Michael KreponEscalating to De-escalate

Quote of the week:

“In light of the certain prospect of retaliation there has been literally no chance at all that any sane political authority, in either the United States or the Soviet Union, would consciously choose to start a nuclear war. This proposition is true for the past, the present, and the foreseeable future…”

“There is an enormous gulf between what political leaders really think about nuclear weapons and what is assumed in complex calculations of relative ‘advantage’ in simulated strategic warfare…”

“In the real world of political leaders…a decision that would bring even one hydrogen bomb on one city of one’s own country would be recognized in advance as a catastrophic disaster; ten bombs on ten cities would be a disaster beyond history; and a hundred bombs on a hundred cities are unthinkable.”
— McGeorge Bundy, “To Cap the Volcano,” Foreign Affairs, October 1969

Is Bundy still right? Do political leaders in nuclear-armed states want no part of nuclear war-fighting plans? We know that as long as nuclear weapons exist, these plans will exist. They will be briefed to national leaders who prefer to keep them in locked safes. One extreme case: I am told on responsible authority that when Indian Prime Minister Narasimha Rao was first handed these plans, he demanded that only one copy be made, that it be placed in a locked safe and that no-one else see it. He was later persuaded otherwise.

National leaders that have murdered millions, like Mao and Stalin, declined to employ nuclear weapons on battlefields. Will Donald Trump break this crucial norm of national leadership? His public statements about the use of nuclear weapons have given us all pause, and have prompted useful ideas constraining the first use of nuclear weapons by the United States, including a three person rule to authorize first use.

Has Bundy’s dichotomy between the mind of the nuclear enclave and the political mind eroded? Scott Sagan’s recent public opinion research suggests changes are afoot in thinking about the Bomb’s use. Is this also the case among new national leaders? If so, we are in even more trouble than we think.

Bundy, who sat with President Kennedy during deliberations on the Cuban missile crisis and heard proposals for military action against targets where nuclear weapons were wrongly presumed to be absent, was fortunate to work for a president whose utterly risky personal choices did not carry over to strategic calculations. Bundy came to the conclusion that nuclear weapon strategists were living in an “unreal world.” He wrote that, “In sane politics,” no level of superiority and no perceived tactical advantage would make first use “anything but an act of utter folly.”

Speaking of utter folly, have the Kremlin’s war planners embraced the gonzo nuclear war-fighting concept of “escalating-to-de-escalate” by means of tactical nuclear weapons? Granted, the Kremlin opposes the post-Cold War geo-political status quo and is engaged in nefarious activities to disrupt democracies, but there is scant evidence that Russia is a revisionist nuclear power. Instead, it is looking to increase its leverage by muscling its neighbors — including by circumventing the INF Treaty. As was the case in the late 1970s and early 1980s when it last attempted this gambit, it is likely to backfire.

Russia has also recapitalized its strategic forces – as the United States is now doing — but within the framework of limits agreed upon in the 2010 New Strategic Arms Reduction Treaty. Moscow has proposed extending these constraints, suggesting that Putin recognizes the benefits of the strategic nuclear status quo with the United States.

It is not surprising that the mind of the nuclear enclave would dwell on the first use of low-yield nuclear weapons to “escalate-to-de-escalate.” But how exactly would the crossing of the nuclear threshold in this way – by definition a highly freighted symbolic act – tip the scales toward de-escalation rather than escalation? Would a political leader accompany first use with signals that subsequent use would not follow? How reassuring would that be? If, instead, first use would be accompanied by the threat of subsequent use to reinforce deterrence, how likely would this serve de-escalatory purposes? Can the aggrieved state accept restraint when it can retaliate in kind and up the ante? Whatever the scenario, the notion of crossing the nuclear threshold first as a de-escalatory step sounds hallucinatory.

It is hard for the mind of one insular nuclear enclave to read the mind of another. This makes mirror imaging a real problem. During the Cold War, U.S. nuclear strategists devised graduated targeting options on the assumption that the Kremlin was similarly inclined. After gaining access to Soviet nuclear war-fighting plans after the Cold War ended, U.S. nuclear strategists discovered that they were wrong about their assumptions. In fact, the Soviet military, if given the green light, didn’t believe in escalation control.

Fast forward to speculative assessments that Russia’s targeting plans have adopted a war-fighting posture to “escalate-to-de-escalate.” On what basis does this hypothesis exist? True, Russia possesses more “tactical” nuclear weapons than the United States, but this is not dispositive evidence. The Soviet Union also possessed more low-yield weapons than the United States, and Soviet doctrine was not postured to escalate to de-escalate; it was to escalate to win.

Do Russian nuclear doctrine and military exercises indicate an “escalate-to-de-escalate” posture? The highly respected French strategic analyst Bruno Tertrais has reviewed available evidence and has concluded otherwise. His essay “Does Russia really include limited nuclear strikes in its large-scale military exercises?” in the Survival Editors’ Blog deserves a wider audience.

Bruno’s assessment questions the hypothesis that “Russia’s doctrine of ‘escalate-to-de-escalate’ is real, as it is not reflected in large-scale military exercises, official doctrine and public statements:

“All the elements of this narrative, however, rely on weak evidence – and there is strong evidence to counter most of them. This applies to the role of nuclear weapons in Russian military exercises.”

“Exercises are important in understanding Russian nuclear posture, because, as the saying goes, Moscow trains as it fights and fights as it trains. So what do large-scale ones such as Zapad(Western front) and Vostok (Eastern front) tell us?”

“What they tell us is that the last time a Zapad included nuclear use was almost 20 years ago, in 1999 – Russia was explicit about it – and that no known large-scale theatre military exercise has included nuclear-weapons use for at least a decade. This is unsurprising: Russia now ‘wins’ – or at least ‘resists’ – without nuclear weapons.”

Another way for the Kremlin to “strengthen” deterrence by warning that it is inclined to “escalate to de-escalate” is through official pronouncements. While reserving the right of first use, Russian doctrine makes no mention of an “escalate to de-escalate” posture. Here is a relevant passage from the 2014 Russian Military Doctrine:

“The Russian Federation shall reserve the right to use nuclear weapons in response to the use of nuclear and other types of weapons of mass destruction against it and/or its allies, as well as in the event of aggression against the Russian Federation with the use of conventional weapons, when under threat the very existence of the state.”

Here’s what the 2010 Russian Military Doctrine has to say:

“The Russian Federation reserves the right to utilize nuclear weapons in response to the utilization of nuclear and other types of weapons of mass destruction against it and/or its allies, and also in the event of aggression against the Russian Federation involving the use of conventional weapons when the very existence of the state is under threat.”

Perhaps doctrine has been updated. If so, it would make sense for Russian officials to reinforce this deterrent message with public statements. They have not done so. Instead, they have denied an interest in escalating to de-escalate. Those who assert that Russia now embraces this posture – and the need for new low-yield U.S. options to counter it — are obliged to provide supportive evidence for their claims. I’m skeptical. The analysis offered by Olga Oliker seems more on the mark:

“[T]he evidence that Russia’s nuclear strategy is one of “de-escalation,” or that it has lowered its threshold for nuclear use, is far from convincing. Rather, Russia’s statements and behavior indicate more a desire to leverage its status as a nuclear power—less a lowering of the threshold than a reminder that escalation is possible, and that Russia must therefore be taken seriously.”

Dear Readers: If there is hard evidence to substantiate the Russian “escalate to de-escalate” hypothesis — something beyond un-sourced commentary in Polish media and off-the-cuff remarks by U.S. officials — to substantiate this folly, please provide it here or elsewhere. Or has the “escalate-to-de-escalate” hypothesis been concocted mostly out of thin air to support new low-yield nuclear options long cherished by the U.S. nuclear enclave?


  1. Davey (History)


    “National leaders that have murdered millions, like Mao and Stalin, declined to employ nuclear weapons on battlefields. Will Donald Trump break this crucial norm of national leadership?”

    This article is about Russian doctrine, specifically the Russian “escalate to de-escalate” hypothesis. It is worthy of discussion, since it may inform our response to Russian nuclear weapon deployment. Yet somehow, you managed to find a reason to compare Donald Trump to Mao and Stalin. Seriously?

    This is supposed to be a world-class forum. Please try to keep your personal dislike for the President out of an academic discussion. It reflects poorly on you.

    • Michael Krepon (History)

      Fair point, Davey.
      What I wrote was far too compressed. What I meant to say is that predicting the behavior of national leaders when it comes to nuclear weapons use is very hard. The ones with blood on their hands can steer clear of nuclear weapons’ use, and the US leaders who say very troubling things (think of Reagan before Trump) can surprise us.
      Thanks for your comment —

    • Phoenix Woman (History)

      Davey, do you honestly think Donald John Trump is anything other than a weak reed who does whatever the last person he’s talked to has told him to do – or whatever he thinks will win him applause from people he respects?

      Look at how quick he was to call up Chuck Schumer for a meeting on DACA when he saw the first batch of poll numbers that morning showing he’d get blamed for a government shutdown. Then see how quick he reneged on his promises to Schumer once the rabidly anti-immigrant John Kelly got hold of him after the meeting was over.

      That’s the man entrusted with the football.

  2. Jonah Speaks (History)

    I believe Bundy was wrong, both then and now. During the Cuban missile crisis, Bundy thought the risk of nuclear war was only 1%, whereas Kennedy thought it was between 33% and 50%. Despite this huge risk of utter catastrophe, Kennedy was unwilling to back down on his demand for immediate removal of the nuclear missiles.

    Was this a risk worth taking? Almost certainly not, as it would have been better to allow the missiles to remain in Cuba and to deter their launch, than to risk a 40% chance of immediate nuclear war. Yet Kennedy (and Khrushchev) undertook that highly unreasonable risk for 13 days.

    In my view, “escalate to de-escalate” is simply a rather colorful description of Russian willingness to contemplate nuclear first use in the unlikely event that a conventional attack from NATO or China threatens “the very existence of the [Russian] state.”

    • Michael Krepon (History)

      There is already an existing nuclear posture to escalate to de-escalate: Pakistan has adopted it to deal with India’s Cold Start conventional doctrine.
      For deterrence purposes, Pakistani leaders have been open and clear about this. (Whether they would actually employ tactical nuclear weapons first is another matter, given the risks involved.)

  3. Akshat Upadhyay (History)

    You are right with respect to Pakistan as far as their alleged Full Spectrum Deterrence is considered. India has also upped the ante by tweaking its nuclear doctrine. It’s now massive retaliation instead of credible.

  4. Trevor30024 (History)

    Have you discussed this with Pavel Podvig? I’d be really interested to hear his input.

    • Michael Krepon (History)

      Will ask.

  5. Bruno Tertrais (History)

    Thanks, Michael. This debate should not go away, given that (my friends) Frank M and Greg W are doubling down publicly (articles, tweets). I have a longer piece coming in Survival on this very issue;

  6. Sultan (History)


    Agree with most of the arguments in the main article, except the comparison with Pakistan in the comments section. Russia enjoys conventional advantage over its imemdiate neighbors and the threat of ‘escalate to de-escalate’ (if at all) could be to raise the costs for the US involvement. Pakistan, on the other hand, is at relative conventional disadvantage, and therefore, has to rely on its TNWs to deter Cold Start. At best one could argue it as ‘escalate to deter’ and not ‘escalate to de-escalate’ – former is ‘war prevention’ and the latter ‘nuclear war fighting’.

    • Michael Krepon (History)

      If deterrence fails after first use, then what?
      Best wishes,

    • Sultan (History)


      It depends upon the credibility of deterrence, but yes it can fail, if the adversary is continuously looking for war fighting options. Responsibility lies more with the aggressor than the defender.
      As Quinlan argued: ‘the role of NWs is to prevent all wars and not just the nuclear war’, and ‘deterrence works by the very fact that these weapons remain useable’ – unfortunate but true.


    • Rabia Akhtar (History)

      @MK @Adil Thought provoking piece, Michael. But I would also take issue with categorizing Pakistan’s ‘threat’ to use tnws as ‘escalate to de-escalate’. At best it can be categorised as threatening escalation to deter the adversary from pushing deterrence breakdown thus ‘escalate to deter’. deter what exactly is debatable. At the face of it, deterring Indian retaliation that will lead to deterrence breakdown. To your question, the assumption is that deterrence would have broken down right before any use of nukes by Pakistan anyway..not after the first use.

    • Jonah Speaks (History)

      The defenders of Pakistani nuclear policy need to consider the high-risk Kennedy-Khrushchev confrontation quite seriously. Maybe it is Pakistan’s sole intention that deterrence “shall work” (and if it fails it must be India’s “fault”) but unforeseen circumstances have a way of undoing the best-laid plans.

      If Pakistani policy were like the Russian policy, Pakistan would consider nuclear first use only against threats to “the very existence of the [Pakistani] state.” This is a high threshold. It is highly unlikely that any military action by India would do more than punish Pakistan for some future act of terrorism – hence, not an existential threat and not worthy of a nuclear response.

  7. Nikolai Sokov (History)

    Here is the source for deescalation discourse: a collection of pieces I wrote in 2000-04: http://www.nti.org/analysis/articles/russias-nuclear-doctrine/. I still stand by it. Name is inconsequential (originated in a Voyennaya Mysl article in 1999). From day 1, it was a deterrence strategy to prevent exercise of US/NATO conventional superiority in a Kosovo-style conflict. The doctrine remained unchanged until 2014 (except for enhancing threshold in 2010). Always pitched as a stop-gap strategy until conventional forces are modernized. In 2014, they introduced conventional deterrence, so now it’s different (to an extent) and in any event threshold is higher.

  8. JStreet (History)

    I would also suggest that you check out the article by Dmitry Adamsky entitled “If War Comes Tomorrow: Russian Thinking About ‘Regional Nuclear Deterrence’ ” (doi: https://doi.org/10.1080/13518046.2014.874852 ). It is a useful piece that explores some of the Russian internal posture and policy debates and how they connect with an asymmetric escalation posture (which, though the typology was specifically framed to encompass regional nuclear powers, fits Russian signalling and technical capacity).