{"id":1042,"date":"2024-10-29T12:00:00","date_gmt":"2024-10-29T12:00:00","guid":{"rendered":"https:\/\/forecastingresearch.org\/?post_type=research&p=1042"},"modified":"2026-04-28T14:07:39","modified_gmt":"2026-04-28T14:07:39","slug":"nuclear-risk","status":"publish","type":"research","link":"https:\/\/forecastingresearch.org\/research\/nuclear-risk","title":{"rendered":"Can Humanity Achieve a Century of Nuclear Peace?"},"content":{"rendered":"\n

Abstract<\/h2>\n\n\n\n

While the world has avoided large-scale nuclear war, questions remain\nabout the role of chance versus policy choices in preventing such\nevents. This study systematically assesses expert beliefs about the\nprobability of a nuclear catastrophe by 2045, the centenary of the\nbombings of Hiroshima and Nagasaki. We define a nuclear catastrophe as\nan event where nuclear weapons cause the death of at least 10 million\npeople. Through a combination of expert interviews and surveys, 110\ndomain experts and 41 expert forecasters (\u201csuperforecasters\u201d) predicted\nthe likelihood of nuclear conflict, explained the mechanisms underlying\ntheir predictions, and forecasted the impact of specific tractable\npolicies on the likelihood of nuclear catastrophe. Experts assigned a\nmedian 5% probability of a nuclear catastrophe by 2045, while\nsuperforecasters put the probability at 1%. Factors contributing to\nhigher risk estimates included ongoing geopolitical tensions, the\nproliferation of nuclear weapons, and technological vulnerabilities.\nLower risk estimates highlighted the continued effectiveness of nuclear\ndeterrence. Although Russia and NATO was the adversarial domain thought\nmost likely to cause a nuclear catastrophe, experts believe that risks\nare dispersed roughly uniformly across regional conflict theaters\n(Russia and NATO, China and the USA, the Korean Peninsula, India and\nPakistan, and Israel and Iran). Participants believe that the\nimplementation of a bundle of six tractable policies, including the\nestablishment of a crisis communications network and the implementation\nof failsafe reviews, would together halve the risk of a nuclear\ncatastrophe.<\/p>\n\n\n\n

\n
View the full PDF report \n \n<\/svg>\n\n \n<\/svg><\/a><\/div>\n<\/div>\n\n\n\n
\n
Acknowledgments<\/summary>\n

This research would not have been possible without the generous support of Open Philanthropy.<\/p>\n\n\n\n

We thank the following people, who kindly served as expert interviewees: James Acton, Catherine Dill, Robert Einhorn, Peter Hayes, Feroz Khan, Frank O\u2019Donnell, David Santoro, Manpreet Sethi, Sir Graham Stacey, Dmitry Stefanovich, Tatsujiro Suzuki, Jon Wolfsthal, and Tong Zhao. Their expertise helped to inform the direction of the project, but we note that the content of the project and this report don\u2019t necessarily reflect their views. <\/p>\n\n\n\n

We also thank Peter Scoblic for helpful feedback on survey questions. We greatly appreciate the assistance of Josh Rosenberg, Kayla Gamin, Sam Glover, Harrison Durland, Nikitas Angeletos Chrysaitis, Catherine Wu, Amory Bennett, Kaitlyn Coffee, Coralie Consigny, Rhiannon Britt, and Rebecca Ceppas de Castro throughout the project. <\/p>\n\n\n\n

Lastly, we extend our gratitude to our research participants for their invaluable contributions.<\/p>\n<\/details>\n<\/div>\n\n\n\n

Executive Summary<\/h2>\n\n\n\n

This study is the largest systematic survey of subject matter experts\non the risk posed by nuclear weapons. In addition to experts, we\nsurveyed forecasters with a strong track record of accuracy\n(\u201csuperforecasters\u201d). The study summarizes responses on two\ncomplementary surveys taken a month apart: the first survey focused on\nrisk pathways and the second on policy responses.<\/p>\n\n\n\n

Participants<\/h3>\n\n\n\n

A total of 151 participants (110 experts and 41 superforecasters)\ncompleted the full first survey, and 148 participants completed both\nsurveys (109 experts and 39 superforecasters). Most respondents engaged\ndeeply with the questions. The median expert or superforecaster\nparticipant reported spending nine hours completing the two surveys and\nwrote around 4,200 words to explain their forecasts.<\/p>\n\n\n\n

Key results<\/h3>\n\n\n\n

Forecasts of nuclear\ncatastrophe by 2045<\/h4>\n\n\n\n

We asked participants to estimate the probability that, before 2045,\none or more incidents involving nuclear weapons will cause the death of\nat least 10 million people. The median expert forecast was 5% and the\nmedian superforecaster response was 1%. The median forecast from a\nsample of the US public was 10% (see Figure 1). Respondents thought that\na nuclear conflict between Russia and NATO was the adversarial domain\nmost likely to be the cause of a nuclear catastrophe of this scale;\nhowever risk was dispersed relatively evenly among all adversarial\ndomains we asked about, which also included China and the USA, the\nKorean Peninsula, India and Pakistan, and Israel and Iran.<\/p>\n\n\n\n

Participants who were more concerned about nuclear risk often\nmentioned ongoing military conflicts between nuclear powers, the\nproliferation of nuclear weapons, the development of new military\ntechnologies, and the weakening of international arms control\nagreements. Participants less concerned about nuclear risk emphasized\nthe long-standing effectiveness of deterrence, improvements in safety\nmechanisms, and the assumption that most nuclear states will act\nrationally.<\/p>\n\n\n\n

\"Figure
Figure 1:<\/strong> Distribution of forecasts of the probability of nuclear catastrophe by 2045. We jitter the data points horizontally to allow for better visualization of the distribution of forecasts. Horizontal variation within each group serves no other empirical purpose. The y-axis uses a logarithmic scale to informatively show variation in forecasts in the 0\u201310% range.<\/figcaption><\/figure>\n\n\n\n

Factors influencing risk<\/h4>\n\n\n\n

Participants saw conflict between adversarial countries and new\nactors acquiring nuclear weapons as the major drivers of nuclear risk.\nWe discuss a large set of events in this study, but to summarize three\nkey events:<\/p>\n\n\n\n

    \n
  1. The median participant in this study reported that violent\nconflict between Russia and NATO would triple the risk of a nuclear\ncatastrophe. Both experts and superforecasters reported a low\nprobability of this event occurring: median forecasts of 5% and 1.8%,\nrespectively;<\/li>\n\n\n\n
  2. Participants were more concerned about the likelihood of a\nChinese invasion of Taiwan, with the median expert estimating a 25%\nchance of this occurring by 2030. The median expert reported that this\nevent would roughly double their forecast of the risk of nuclear\ncatastrophe;<\/li>\n\n\n\n
  3. The median expert also put a 25% chance of Iran acquiring nuclear\nweapons by 2030. Should this event occur, their forecast of the risk of\ncatastrophe would increase by 50%.<\/li>\n<\/ol>\n\n\n\n

    Many of the events we asked about did not influence the median\nrespondent\u2019s forecast of the risk of nuclear catastrophe. These events\ninclude: summits between adversarial countries, a nuclear weapons test\nby North Korea, ballistic missile submarines becoming more detectable,\nand an accidental non-test detonation of a nuclear weapon.<\/p>\n\n\n\n

    Effects of policies<\/h4>\n\n\n\n

    We asked participants about their beliefs on the effectiveness of\nseveral policy options aimed at reducing the risk of a nuclear\ncatastrophe. The distribution of relative risk scores for the six\npolicies we included are shown in Figure 2. We also asked participants\nto rank the policies by how much they would like to see each policy\nimplemented and how much they would support funding aimed at\nimplementing the policies.<\/p>\n\n\n\n

    \"Figure
    Figure 2:<\/strong> Violin plots showing distribution of relative risk associated with each policy, and all six of these policies implemented together. The relative risk is the relative change in probability of nuclear catastrophe conditional on policy implementation. The group median is shown in text. The thicker bar within each violin shows the interquartile range (25th<\/sup> to 75th<\/sup> percentile forecasts), and the thin line shows the range of forecasts minus outliers.1<\/a><\/sup><\/figcaption><\/figure>\n\n\n\n

    Two policies emerged as clear favorites: a crisis communications network2<\/a><\/sup> and nuclear-armed states implementing failsafe reviews.3<\/a><\/sup> The median expert thought that a crisis communications network would reduce the risk of a nuclear catastrophe by 25%, and failsafe reviews would reduce it by 20%. The superforecasters were less optimistic about the effects of the policies, with median relative risk of 15% and 10%, respectively. We also asked experts to say how their forecast of nuclear catastrophe by 2045 would change if all six policies we described were implemented. The median expert thought that the combined bundle of policies would halve the risk of a nuclear catastrophe and the median superforecaster thought it would reduce risk by 42%.<\/p>\n\n\n\n

    Probability of policy\nimplementation<\/h4>\n\n\n\n

    We asked participants to forecast the probability that each policy\nwould be implemented (on a time frame consistent with a decision to\nimplement the policy within the next three years). The median expert\nforecasted a 15% probability of implementation for the crisis\ncommunications network and a 15% probability of implementation for the\nfailsafe reviews policy. Superforecasters estimated probabilities of 10%\nand 7%, respectively. Both groups thought that funding could improve the\nprobability of policy implementation. Conditional on $500 million of\nfunding being put towards the goal of having the policy implemented, the\nmedian expert forecasted a 25% chance that a crisis communications\nnetwork would be established and a 30% chance of the failsafe reviews\npolicy being implemented. The median superforecaster\u2019s forecast rose to\n18% and 10%, respectively.<\/p>\n\n\n\n

    Limitations and next\nsteps<\/strong><\/h3>\n\n\n\n

    While this study is the largest and most comprehensive survey of\nnuclear experts\u2019 beliefs about nuclear risk, there are some important\nlimitations that we hope to address in follow-up work. First, most\nparticipants were based either in the USA or Western Europe. The number\nof expert participants from South Asia was similar to the number from\nthe USA, but there were very few participants from East Asia, Eastern\nEurope, or the Middle East. Although this was the largest survey of its\nkind, the sample size was still relatively small, limiting the\nstatistical inferences we can make from the results. The lists of\npolicies we included may not have represented the full range of\nviewpoints on reducing nuclear weapons risk.<\/p>\n\n\n\n

    Despite these limitations, this study provides important insight into\nviews on the risk of nuclear war. It clarifies experts\u2019 views on how the\nworld could best mitigate those risks and maintain nuclear peace until\nand beyond the centennial anniversary of the first and so far only use\nof nuclear weapons in warfare. We believe there is a role for\nquantitative forecasts to build upon this work, improving our\nunderstanding of beliefs about nuclear weapons use and how the\nassociated risks change over time.<\/p>\n\n\n\n

    1. Background<\/h2>\n\n\n\n

    Since 1945, the world has lived under the threat of nuclear weapons.\nSo far, we have managed to avoid the disaster that would be nuclear war.\nHave we been lucky? Or was the probability of nuclear war always low?\nHow confident should we be that humanity will make it to the\nhundred-year anniversary of the bombings of Hiroshima and Nagasaki\nwithout any further nuclear catastrophes?<\/p>\n\n\n\n

    These questions are difficult to answer. However, understanding views\non the magnitude and nature of this risk can inform decisions about how\nto best prioritize resources to improve humanity\u2019s prospects.<\/p>\n\n\n\n

    The goal of this study was to understand current views about the potential causes of a nuclear catastrophe. We systematically collected the views of subject matter experts, highly accurate forecasters (\u201csuperforecasters\u201d),4<\/a><\/sup> and members of the public on the threat of nuclear weapons. Specifically, we asked participants to forecast the probability that we will survive a century without another nuclear catastrophe, and the effects of events and policies that may alter this outlook. The result is the largest survey ever conducted of policy experts\u2019 views on the magnitude of risks posed by nuclear weapons. A total of 110 experts collectively spent over 1,300 hours and wrote 520,000 words developing forecasts and explaining their reasoning.<\/p>\n\n\n\n

    1.1 Previous\nforecasts of nuclear weapons risks<\/h3>\n\n\n\n

    While our study is the largest survey collecting forecasts from\nnuclear weapons policy experts, it is not the first. Over the past three\ndecades, there have been at least four other surveys asking subject\nmatter experts to make predictions about nuclear weapons risk:<\/p>\n\n\n\n