By NOW IN QUESTION EditorialSourcing & correctionsReport an error
Survey Finds 40% of Students Skeptical of More Foreign Residents—Can the Internet Be Blamed?
A study of about 2,000 students at ten Tokyo-area public schools tests the limits of headlines, causation and coexistence
In a 2025 survey of about 2,000 students at ten public junior high schools in the Tokyo metropolitan area, more than 40% leaned toward saying that an increase in foreign residents would not bring many good things to society. The share was in the 20% range in 2022. But does that change prove a rightward shift—or that the internet caused it?
Japanese news organizations reported on November 7, 2026 that a 2025 social-attitudes survey at ten public junior high schools in the Tokyo metropolitan area found more than 40% of respondents leaning toward the view that an increase in foreign residents would not bring many good things to society. In the comparable 2022 survey, the share had been in the 20% range.
Third-party public records of Japanese-language X trends displayed related phrases including “40% of junior high students,” “increase in foreigners,” “internet influence” and “the internet's fault” during the same news cycle. These records establish a visible conversation, not a complete national X dataset, an official post total or a measure of what all Japanese speakers believe.
The headline invites two rapid conclusions: young people have become exclusionary, and social media caused the shift. Yet the academic presentation was titled “Liberalization and Conservatization in Junior High School Students' Social Attitudes: The Turn of 2025.” Reporting says support for human rights and diversity remained above 90% on several items, and the researcher warned against reducing the results to simple conservatization.
The useful question is therefore not which label to attach to adolescents. It is: what can one changed survey response tell us, and what remains unproved?
What the available evidence establishes
| Item | What reporting and conference materials establish |
|---|---|
| Sample | About 2,000 students at ten metropolitan-area public junior high schools |
| Survey year | 2025, compared with 2018–19 and 2022 waves |
| Foreign-resident item | Whether an increase would bring “many good things” to society |
| 2025 result | More than 40% on the “do not think so” side |
| 2022 result | In the 20% range on that side |
| Other attitudes | High support on multiple human-rights and diversity items |
| Presentation | Japan Sociological Society conference report by University of Tokyo researcher Norihiro Nihei |
The respondents were not asked simply whether “foreigners are bad.” The item concerned whether an increase in foreign residents would produce many social benefits. Different reasons can arrive at the same response: expectations about the scale of benefits, anxiety about local capacity, language barriers, perceptions of security, or experience in a particular community.
About 2,000 respondents is a meaningful sample, but ten metropolitan-area schools are not automatically every junior high student in Japan. The sampling design, response rate, precise questionnaire and regional composition determine the population to which the finding may reasonably be generalized.
“Internet influence” is a hypothesis, not a causal finding
Reports quoted the researcher raising the possibility that online information—amid debate over foreign-resident policy in the 2025 upper-house election—affected the 2025 results. A plausible mechanism is not the same as proof.
If internet use and skeptical responses rose in the same period, their movement together does not by itself show that one caused the other. A stronger causal claim would need evidence about:
- what each respondent saw, on which service and how often;
- whether attitudes changed before or after exposure;
- conversations at home, school and in peer groups;
- direct contact with foreign residents;
- alternative influences such as prices, employment, crime coverage and election campaigning.
“The internet” is not one message. Exclusionary posts, first-person accounts by migrants, municipal guidance and fact-checking all travel through the same networks. Treating the container as the cause hides the harder questions: which claims, recommendation systems and relationships made information persuasive?
Liberal norms and practical anxiety can coexist
It can look inconsistent to support rights and diversity while doubting the benefits of a larger foreign-resident population. Human attitudes, however, do not sit on one left-right line.
A student may endorse the norm that discrimination is wrong and still worry about whether schools or local services are prepared. Another may favor coexistence in the abstract without having considered who funds language support or labor protection. The value of a multidimensional survey is not to erase these tensions but to reveal how several principles coexist in one person.
Reporting also described an association between support for stricter punishment and skepticism toward the increase in foreign residents. That association does not establish that punitive attitudes directly cause exclusion. Perceived disorder, trust in institutions or general risk sensitivity could influence both.
How a headline creates a second reality
“Forty percent of students” is compact and shareable. Detached from its questionnaire, the figure can support opposing but equally unwarranted stories:
- “Young Japanese people reject foreigners,” turning a limited item into a mandate.
- “Students have been brainwashed online,” denying adolescents' agency.
- “It is only 40%, so nothing changed,” ignoring the movement since 2022.
- “The other 60% favor more immigration,” inventing a binary without seeing the full response distribution.
None follows directly from the reported result. Responsible survey reporting should state not only what a number measures, but also the questions it cannot answer.
What readers and institutions can do
If advocates want support for coexistence, moral insistence that diversity “must be beneficial” is insufficient. The public needs concrete discussion about Japanese-language support in schools, interpretation in public services, housing, labor advice, community mediation and remedies for discrimination.
Students likewise need more than an approved conclusion. They need practice checking sources, reading denominators, separating correlation from causation, hearing affected people and comparing policy choices. Media literacy should not mean “never trust the internet.” It should mean learning to test claims across the internet and other institutions.
The philosophical question: who owns an opinion?
When an opinion disturbs us, we tend to attribute it either to character or manipulation. The student is described as prejudiced, or as a passive victim of an algorithm. Both explanations fail to treat adolescents as people who interpret, hesitate and learn.
An opinion is neither wholly private property nor wholly manufactured by an environment. It forms where family, school, peers, news, recommender systems and direct experience intersect. Responsibility is therefore distributed: individuals must reconsider evidence; schools must cultivate verification; platforms must explain amplification; and politics must address anxiety through facts and workable institutions.
Taking the survey seriously does not mean condemning students or promoting the internet into an all-purpose cause. It means observing change, preserving the limits of the question and moving coexistence from abstract goodwill to concrete institutions. The number is not a conclusion. It marks where the public should investigate and deliberate next.
References
- Japan Sociological Society: 99th annual meeting general research presentations, Oct. 6, 2026
- Kumamoto Nichinichi Shimbun / Kyodo News: report on the student survey, Nov. 7, 2026
- Saitama Shimbun / Kyodo News: report on the student survey, Nov. 7, 2026
- Snaplytics: public Japan X trend record for Nov. 7, 2026
Information checked November 8, 2026, Japan time. The public trend source is a limited third-party snapshot, not a complete X ranking or post count. This article treats internet influence as a reported hypothesis and does not claim that the survey established causation.