Responding to a nuclear attack

This brief note is on Wikiversity to invite others to provide alternative responses to this question, adding relevant, substantive references, moderated by the Wikimedia rules that invite contributors to “be bold but not reckless,” contributing revisions written from a neutral point of view, citing credible sources -- and raising other questions and concerns on the associated '“Discuss”' page.
This article uses ISO 8601 dates except for References, which are controlled by standard Wikidata formatting, and direct quotes. In the initial author's experience, ISO 8601 dates seem to make it easier to remember dates and to compute differences between them.

What's the best response to a nuclear attack?

That's a difficult question. The opposite is much easier:

  • What's the worst response to a nuclear attack?
Simulation of a nuclear war between Russia and the US.[1]
The evidence summarized in this article suggests that the worst worst response to a nuclear attack would be a nuclear response.
If you think otherwise, please revise this article accordingly, subject to the standard Wikimedia Foundation rules of writing from a neutral point of view citing credible sources. Or post your concerns to the "Discuss" page associated with this article.
Percent of the world's population dead from a nuclear war per simulations by an international team of 10 scientists who specialize in modelling climate, food production, and economics[2] with models fit thereto. The vertical axis is the percent of the world's population expected to die within a few years after a one-week long nuclear war that injects between 1.5 and 150 Tg (teragrams = million metric tons) of smoke (soot) into the stratosphere, shown on the top axis.[3] The bottom axis is the total megatonnage (number of nuclear weapons used times average yield) simulated to produce the quantity of soot plotted on the top axis. "IND-PAK" marks a range of hypothetical nuclear wars between India (IND) and Pakistan (PAK). "USA-RUS" marks a simulated nuclear war between the US (USA) and Russia (RUS). "PRK" = a simulated nuclear war in which North Korea (the People's Republic of Korea, PRK) used their existing nuclear arsenal estimated at 30 weapons with an average yield of 17 kt[4] without nuclear retaliation by an adversary, as recommended in this article.

This conclusion is supported by the accompanying plot summarizing climate simulations by an international interdisciplinary team of 10 scientists who specialize in mathematical and statistical modeling of climate, food production, and economics. Five of their scenarios describe hypothetical nuclear wars between India and Pakistan that loft between 5 and 47 Tg (teragrams = millions of metric tons) of smoke (soot) to the stratosphere, where it will linger for years covering the globe and reducing the amount of solar radiation reaching the earth. That in turn will substantially reduce the production of food for humans. The resulting impact on the global economy means that between 4 and 40 percent of humanity will likely starve to death if they do not die of something else sooner. A hypothetical nuclear war between the US and Russia could lead to the deaths of over 80 percent of humanity with death tolls of roughly 99 percent in the US, Russia, Europe, and China. In any of these scenarios, between 90 and 95 percent of the deaths would be in countries not officially involved in the nuclear exchange.[5]

This claim is clearer, more succinct, and stronger than the Joint Statement of the Leaders of the Five Nuclear-Weapon States on Preventing Nuclear War and Avoiding Arms Races, "that a nuclear war cannot be won and must never be fought", issued 2022-01-03 by the leaders of the first five nuclear-weapon states.[6] This repeated a statement made 1987-12-11 by US President Ronald Reagan and USSR head of state Mikhail Gorbachev.[7]

In the following we review the evidence for and against this claim and then comment on the credibility of the logic that led to the creation of the world's current nuclear arsenals and seems to be driving the current "modernization" programs in the US, Russia, China and elsewhere.

Summary of research on the consequences of a nuclear war

edit

It is theoretically possible that a nuclear exchange would end like World War II with no more than two nuclear weapons being used. It is also theoretically possible that nuclear weapons in a new war would only target deserted areas like the locations where more than 2,000 tests of nuclear weapons have been conducted so far.[8] Either of those scenarios would increase the level of harmful background radiation worldwide leading to increases in the rates of cancer, birth defects and genetic mutations, but would otherwise not likely have an immediate impact a large portion of humanity.[9]

However, a nuclear war with such negligible results is highly unlikely. More likely is the deaths in a few hours or days of tens or hundreds of millions of humans.[10] More would die of radiation poisoning over the next few months and years.[11] If more than a few dozen nuclear weapons are used, then "nuclear war would also produce nearly instantaneous climate change that among other effects, would threaten the global food supply. Even a regional nuclear war ..., such as between India and Pakistan,[12] in which less than 3% of the world’s nuclear weapons stockpiles were detonated in urban areas, would suddenly decrease the average global temperature by 1°C–7°C [2°–13°F], precipitation by up to 40%, and sunlight by up to 30%. ... Such a conflict would decrease crop production to an extent that it could seriously threaten world food security and even trigger global famine",[13] according to Robock and Prager (2021). In theory, crop losses of between 10 and 25 percent for 5-10 years[14] might not threaten a global famine or even an increase in malnutrition if people ate more plant-based foods and less meat. In practice, famines never work that way: There is hoarding, and many who do not die of starvation succumb to diseases or secondary wars driven by the food insecurity, according to Helfand (2013). Nobel Prize Economist Sen observed that, "no famine has ever taken place ... in a functioning democracy".[15] This generalizes the observation that Ireland was a net food exporter during its infamous potato famines of the nineteenth century.[16] Xia et al. (2022, Table 1) estimated that between 4 and 85 percent of humanity would starve to death if they did not die of something else sooner in the nuclear wars they simulated, with between 90 and 95 percent of the fatalities being in countries not directly involved in the hostilities.

In the spring of 1961, "The total death toll as calculated by the Joint Chiefs of Staff [top US military leaders], from a U.S. first strike aimed at the Soviet Union, its Warsaw Pact satellites, and China, would be roughly six hundred million dead. A hundred Holocausts", according to Daniel Ellsberg, who served as a nuclear war planner for presidents Eisenhower, Kennedy, Johnson and Nixon[17] before releasing "The Pentagon Papers" in 1971. Six hundred million was roughly 20 percent of the total human population on earth in 1961, and that didn't count any in the US who might be killed in retaliation. In 1957, roughly 4 years earlier, Mao Zedong, then the Chairman of the People's Republic of China, had reportedly said that a nuclear war could kill a third of humanity, perhaps half, "but imperialism would be razed to the ground, and the whole world would become socialist."[18]

Turco et al. (1983) published the first predictions of a nuclear winter based on climate modeling that considered smoke anticipated from fires started by a massive nuclear weapons exchange between the US and the Soviet Union. They found that "average light levels can be reduced to a few percent of ambient and land temperatures can reach -15° to -25°C [5° to -4°F]" with smoke transported from the Northern to the Southern Hemisphere, all of which "could pose a serious threat to human survivors and to other species." Various teams have published comparable analyses since then with different and increasingly sophisticated models, beginning with Aleksandrov and Stenchikov (1983), with similar conclusions.[19] Coup et al. (2019) predicted hard freezes in the summer in most of the Northern Hemisphere including the US, Russia, and most of Europe during the first three years following such a war, where temperatures drop below −4°C [25°F], making it impossible to grow crops in those regions. China would suffer a similar fate, with only its southeast portion remaining above freezing in the summer. Much of Southern Mexico, Central and South America, and the Southern Hemisphere would also be negatively impacted, but not to the same extent. These climate modeling results make Mao's predictions from 1957 seem wildly optimistic: Any humans in the US, Canada, or most of Eurasia who survived the nuclear exchange would have extreme difficulties finding enough to eat -- "imperialism razed to the ground", according to Mao. However, crop yields in most of the rest of the world would also be extremely depressed, which Mao had not considered. The results would threaten famine vastly worse than what has been predicted following a nuclear war between India and Pakistan.[20]

Of course, no one knows for sure how many people would die directly and indirectly from a nuclear war. However, it should be obvious to at least some if not most people that the worst response to a nuclear attack would be a nuclear response:

  • A nuclear response to a nuclear "warning shot" with minimal destruction could too easily escalate until the nuclear arsenals of all parties were expended and the life expectancy of all survivors worldwide was dramatically reduced.
  • Alternatively, a nuclear response to a massive first strike against a thousand cities would most likely increase the death toll and reduce the life expectancy of survivors in the country responding with nuclear weapons (and, of course, in other countries not officially involved).
  • It is possible that a nuclear response could deter further uses of nuclear weapons and reduce the length and severity of the war and its global impact. However, this outcome seems unlikely given the record of history.

Turcotte (2022) concluded that if the 2022 Ukraine 'conflict ends without the annihilation of our species, it should nonetheless be regarded as a planet-wide near-death experience, and the “Peoples of the United Nations” should demand the total elimination of nuclear weapons as quickly as humanly possible, as well as the establishment of new common security measures that will move us much closer to sustainable peace throughout the world.' In spite of this concern, Turcotte recommended military action to support Ukraine but short of declaring war on Russia.

Leading experts have made alarming comments about the likelihood of a nuclear attack, possibly by a terrorist organization. In 2004 Bruce Blair, president of the Center for Defense Information wrote: "I wouldn't be at all surprised if nuclear weapons are used over the next 15 or 20 years, first and foremost by a terrorist group that gets its hands on a Russian" or Pakistani nuclear weapon.[21]

Other experts seemed even more concerned: A nuclear terrorist attack in the US was considered "more likely than not" within the next five to ten years, according to Professor Robert Gallucci of the Georgetown University School of Foreign Service in 2006 or in the next decade per former U.S. Assistant Secretary of Defense Graham Allison in 2004.[22]

The Wikipedia article on "National Response Scenario Number One" describes "the United States federal government's planned response to a nuclear attack." It focuses primarily on "the possible detonation of a small, crude nuclear weapon by a terrorist group in a major city, with significant loss of life and property."[23] That article discusses preparing for a nuclear attack but not how to respond.

Nevertheless, if the worst response to a nuclear attack is a nuclear response, that has other policy implications for leaders of nuclear and non-nuclear countries world wide. However, an analysis of those implications will be left for future work.[24]

Credibility of military leaders and national security experts

edit
  • Never attribute to malice that which is adequately explained by stupidity. (Hanlon's razor)
  • Never attribute to malice or stupidity that which can be explained by moderately rational individuals following incentives in a complex system. (Hubbard's clumsier correlary.[25])

The history of armed conflict should raise questions about the credibility of those advocating use of military force: In all major armed conflicts in history, at least one side has lost. Often the official winners lost substantially more than they gained.

Research on expertise

edit

The history of armed conflict is consistent with the research by Kahneman and Klein (2009) in their conclusion that

expert intuition is learned from frequent, rapid, high-quality feedback.

In particular, military leaders in combat can get frequent, rapid high-quality feedback on their ability to deliver death and destruction to designated targets. However, no one can get such feedback about how to win wars or how to promote broadly shared peace and prosperity for the long term. This is discussed in more detail in the Wikiversity article on "Expertise of military leaders and national security experts". That article documents how experts without such feedback can be beaten by simple rules of thumb developed by intelligent lay people.[26]

As the time since the atomic bombings if Hiroshima and Nagasaki increases, the intuition that political and military leaders have about nuclear weapons gets worse, because that history tells them that they can use more military force, even threatening to use nuclear weapons, without seriously risking a nuclear war. That intuition increasingly threatens the entirity of humanity.

Increasing risks with nuclear proliferation

edit

Narang and Sagan, eds. (2022) The Fragile Balance of Terror: Deterrence in the New Nuclear Age includes 8 chapters by 12 authors reviewing the literature on different aspects of nuclear deterrance today. They raised many questions about the applicability of Cold War analyses of deterence in an age with an increasing number of nuclear weapon states. They mentioned numerous concerns including the following:

  • During terrorist attacks in Mumbai in 2008, someone called called Pakistani president Zardari claiming to be Indian foreign minister Mukherjee threatening to attack Pakistan. That crises was diffused without escalation after US secretary of state Condoleezza Rice called Mukherjee, who assured her that he had not placed such a call, and India was not planning to attack Pakistan. If someone claiming to be a US official had placed a similar call to Kim Jong Un while Donald Trump was President of the US, the result may not have been as benign.[27]
  • "In January 2018, the Hawaii emergency management system issued an incoming missile warning alert adding, 'this is not a drill.'" The US did not respond, because (a) they had redundant early warning systems that did not indicate an incoming missile, (b) professional operators in Hawaii promptly acknowledged the mistake, and (c) no one in the US seriously expected such an attack. If this had happened in North Korea, none of these three restraining conditions were present: (a) They did not have redundant warning systems. (b) Operators are killed, not just fired in North Korea for making a mistake like that. (c) US "President Trump was threatening 'fire and fury' if North Korean nuclear and missile tests continued."[28]
  • In 2019 India bombed an alleged terrorist training camp in Balakot, Pakistan. This was "the first time a nuclear weapons state has bombed the undisputed territory of another nuclear weapons state."[29]
  • In 2020, Chinese and Indian troops engaged in hostilities along their disputed border with fatalities on both sides, "for the first time in almost half a century. Intense conflict between three nuclear powers simultaneously is no longer a remote possibility.[30]

Beyond this, Richard Ned Lebow said, "There’s all kinds of empirical evidence that a deterrence strategy is as likely to provoke the behavior it seeks to prevent as not."[31]

System accidents

edit

The concept of "normal accidents" or "system accidents" seems important here. Research in that area has established that it is impossible to design and manage complex systems to ultra-high levels of reliability. Maintenance on redundant systems is often deferred, because responsible managers are often reluctant to spend money fixing something that works.[32] And procedures are sometimes secretly modified by people with different priorities from their management. For example, at least between 1970 and 1974 the codes in US Air Force launch control centers for Intercontinental ballistic missiles were all set continuously to 00000000.[33] This clearly negated the claim that only the President of the US could order the use of US nuclear weapons, secured by secret codes carried in a briefcase (called the "nuclear football") near the President at all times. Similarly, former US Secretary of Defense William J. Perry has said an actual nuclear attack on the US is far less likely than a report of one generated by a malfunction in the US nuclear command, control, and communications systems.[34]

A tragic example of a system accident is the Sinking of MV Sewol, 2014-04-16. It sank with over twice its rated load under the command of a substitute captain. The regular captain had complained of deferred maintenance threatening the stability of the vessel; he said the company had threatened to fire him if he continued to complain.

As of this writing, it has been over 77 years since nuclear weapons were detonated in hostilities. As noted above, that history feeds human intuition that we can safely be more aggressive in developing, deploying and threatening the use of nuclear weapons without seriously risking nuclear Armageddon. People who disagree like the Union of Concerned Scientists with their Doomsday Clock are dismissed as unrealistic, like Chicken Little.

Human psychology and the role of the media

edit

When people are attacked, it can sometimes be difficult to control their responses, which are driven by instinctive reactions often characterized as irrational. Johnson (2004) documented how these instinctive reactions exist, because they provided survival benefits to our ancestors over hundreds of thousands and millions of years of evolutionary history. These instincts may, however, push us into the worst possible response to a nuclear attack.

Worse, major media everywhere have a conflict of interest in honestly reporting on anything (like these research results) that might threaten those who control the money for the media.[35] Everyone thinks they know more than they do,[36] which makes them easily misled by the media they find credible.[37]

Recapitulation

edit

In sum, the worst possible response to a nuclear attack would seem to be a nuclear response.

Existing nuclear weapons policies appear to be supported by propaganda that is effective, because it supports the preferences of those who control the money for the media,[35] and because everyone thinks they know more than they do.[36]

Acknowledgements

edit

Thanks to Owen B. Toon, Alan Robock, and presenters at their irregular webinar series on impact on climate of a nuclear war. Of course, any errors and other deficiencies in this article are solely the responsibility of the author.

See also

edit

References

edit

Notes

edit
  1. Tegmark (2023).
  2. Xia et al. (2022; see esp. their Table 1).
  3. Xia et al. (2022, Table 1) reported "Number of direct fatalities" and "Number of people without food at the end of year 2" out of a total population of 6.7 billion for their simulated year 2010. Two issues with this: First, Xia et al. (2022, Fig. 1) show that the climate impact does not start recovering until year 5 after the nuclear war and has not yet fully recovered 9 years after the war. Thus, few people still alive without food at the end of year 2 will not likely live to year 9. Second, the percentages plotted here are the sums of those two numbers divided by 6.7 billion. The Wikipedia article on World population said the world population in 2010 was 6,985,603,105 -- 7 billion (accessed 2023.08-12). The difference between 6.7 and 7 billion seems so slight that it can be safely ignored, especially given the uncertainty inherent in these simulations and the likelihood that the small populations excluded were probably not substantively different from those included.
  4. Estimates of North Korea's nuclear weapons stockpile vary widely, as summarized in the Wikipedia article on North Korea and weapons of mass destruction, accessed 2023-08-07. The estimate of 30 weapons averaging 17 kt each seems not far from the middle of the estimate cited in that article. That totals 510 kt (0.51 megatons), roughly a third of smallest nuclear war simulated by Xia et al. (2022).
  5. Xia et al. (2022, esp. their Tables 1 and 2).
  6. Wikisource:Joint Statement of the Leaders of the Five Nuclear-Weapon States on Preventing Nuclear War and Avoiding Arms Races. See also Borger (2022). Douthat (2022) discussed the current Ukraine crisis in The New York Times. He concluded that for us (presumably the US and perhaps its NATO allies) "To escalate now against a weaker adversary [Russia], one less likely to ultimately defeat us and more likely to engage in atomic recklessness if cornered, would be a grave and existential folly."
  7. Ronald Reagan; Mikhail Gorbachev (11 December 1987). "Joint statement by Reagan, Gorbachev". The Washington Post. Wikidata Q111845607. ISSN 0190-8286. https://www.washingtonpost.com/archive/politics/1987/12/11/joint-statement-by-reagan-gorbachev/cd990a8d-87a1-4d74-88f8-704f93c80cd3/.  Reagan made that same statement 1984-01-25 in his fourth State of the Union Address.
  8. For a "List of nuclear weapons tests", see the Wikipedia article by that title (accessed 2023-07-06).
  9. Johnston (2001) reported that only 521 of the more than 2,000 nuclear weapons tests were above ground. If 521 explosions of nuclear weapons in deserted places have not generated a substantive impact on human health, it seems unlikely that a nuclear war involving a few thousand explosions of nuclear weapons in deserted areas would be dramatically worse.
  10. The "Number of direct fatalities" in a nuclear war lasting a week ranged from 27 to 360 million in simulations summarized in Xia et al. (2022, Table 1).
  11. Ellsberg (2017, pp. 2-3) includes a graph that the Joint Chiefs Joint Chiefs of Staff produced in the Spring of 1961 to answer President Kennedy's question, "If your plans for a general [nuclear] war are carried out as planned, how many people will be killed in the Soviet Union and China?" This graph was a straight line beginning at 275 million who would die during the initial nuclear exchange with another 8.25 million dying each month for the next six months, totaling 325 million deaths.
  12. Robock et al. (2007); Toon et al. (2019). Of course, a nuclear war could be started accidentally by any nuclear-weapons state, as suggested in the report of an Indian cruise missile that landed 2022-03-10 in Pakistan (Mashal and Masood 2022). See also Xia et al. (2022).
  13. Jägermeyr et al. (2020).
  14. as predicted by Jägermeyr et al. (2020) and others.
  15. Sen (1999, p. 32). Later on p. 178, he stated similarly, "there has never been a famine in a functioning multiparty democracy."
  16. e.g., Woodham-Smith (1962).
  17. Ellsberg (2017, esp. pp. 2-3) noted that 325 million would die in the Soviet Union and China and another couple hundred million in neighboring countries, totalling six hundred million.
  18. Dikötter (2010). See also Halimi (2018), which gives the date as 1957. There is some controversy about this quote; see the Wikipedia article on "Mao Zedong", accessed 2022-03-02.
  19. Coup et al. (2019, p. 8522).
  20. Ellsberg said that 98 or 99 percent of humanity would starve to death if they did not die of something else sooner (Ellsberg et al. 2017). Coup et al. (2019) and Xia et al. (2022) conclude that it won't be quite that bad but will still pretty grim.
  21. Nicholas Kristof (10 March 2004). "A Nuclear 9/11". The New York Times. Wikidata Q111906710. ISSN 0362-4331. https://www.nytimes.com/2004/03/10/opinion/a-nuclear-9-11.html. 
  22. Orde Kittrie (22 May 2007). "Averting Catastrophe: Why the Nuclear Non-proliferation Treaty is Losing its Deterrence Capacity and How to Restore It". University of Michigan Law School. Wikidata Q111906652. https://web.archive.org/web/20091222051825/http://students.law.umich.edu/mjil/article-pdfs/v28n2-kittrie.pdf. 
  23. Accessed 2022-05-08, when it cited Jay Davis (25 March 2008). "After A Nuclear 9/11". The Washington Post. Wikidata Q111905675. ISSN 0190-8286. https://www.washingtonpost.com/wp-dyn/content/article/2008/03/24/AR2008032402291_pf.html. , Brian Michael Jenkins (12 September 2008). "A Nuclear 9/11?". The RAND Blog. Wikidata Q111906145. https://www.rand.org/blog/2008/09/a-nuclear-911.html. 
  24. Turcotte (2022) offered some suggestions. Recommendations more consistent with the analysis here is the Veterans For Peace Nuclear Posture Review (PDF), Veterans for Peace, January 2022, Wikidata Q111141993 They mention the "Treaty on the Prohibition of Nuclear Weapons", supported by the International Campaign to Abolish Nuclear Weapons (ICAN).
  25. Hubbard (2020, pp. 81-82).
  26. Kahneman et al. (2021) report that with some data, a statistical model fit often does better. With lots of data, artificial intelligence systems can do even better. This extends the work of Meehl (1954). Hubbard (2020) and [[w:Superforecasting: The Art and Science of Prediction|Tetlock and Gardner (2015) describe things one might do to improve their intuition.
  27. Narang and Sagan (2022, p. 241).
  28. Narang and Sagan (2022, p. 232).
  29. Narang and Sagan (2022, pp. 231-232).
  30. Narang and Sagan (2022, p. 232).
  31. Lebow et al. (2023). See also Lebow (2020, ch. 4).
  32. e.g., Sagan (1993).
  33. Ellsberg (2017, p. 61).
  34. Perry and Collina (2020). Of course, a nuclear war could be started accidentally by any nuclear-weapons state, as suggested in the report of an Indian cruise missile that landed 2022-03-10 in Pakistan (Mashal and Masood 2022).
  35. 35.0 35.1 McChesney (2004). Cagé (2016). Rolnik et al. (2019). See also "Confirmation bias and conflict".
  36. 36.0 36.1 Kahneman (2011).
  37. Confirmation bias and conflict. See also McChesney (2004), Cagé (2016), and Rolnik et al. (2019).