Doomsday Clock and ICAN Forecasting
2
The [Doomsday Clock](https://thebulletin.org/doomsday-clock/), which holds at 100 seconds to midnight as of today, suggests the world has never been closer to a catastrophe involving nuclear weapons. But another advocacy group has [said](https://d3n8a8pro7vhmx.cloudfront.net/ican/pages/2165/attachments/original/1623235224/ICAN-NATO-report-final.pdf?1623235224) it’s “only a matter of time” until some of the world’s most powerful countries agree to make them illegal. That second prediction, made by the International Campaign to Abolish Nuclear Weapons (ICAN) in the lead-up to the 47th G7 summit, cites “numerous” signs of support amongst NATO countries for the UN Treaty on the Prohibition of Nuclear Weapons — the first binding international prohibition on nukes. And while no NATO members, nor any of the nine nuclear powers, have signed on so far, The New York Times [observed](https://www.nytimes.com/2021/06/10/world/europe/nuclear-weapons-treaty-nato.html) that the contingent’s opposition to the ban “appears to have weakened,” calling that a “challenge” that President Joe Biden and his NATO counterparts “may” have to contend with. It’s worth probing both the Doomsday Clock and ICAN frameworks, asking whether they provide any useful signals about the way forward on nuclear risk and proliferation. First, the Doomsday Clock and ICAN worldviews are not completely at loggerheads. The Bulletin of the Atomic Scientists, which sets the clock each year (and which has faced [criticism](https://www.scientificamerican.com/article/how-the-doomsday-clock-could-help-trigger-the-armageddon-it-warns-of/) for its methods), has acknowledged that the treaty brought “much-needed attention to the risks posed by nuclear weapons,” and ICAN similarly referenced the Bulletin in noting the risk of a nuclear weapon being used “is widely considered to be growing.” One could broadly characterize the former as a “pessimistic” view, and the latter as more “optimistic” about the immediate future of nuclear weapons in the world. But they are not mutually exclusive. A scenario where more — or even many — NATO countries agree to the ban, but one of the current nuclear powers or even a non-state actor, ultimately pulls the nuclear trigger, is certainly plausible. As forecasts, however, these reports (and often the news coverage they generate) lack precision, making it difficult to judge whether either worldview is truly “correct.” Although today’s statement by the Bulletin indicated the potential for the “start of a dangerous new nuclear arms race,” it didn’t make a corresponding adjustment to its doomsday prediction like moving the clock closer to “Midnight.” Nor does it contemplate a specific time horizon. Similarly, ICAN doesn’t state how many countries might sign on to the weapons ban, or exactly how long it might take for any of them to do so. Both proclamations could still have value for forecasters — both for thinking through priors, and operationalizing actual forecasts. In a world that’s as close — or closer — to nuclear annihilation than it was during the depths of the Cold War, the chances that several countries, say, [increase their stockpile by 10% or more over the next two years](https://www.metaculus.com/questions/7463/-countries-increase-nuclear-arsenal-by-10/), or that there is at least one death due to [accidental](https://www.metaculus.com/questions/7455/unauthorised-nuclear-detonation-by-2024/) or [deliberate](https://www.metaculus.com/questions/7404/nuclear-detonation-fatality-before-2024/) detonations, could very well be higher than in a world where more and more NATO countries are clamoring to sign an anti-nuke pact, and indeed, are successfully applying pressure to nuclear powers like the United States to disarm. Although there’s still time left before resolution, the Metaculus community isn’t exactly bullish about the treaty’s long-term prospects at quashing nuclear weapons. Currently, the median forecast on the question of whether a country that actually has nuclear weapons will agree to give them up by 2035 [is only 20%](https://www.metaculus.com/questions/492/with-an-international-nuclear-weapons-ban-in-place-will-at-least-one-nuclear-state-actually-give-them-up-by-2035/), while it’s 40% likely that a signatory [will renege on the treaty by 2060](https://www.metaculus.com/questions/5808/signatory-to-ban-on-nukes-to-break-treaty/) — by developing a nuclear weapon anyway. To better evaluate ICAN’s position, we look to this Metaculus question asking *Will at least one NATO state sign on to the Treaty on the Prohibition of Nuclear Weapons by July 15th?* (Note there is [a difference between signing and ratifying the treaty](https://www.icanw.org/campaigner_guide_to_tpnw_signature_and_ratification).) The Doomsday Clock is meant to signal more than just nuclear threats, it’s true, but we can stress-test its nuclear-specific claims—particularly its arguments that “nuclear nations” have undermined tools to manage nuclear risks, and that the chances the world will “stumble into nuclear war…increased,” Will there be at least one fatality due to accidental or unauthorized nuclear detonation by 2024?