COVID-19 health certification reduces outgroup bias: evidence from a conjoint experiment in Japan

The study explores the impact of health certifications on reducing outgroup bias during the COVID-19 pandemic. Psychological theories suggest that serious threats, like pandemics, can exacerbate negative attitudes towards outgroups. While numerous studies have documented increased outgroup bias during the pandemic, fewer have investigated potential mitigating factors. This research focuses on the effect of health certifications—proof of vaccination or a negative COVID-19 test result—on reducing negative attitudes towards inbound travelers (foreigners, immigrants, and tourists) in Japan. The study's significance lies in its potential to inform policy decisions regarding border controls and international travel during pandemics, particularly concerning the ethical and practical aspects of vaccine passports and alternative measures.

The behavioral immune system (BIS) theory posits a link between infectious disease threats and outgroup bias. Studies have shown a correlation between perceived disease vulnerability, germ aversion, and negative attitudes towards outgroups. The COVID-19 pandemic has provided ample evidence of this bias globally, targeting various outgroups such as racial/ethnic minorities, immigrants, and tourists. However, research exploring factors that could mitigate this bias remains limited. Existing studies suggest that individualism, shared norms, and pathogen avoidance motivations may influence intergroup attitudes. This study aims to address this gap by examining the effect of health certification on reducing outgroup bias during a pandemic.

A pre-registered population-based discrete choice experiment with a randomized conjoint design was conducted in Japan in February 2021. The study employed a conjoint experiment, presenting participants with pairs of hypothetical inbound traveler profiles and asking them to choose which traveler should be admitted to Japan or to rate their acceptability on a seven-point scale. Ten attributes were used to define each traveler profile: Nationality (Japanese, foreigner with permanent residency, foreigner without permanent residency), Quarantine_Certificate (vaccination certificate, negative test certificate, moving plan, self-isolation, nothing), Region (Taiwan, China, South Korea, US, Brazil), Duration (1 month, 3 months, 1 year, 4 years, indefinite), Purpose (Tokyo Olympics, migration, study, business, tourism), Sex, Age, Speaking Japanese, Education, and Income. The survey was conducted online with a quota sampling method to ensure national representativeness. A total of 1066 participants completed the survey. Three outcome variables were measured: choice-based question (3 options including "Neither should be admitted"), rating-based question (7-point scale), and a secondary choice-based question (2 options) for those who chose "Neither" in the first question. Subgroup analyses explored the impact of host residents' risk perceptions of COVID-19, associations with foreigners, and political partisanship on the effectiveness of health certification in reducing outgroup bias. Linear regression models were used to analyze the data, estimating marginal means (MM) and uniform average marginal component effects (uAMCE).

The results revealed that travelers with health certificates (vaccination or negative test certificates) had a significantly higher probability and rating of being admitted compared to travelers without any quarantine measures. This effect was comparable to the effect of self-isolation. Specifically, the probability of admission was significantly higher for those with health certificates. The effect of health certification was particularly pronounced in reducing the disparity in admission rates between Japanese nationals and foreign nationals with permanent residency. Subgroup analyses showed that the effectiveness of negative test certificates, but not vaccination certificates, in reducing outgroup bias was moderated by respondents' risk perceptions of COVID-19 (serious risk and discomfort) and conservative political partisanship. Those with higher risk perceptions or stronger conservative leanings showed a greater increase in acceptance of travelers with negative test certificates.

The findings support the hypothesis that health certification can mitigate outgroup bias during the pandemic. The study's results corroborate previous research on the BIS theory, demonstrating the potential of health measures to reduce negative attitudes towards outgroups. The consistent positive effect of both vaccination and negative test certificates on reducing bias highlights the potential value of these certifications as policy tools. The significant reduction in the disparity between acceptance rates of Japanese nationals and foreign nationals with permanent residency strengthens this conclusion. While both types of certificates proved effective, the differential moderating effects of risk perception and political orientation highlight the importance of considering contextual factors when implementing such policies. Negative certificates may be less effective in certain social or political contexts.

This study demonstrates that health certification, including vaccination and negative test certificates, effectively reduces outgroup bias against inbound travelers during a pandemic. This provides crucial evidence to support the combined use of both approaches in managing international borders and mitigating pandemic-related anxieties. Future research could explore the role of trust in government institutions and the impact of specific border control policies on this phenomenon. Longitudinal studies are also needed to confirm the stability of these findings over time and across various contexts. Further research should also address methodological refinements, such as using different profile distributions in the conjoint design.

The study's findings are limited by its focus on a specific time and context (Japan during the second State of Emergency and in anticipation of the Olympics). The mechanisms underlying the heterogeneity of effects were not fully explored, and factors such as trust in institutions were not included. The use of a uniform distribution in the conjoint design might also affect the generalizability of the results. Therefore, further research using different methodologies and broader contexts is warranted to confirm and extend these findings.