing unwillingness are coded as 0, whereas those indicating willingness are coded as 1. The primary independent variables stem from the questions: ‘Do you have acquaintances from South Korea?’ and ‘Have you watched Korean dramas?’. Korean dramas hold a pivotal role in South Korean culture, serving as a powerful medium for both national identity expression and global cultural influence (Kim et al., 2017). Answers indicating no acquaintances or lack of exposure to Korean dramas are coded as 0, whereas affirmative responses are coded as 1. For the former question, the population density of South Korean residents in each prefecture is employed as an instrumental variable (SBJ 2023). For the latter, the number of Korean dramas in the survey year is used as the instrumental variable. Specifically, the dramas considered must have been broadcasted at least once on NHK, KBS World, and the five major private broadcasters (TBS, TV Tokyo, TV Asahi, Nippon TV, and Fuji Television) (NHK 2023; KBS World 2023; TBS 2023;  Nippon TV 2023; TV Asahi 2023; TV Tokyo 2023; Fuji Television 2023; Kdrama 2023). In addition, HLM is incorporated in the regression analysis for the variable representing the population density of South Korean residents in each prefecture. Both instrumental variables have passed the weak instrument test, with F-values of 21.83 for Models 1 and 3 in Table 3, and 20.37 for Models 2 and 4 in the same table.

The findings, as revealed by Models 1 and 3 in Table 3, suggest that having South Korean acquaintances does not substantially increase acceptance among the Japanese population (3.134 in Model 1 and -1.014 in Model 3). Conversely, the outcomes from Model 2 and Model 4 indicate that exposure to Korean dramas positively influences acceptance (1.657*** in Model 2 and 1.237* in Model 4). Thus, based on the data presented in both Table 2 and Table 3, I can conclude that the direct contact hypothesis is not supported in the context of Japan’s interactions with its neighbouring countries, whereas the Parasocial Contact Hypothesis is upheld.

4.4. Robustness checks

After completing the regression analysis, I perform robustness checks to confirm the validity and reliability of my results. Such checks are instrumental in ascertaining the consistency of the findings across varying model specifications and data subsets, thus reinforcing the soundness of the conclusions. As detailed in Table 4, my strategies for robustness checks involve two distinct methods: altering the datasets and modifying the coding techniques.

First, to mitigate the potential bias arising from reliance on a single survey, I replicate the analysis using an alternative dataset, the Public Opinion Poll on Internationalization and Citizens’ Political Participation (ICPP 2022). Conducted in 2009 and 2013, the ICPP gathered data from 7,744 participants across Japan. Utilizing this dataset, I define the dependent variable as, ‘Would you welcome an increase in Chinese/South Koreans in your vicinity? (Would not=0, Would=1)’, and identify the primary independent variable as ‘Have you befriended a Chinese/South Korean? (No=0, Yes=1)’. These variables’ descriptive statistics are shown in Appendix Table A1 of the Supplementary Material. Analysis of Model 1 and Model 2 in Table 4, which focuses on attitudes towards the influx of immigrants, reveals that befriending South Koreans does not improve Japanese attitudes; surprisingly, even befriending Chinese individuals can significantly deteriorate them. Consequently, the regression results from the ICPP dataset also refute the direct contact hypothesis.

Second, I evaluate the broader sentiments of Japanese individuals towards their East Asian neighbours. Continuing to utilize JGSS datasets, I modify the dependent variable to measure the favourability of Japanese views towards both individuals from the People’s Republic of China (PRC) and the Republic of Korea (ROK) (acceptable to none = 0, acceptable to either Chinese or South Koreans = 1, acceptable to both = 2). I also adjust the key independent variable to examine whether Japanese people have acquaintances from either or both countries (none = 0, either = 1, both = 2). In addition, I adapt the instrumental variable to reflect the proportion of Chinese and South Korean residents in each prefecture. The regression findings from Model 3 and Model 4 consistently demonstrate an absence of any notable positive change in attitudes, even when incorporating more nuanced forms of contact. Consequently, my scepticism regarding the direct contact hypothesis is further substantiated in the context of Japanese perceptions towards its East Asian neighbours.

5. Discussion and conclusion

Nowadays, forms of interaction are becoming increasingly diverse. Developments in transportation have made face-to-face communication more accessible, while the proliferation of media enables greater exposure to diverse cultural products. Both forms of interaction can potentially invite us to lower or increase prejudice or acceptance toward other communities or nations. In Japan, intergroup contacts facilitated by media seem more effective than face-to-face interactions at limiting citizens’ prejudice against Chinese and South Korean. This finding challenges the prevailing IC Hypothesis, which typically holds that face-to-face contact is superior in diminishing group prejudice.

This counter-intuitive conclusion might be attributed to two distinct factors. First, it is partially explicable through the lens of Cognitive Dissonance Theory. As Festinger (1957) articulated, this theory suggests individuals tend to engage with information that aligns with their existing beliefs, thus avoiding cognitive dissonance. In real-world interactions, Japanese individuals may avoid positive engagements with Chinese or South Korean people due to entrenched prejudices or seek evidence that confirms their biases. Moreover, Orsini and Magnier-Watanabe (2022) empirically show that the effect of direct contact on perceptions is not linear. Compared with no contact, minimal contact does not effectively challenge existing prejudices. Conversely, exposure to entertainment forms like films or television dramas is likely to place viewers in a more relaxed state, making them more open to new perspectives (Zillmann 1991). Subsequently, the concept of ‘Narrative Resonance’ (Green and Brock 2000) can potentially explain why parasocial interactions are found associated with lower prejudice on Chinese and South Korean people. In real-life encounters with foreigners, individuals might draw associations with complex international dynamics, potentially reinforcing existing prejudices (Pettigrew and Tropp 2006). However, cultural productions like films and TV series tend to present idealized stories and characters, softening harsh realities and fostering emotional resonance and empathy (Green et al., 2004). These idealized portrayals can help develop positive emotional connections among Japanese audiences towards Chinese and South Korean people and cultures.

In examining the reasoning process, I uncover two primary limitations. First, my analysis recognizes the influence of cultural exposure but falls short in distinguishing the nuances between different types of cultural products. More precisely, Chinese films and Korean dramas show unique qualities and broadcast modes. As previous studies have mentioned, to promote overall positive attitudes, a likable cultural persona needs to be perceived as typical of the out-group (Bodenhausen et al., 1995). Therefore, it is important to determine whether the positive and likable characters in Chinese films and in Korean dramas can respectively represent Chinese and South Koreans as a whole, thereby fostering the potential for positive feelings to extend to these groups (Tukachinsky et al., 2015). Future research should not only focus on differentiating the forms of these cultural media but also concentrate on whether Japanese audiences agree with the political and cultural values embedded in these forms. By comparing cultural media from different countries, it seems possible to better evaluate the effectiveness of the country’s soft power strategy (Chitty 2021).

Second, although the surveys used in this study have laid a solid foundation for discovering Japanese perceptions of foreign countries, some flaws still exist. The surveys might not fully address the problem of consistency bias, which may artificially inflate correlations between variables (Podsakoff et al., 2003). For example, in JGSS 2008, questions about perceptions towards China and South Korea are mainly concentrated in Question Numbers 51–55, potentially motivating respondents to unify their attitudes (JGSS 2008). In addition, the survey lacks questions that allow for in-depth exploration of psychological mechanisms such as enhanced empathy, increased openness, and reduced fear, which impacts the analysis of contact and perception. This limitation prevents this study from conducting factor analysis to assess the significance of these factors. Finally, the survey design restricts the understanding of Japanese attitudes towards other neighbours like North Korea and Russia, potentially affecting the applicability of my conclusions.

Despite these potential shortcomings, this research contributes significantly to two areas. First, by comparing different types of contact, we can find that direct interaction is not always superior to parasocial contact in improving intergroup relations. This challenges the prevailing assumption that direct contact is invariably more effective, and suggests the effects of face-to-face interaction and parasocial interaction on group prejudice might be context specific. Second, this study of xenophobia in Japanese society provides a sociological basis at the micro-level for bettering Sino-Japanese and Japan-South Korea relations. This not only contests the common diplomatic view that stronger people-to-people ties directly lead to better bilateral relations but also ties in with international relations theory, such as soft power (Nye 2004), through our exploration of cross-cultural communication.

Conflict of Interest

None reported.

Funding

None reported.

Acknowledgements

I am grateful to Ken’ichi Ikeda and Xiaoguang Fan for their immense support. I also extend my sincere thanks to Shu Keng, Angran Li, Shengyu He, Youxing Lang, Duofen Chang and the two anonymous reviewers for their insightful comments and constructive suggestions. I thank the organizers of the two databases for granting access to their data. The Japanese General Social Survey (JGSS) is managed by the JGSS Research Center at Osaka University of Commerce with support from the university. JGSS-2008 was funded by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) through the Gakujutsu Frontier Grant, in collaboration with the University of Tokyo’s Institute of Social Science. The JGSS-2017G/2018G, supported by MEXT’s Distinctive Joint Research Center Program, was conducted with Kyoto University’s Division of Sociology of Education. Data curation was supported by the Japan Society for the Promotion of Science (JSPS) under the Humanities and Social Sciences Data Infrastructure Program, Grant JPJS00218077184. Data for the secondary analysis titled ‘Public Opinion Poll on Internationalization and Citizens’ Political Participation’, 2009, 2013, conducted by Shunsuke Tanabe and Mitsuru Matsutani, was sourced from the Social Science Japan Data Archive at the Center for Social Research and Data Archives, Institute of Social Science, The University of Tokyo.

References

Footnotes