Defence policy
Credible retaliation can deter attack.
What this view explainsWhy governments value an allied nuclear guarantee.
What it may missA counterfactual is difficult to prove, and risk is borne unevenly.
Start by separating the choices
An umbrella is not a weapon in the hallway
Japan's 2026 defence white paper says the country receives United States protection that includes nuclear capabilities. That is the government's policy. It is not proof that a weapon prevented a particular attack.
The University of Tokyo survey asked different questions. Combined support for strengthening the American alliance was 45.6%. Support for Japan developing nuclear weapons was 18.9%; support for hosting American weapons was 19.3%. Opposition to the last two proposals was 57.2% and 53.7%. The survey cannot tell us whether the same respondents chose each answer, or why.
The argument depends on where you stand
NATO says nuclear weapons deter attack because an opponent believes they could be used. Nagasaki's mayor starts at the other end of that logic: a leader may decide to use them, so a security policy built on that possibility is dangerous and fragile.
Both are stated positions. Neither supplies the missing counterfactual—what would have happened without the weapons. That is why a claim that deterrence 'worked' needs more than the absence of an attack.
A percentage is not a portable public
Seoul National University's institute reported 61.1% support for South Korean nuclear possession. The institute also warned that general support does not establish willingness to accept economic or diplomatic costs.
That survey used in-person interviews with 1,200 adults. The Japanese survey was online and used a different sample and question set. Putting the percentages side by side can raise a useful question, but it cannot create a league table of national enthusiasm.
The decision does not end when the test ends
British nuclear testing in Australia included 12 atomic detonations from 1952 to 1963. ARPANSA says its 2025 Maralinga survey found no significant change in radioactivity and that earlier dose assessments still apply for the uses they cover.
That conclusion has a boundary. The public summary does not publish the underlying measurements by location, and ARPANSA still describes a precautionary restriction on permanent occupancy near Taranaki. A current scoped safety assessment does not erase past harm or the work of managing Country.