A nuclear exchange may have the potential to kill millions or billions of people, and possibly lead to human extinction.
In this post, I rank plausible nuclear exchange scenarios in terms of their potential to cause harm based on three factors: 1) The size of the involved countries’ nuclear arsenals; 2) The size of the involved countries’ populations; 3) The probability of the given nuclear exchange scenario.
Based on my rough prioritization, I expect the following nuclear exchange scenarios have the highest potential for harm:
- Russia and the US
- India and Pakistan
- China and either the United States, India, or Russia
This is the first post in Rethink Priorities’ series on nuclear risks. In this post, I look into which plausible nuclear exchange scenarios should worry us most, ranking them based on their potential to cause harm. In the second post, I explore the make-up and survivability of the US and Russian nuclear arsenals. In the third post, I estimate the number of people that would die as a direct result of a nuclear exchange between NATO states and Russia. In the fourth post, I estimate the severity of the nuclear famine we might expect to result from a NATO-Russia nuclear war. In the fifth post, I get a rough sense of the probability of nuclear war by looking at historical evidence, the views of experts, and predictions made by forecasters. Future work will explore scenarios for India and Pakistan, scenarios for China, the contradictory research around nuclear winter, the impact of several nuclear arms control treaties, and the case for and against funding particular organizations working on reducing nuclear risks.
Toward a better understanding of nuclear risks
A nuclear exchange may have the potential to kill millions or billions of people, and possibly lead to human extinction. There have been many cases where nuclear weapons have almost been launched by mistake (Baum, de Neufville & Barrett, 2018). And if a nuclear exchange — started by accident or on purpose — were to escalate to a full-scale nuclear war, the nuclear detonations could lead to a nuclear winter, a state where soot launched into the atmosphere blocks out enough sunlight to cause a famine so severe and long-lasting that almost everyone on Earth could starve to death before its end.
Because there seems to be a non-negligible probability of a large-scale nuclear exchange, and because the stakes would be so high in the event that a nuclear exchange did escalate, many effective altruists believe reducing nuclear risks should be among the top priorities for the EA community. For example, 80,000 Hours published a problem profile on nuclear security, giving a score of 15 out of 16 on Scale (though it scores relatively low on Solvability and Neglectedness).
But my sense is that some details of the nuclear risks problem area aren’t well-understood by most EAs — for example, how bad nuclear war would actually be, the mechanisms behind nuclear winter, and where EAs that prioritize reducing nuclear risks should donate. In a number of upcoming posts, I’ll try to understand, in somewhat concrete terms, how much harm nuclear war would cause and how plausible nuclear risks are. One of the things I'll do to better understand the risks posed by nuclear winter is review the implications of recent academic literature that is interpreted by some as casting doubt on the science behind the nuclear winter phenomenon. Finally, I’ll also evaluate some of the work being done to reduce nuclear risks. In particular, I’ll focus on a recent treaty that’s been adopted by the United Nations, the Treaty on the Prohibition of Nuclear Weapons (TPNW), which would make the research and use of nuclear weapons illegal in all countries that ratify the treaty.
Focusing on the most troubling nuclear risks
In the first few posts, I’ll consider the probability and severity of several nuclear war scenarios, looking separately at the amount of harm that would be caused from both the short-termist perspective and the long-termist one.
It would be intractable to estimate the impacts of every imaginable nuclear war scenario. So I instead focus on nuclear war scenarios which I expect make up the majority of the expected harm that would be caused by nuclear war. These are scenarios in which:
The specified countries have relatively large nuclear arsenals. This an important consideration for two reasons: first, because the size of the nuclear arsenal is a major factor in how severe the direct effects of a nuclear exchange are — more nuclear weapons can cause many more deaths. Second, whether a nuclear exchange leads to a nuclear winter depends in large part on the number and size of nuclear weapons detonated. It would take a lot of nuclear weapons to produce a nuclear winter severe enough to cause a worldwide famine that could lead to human extinction. Given that I'm most worried about nuclear winter scenarios that pose an extinction risk, I believe we should focus on nuclear exchange scenarios that would involve large nuclear arsenals.
The countries involved have large populations. This is an important consideration for similar reasons: first, because the population size of a country is an important factor in determining how many people could die as a direct result of the nuclear detonations. Second, whether a given nuclear exchange would lead to a nuclear winter depends on how much smoke is produced from the burning of cities during the exchange. Countries with larger, densely populated cities have much more flammable material. This means that a nuclear exchange involving densely populated countries would be more likely to lead to nuclear winter, all else equal.
The specific conflict scenarios are reasonably probable. This is important because the expected harm is higher in nuclear exchange scenarios that are more likely to actually happen, all else equal.
When I consider these factors for all 9 nuclear weapons possessor states, I get a rough ranking of various plausible nuclear exchange scenarios in terms of their expected dis-value:
The ranking is based on a crude scoring system that approximates the expected harm that would be caused by each scenario. Each of the factors that informs how terrible a given nuclear exchange would be — the size of the involved countries' nuclear arsenals, the size of the populations of the involved countries, and the probability of the specific scenario — were assigned a score of 1 (shaded in green) , 2 (yellow), or 3 (red), where a factor with a score of 1 should worry us less, and a factor with a score of 3 should worry us a lot. For each scenario, the factor scores were summed together to produce the Expected Harm score.
Note that some endnotes are embedded in the table image but can be seen here: source for arsenal size; source for median war probability; note on non-state actors.
Based on this rough prioritization, I’ll spend several posts looking at the amount of harm we would expect to see caused by the following nuclear exchange scenarios:
A scenario where the US and Russia use nuclear weapons, intentionally or accidentally, in a conventional nuclear exchange.
A scenario where India and Pakistan use nuclear weapons, intentionally or accidentally, in a conventional nuclear exchange.
A scenario where China uses nuclear weapons, intentionally or accidentally, in a conventional nuclear exchange with either the United States, India, or Russia.
Edits and Corrections
July 17 2019 — I replaced a simplified version of the table ranking potential conflict scenarios on the basis of nuclear arsenal size, population size, and scenario probability with a more detailed version. I also added a brief explanation of how the Expected Harm scores were calculated.
This essay is a project of Rethink Priorities. It was written by Luisa Rodriguez. Thanks to Peter Hurford, Marinella Capriati, Ida Sprengers, Marcus A. Davis, and Neil Dullaghan for their valuable comments. Thanks also to Matt Gentzel, Seth Baum, and Carl Shulman for providing guidance and feedback on the larger project. If you like our work, please consider subscribing to our newsletter. You can see all our work to date here.
Apps, P. (2015). PS21 Great Power Conflict Report (Rep.). Project for the Study of the 21st Century. Retrieved from: https://www.scribd.com/document/289407938/PS21-Great-Power-Conflict-Report
Baum, S., de Neufville, R., & Barrett, A. (2018). A model for the probability of nuclear war. Social Science Research Network (SSRN). https://doi.org/10.2139/ssrn.3137081
Daniel Ellsberg on the creation of nuclear doomsday machines, the institutional insanity that maintains them, and a practical plan for dismantling them [Audio blog interview]. (2018, September 24). Retrieved April 14, 2019, from https://80000hours.org/podcast/episodes/daniel-ellsberg-doomsday-machines/
Kristensen, H. M., & Korda, M. (2019). Russian nuclear forces, 2019. Bulletin of the Atomic Scientists. https://doi.org/10.1080/00963402.2019.1580891
Kristensen, H. M., & Norris, R. S. (2014). Israeli nuclear weapons, 2014. Bulletin of the Atomic Scientists. https://doi.org/10.1177/0096340214555409
Kristensen, H. M., & Norris, R. S. (2016). Pakistani nuclear forces, 2016. Bulletin of the Atomic Scientists. https://doi.org/10.1080/00963402.2016.1241520
Kristensen, H. M., & Norris, R. S. (2017). Indian nuclear forces, 2017. Bulletin of the Atomic Scientists. https://doi.org/10.1080/00963402.2017.1337998
Kristensen, H. M., & Norris, R. S. (2018a). Chinese nuclear forces, 2018. Bulletin of the Atomic Scientists. https://doi.org/10.1080/00963402.2018.1486620
Kristensen, H. M., & Norris, R. S. (2018b). North Korean nuclear capabilities, 2018. Bulletin of the Atomic Scientists. https://doi.org/10.1080/00963402.2017.1413062
Kristensen, H. M., & Norris, R. S. (2018c). United States nuclear forces, 2018. Bulletin of the Atomic Scientists. https://doi.org/10.1080/00963402.2018.1438219
McIntyre, P. (2016, April 12). How you can lower the risk of a catastrophic nuclear war. Retrieved from https://80000hours.org/problem-profiles/nuclear-security/
Talley, I. (2019, March 22). U.S. says Iran poised to resume work on nuclear weapons. Wall Street Journal. Retrieved May 19, 2019, from https://www.wsj.com/articles/u-s-says-iran-poised-to-resume-work-on-nuclear-weapons-11553263221
Treaty on the prohibition of nuclear weapons – UNODA. (n.d.). Retrieved May 19, 2019, from https://www.un.org/disarmament/wmd/nuclear/tpnw/
See pages 27-32 for an itemized list of the US-Russia near misses (Baum, de Neufville & Barrett, 2018). ↩︎
Note, they now recommend people use their in-depth interview with Daniel Ellsberg as a source of information on nuclear security. ↩︎
Arsenal data from the Bulletin of the Atomic Scientists’ Nuclear Notebooks. ↩︎
I believe the Project for the Study of the 21st Century (PS21) Great Power Conflict Report has several typos (Apps, 2015). I present what I believe to be the correct values (and the values I use in my analysis) here. ↩︎
While a nuclear detonation by a non-state actor (terrorist) looks plausibly quite harmful in expectation, it’d be very difficult to analyze, as there’s no single terrorism scenario to consider. I therefore leave a discussion of the potential harm caused by nuclear terrorism for future work. ↩︎
I see the expected harm scale is logarithmic (as the overall score is the some of the component scores), but what is the ratio between the badness of 2 consecutive scores (e.g. "badness of score 12"/"badness of score 11")?
This post was awarded an EA Forum Prize; see the prize announcement for more details.
My notes on what I liked about the post, from the announcement:
I haven't seen anyone seriously argue for this claim and I don't think it's true or true-adjacent.
Thanks, that's fair. Edited to say 'possibly lead to human extinction.'
I'm excited to read this series!
I don't think this is quite right. Robock 2007 finds a severe nuclear winter effect from an exchange with just 100x 15kt bombs. AFAIK, the only country with an arsenal below that threshold today is North Korea, which would suggest that — on Robock's modelling at least—any bilateral exchange involving nuclear powers other than NK is large enough to pose a significant risk of nuclear winter.
Thanks for raising this! You’re right — Robock 2007 does find that even a relatively small nuclear exchange would have devastating climate effects that would probably cause a famine. But my understanding (from both Robock’s paper and this report on the impact of a regional nuclear exchange on the global food supply) is that a regional nuclear war, while horrible, would not cause a severe enough nuclear winter to risk human extinction.
I’ll clarify in the post that I’m most worried about nuclear exchange scenarios that would lead to a nuclear winter severe enough as to pose an extinction risk.
Are you only worried about extinction, or existential risk more broadly? I talk about a number of ways that a catastrophe the size of a regional nuclear war could lead to long-term impact, including instability leading to full-scale nuclear war.
I hope that the risk of nuclear war between China the US remains low. However, 12 out 16 times that there has been a switch in which is the most militarily powerful country in the world, there has been war (Destined for War by Graham Allison) (though we should not take that literally for the current situation).
Absolutely. I agree that the US and China may be on a collision course for war, and that this should make us a bit more worried about a US-China nuclear exchange. I’m in the process of modeling the impacts of such an exchange so we can get a better sense of just how worried we should be.
For me, the ones we should worry about are the ones which are most likely in the next 5-9 years.
If I understand the history (eg. several occasions when a USSR-USA nuclear war was almost triggered due to software errors) and the scenario analysis (ie. all the ways something similar could happen via the smaller nuclear powers) - the biggest near term likelihood is of an "accidental" regional nuclear war in the Middle East, perhaps because one country erroneously believes it is being attacked, or because one country has had its chain of command hacked or hijacked in an unexpected way.
I believe that the high likelihood from the smaller Middle East powers comes not from high likelihood of any one given scenario, but rather from the large number of pathways to a nuclear exchange, a "basket of risks", each one individually being of low likelihood, but collectively adding up to a likelihood far bigger than any one scenario.
In case anyone thinks that a regional nuclear exchange wouldn't be too bad, Robock et al have discussed direct consequences such as "nuclear autumn", and cascading scenarios are obviously a risk.
Paul Ingram (www.basicint.org/our-staff/paul-ingram) would be excellent to discuss this with, as he specialises in this area and has done since the late 1980s.
This series (#2, #3) has begun as the most interesting-to-me on the Forum in a long time. Thanks very much. If you have written or do write about how future changes in arsenals may change your conclusions about what scenarios to pay the most attention to, I'd be interested in hearing about it.
In case relevant to others, I found your spreadsheet with raw figures more insightful than the discrete system in the post. To what extent do you think the survey you use for the probabilities of particular nuclear scenarios is a reliable source? (I previously distrusted it for heuristic reasons like the authors seeming to hype some results that didn’t seem that meaningful.) I'm interested because, as well as the numbers you use it for, the survey implies ~15% chance of use of nuclear weapons conditional on a conventional conflict occurring between nuclear-armed states, which seemed surprisingly low to me and would change my thinking about conflicts between great powers in general if I believed it.
Thanks for your comments — I’m glad to hear you’re enjoying the series!
I haven’t written about this yet, but I’ll consider working it in as I continue to explore the topic in the next few months. I’ll update this thread if I do.
I’ll be sharing a post on the probability of a US-Russia nuclear war soon. It talks a little bit about the relative merits and weaknesses of some of these probability estimates.
Also, I see you’ve left some great feedback on posts 2 and 3. I’ll be replying to those comments shortly.