Michael KreponMassive Retaliation

Massive retaliation is a siren song that appeals to states that cannot afford a nuclear competition but can afford to let an adversary cross the nuclear threshold first. It’s a money-saver, and it sounds persuasive, until the threat of massive retaliation is actually tested — when a nation’s nuclear bluff is called. What national leader would actually respond to the use of a single nuclear weapon, or just a few, with massive retaliation?

Of course, a single thermonuclear weapon targeted on a major city might be considered massive retaliation when compared to the use of a low-yield, tactical nuclear weapon. Great Britain and France are postured to do far worse – one of the consequences of relying on MIRVed missiles aboard submarines — but it’s hard to imagine their bluff being called, because plausible tripwires are so remote.

No nuclear doctrine can be persuasive when the use of nuclear weapons seems incomprehensible. States possessing nuclear weapons are therefore obliged to suspend disbelief and draw up plans for the unthinkable. Planning occurs in a vacuum until another mushroom cloud appears on a battlefield, whether by accident, inadvertence, or design. Only then will doctrine and declaratory policy be tested. But no possible test can be aced by the option of massive retaliation. Massive retaliation is the antithesis of nuclear planning. Yes, I remember that Lawrence Freedman defined all nuclear strategy as an oxymoron, but massive retaliation makes other nuclear employment options seem downright thoughtful.

The best-laid plans tend to go awry in conventional warfare, and we can only imagine how badly the execution of nuclear planning could go awry. Flexible response and graduated nuclear punishment were conceptualized to make greater sense of weapons in bloated arsenals. The problem was that no one could make a convincing case of escalation control in the smoking, irradiated ruin of a nuclear battlefield. The more rungs of graduated response that Herman Kahn conceptualized, the more he became an object of ridicule.

The Samson option is for losers, not for states with important equities, especially states that can afford to compete. Recoiling from the Korean War, the Eisenhower Administration briefly adopted a declaratory policy of massive retaliation as a deterrence booster and a money-saver. To refresh memories, here are the key passages from Secretary of State John Foster Dulles’s famous speech, delivered to an august assembly of the Council on Foreign Relations on January 12, 1954:

We need allies and collective security. Our purpose is to make these relations more effective, less costly. This can be done by placing more reliance on deterrent power and less dependence on local defensive power… What the Eisenhower administration seeks is a similar international security system. We want, for ourselves and the other free nations, a maximum deterrent at a bearable cost… Local defenses must be reinforced by the further deterrent of massive retaliatory power. A potential aggressor must know that he cannot always prescribe battle conditions that suit him…

The way to deter aggression is for the free community to be willing and able to respond vigorously at places and with means of its own choosing… So long as our basic policy concepts were unclear, our military leaders could not be selective in building our military power…

But before military planning could be changed, the President and his advisers, as represented by the National Security Council, had to take some basic policy decisions. This has been done. The basic decision was to depend primarily upon a great capacity to retaliate, instantly, by means and at places of our choosing.

This declaratory policy began to be qualified soon after Dulles delivered his speech. The United States could afford to compete, but couldn’t afford to have just one, world-ending declaration of nuclear deterrence. Nuclear doctrine is supposed to be credible – a tall order under the best of circumstances – and massive retaliation failed this test, at least for the United States.

Which brings us to India. India’s “draft” nuclear doctrine, prepared by an eclectic group of advisors in 1999, stated that “any nuclear attack on India and its forces shall result in punitive retaliation with nuclear weapons to inflict damage unacceptable to the aggressor.” Perfectly reasonable language. Then, in 2003, the Indian Government put its imprimatur on the draft doctrine, highlighting several refinements. One was that “Nuclear retaliation to a first strike will be massive and designed to inflict unacceptable damage.” The word “massive” might have been chosen to stiffen Indian deterrence, but it opened a trapdoor by either narrowing New Delhi’s options or undermining its credibility.

India, perhaps more than any other state possessing nuclear weapons, might actually have its nuclear doctrine put to the test. One possibility is if, in a limited war, a weapon detonates when struck by conventional means because it lacks adequate safety mechanisms. Another is a breakdown of command and control in the fog of war. A third is if Pakistani military authorities use a detonation to demand stoppage of an Indian advance.

None of these scenarios might come to pass. Previous Indian governments have demonstrated great restraint after suffering attacks originating in Pakistan, preferring to go about the business of economic growth rather than to engage in retaliatory military strikes. The Indian Army’s “Cold Start”-like military plans have many weaknesses and might be left on the drawing boards. And Pakistani military and intelligence authorities might prove capable of preventing the usual suspects from carrying out new explosions on Indian soil during a very hawkish Indian government. These suppositions are conceivable. They are also about as reliable as declaratory nuclear doctrine.

The peculiarity here is that India, unlike the United States facing the Soviet Union, enjoys conventional military advantages over Pakistan – advantages that will grow over time. Pakistan’s nuclear doctrine threatens first use because of India’s conventional edge. This is understandable. But why has New Delhi adopted a posture of massive retaliation? Is it to save money or sound tough, like the Eisenhower Administration? How credible is this posture, and will New Delhi revamp it? And if New Delhi does vocalize the possibility of limited nuclear options, will this be good or bad for deterrence stability and escalation control?


  1. P (History)

    There’s another difference between the second doctrine (“Nuclear retaliation to a first strike will be massive and designed to inflict unacceptable damage”) and the first one (“any nuclear attack on India and its forces shall result in punitive retaliation with nuclear weapons to inflict damage unacceptable to the aggressor”) : it’s not that clear what would constitute a “first strike”. In particular, it doesn’t seems evident to me that the use of a nuclear weapon against Indian forces located outside of pre-war Indian territory would qualify as a “strike”.

    In fact, I would argue that it would be very difficult to make both the Indian public and the international community accept the need or the legitimacy of *any* nuclear retaliation in such a case. This would imply that neither an accidental detonation nor a deliberate strike to stop an Indian advance would actually be possible triggers for this doctrine.

    Also, I believe that this doctrine has some stabilizing effects in case of a breakdown in the chain of command or against unruly generals. That is, if India threatened graduated retaliation, a Pakistani commander could be tempted to launch a nuclear strike at New Delhi, so that India destroys Islamabad (but nothing else, in particular not his army or the rest of the country), after which he could seize the now-vacant power and install his own dictatorship over the rest of the country (without looking as criminal as if he had simply destroyed Islamabad himself, even tough he would actually be far worse). The threat of massive retaliation makes such a plan non-viable.

    • Jonah Speaks (History)

      P, Thank you (and Carey below) for bringing up the distinction between a first use and a first strike – where a first strike requires a massive (not picayune) use of nuclear weapons. If the 2003 Indian policy refers to a massive use of nuclear weapons, then a policy of massive response sounds more reasonable. Does anyone know if the Indian policy statement makes this distinction explicitly?

      The scenario of the unruly general bombing New Delhi, in the hope that India destroys Islamabad and nothing else, is interesting. Give India a week or two to think about it, massive nuclear retaliation against Pakistan is not rational, if the result would be nuclear devastation of India. The power-hungry general knows it is not rational, so he may gamble that India’s threat is not credible. Hence, the threat of massive retaliation may not deter his devious plan.

      Now consider more limited options. To be sure, bombing Islamabad would be the first thing people think of, but it is not the only limited option. For example, India could bomb the Pakistani army, including suspected unruly generals. India knows that Pakistan has a number of unruly, power-hungry generals. India may prefer to destroy the generals and negotiate with the civilian government.

  2. Arch Roberts (History)
    • krepon (History)

      You’ve just given me the next year-end ACW contest: find the second-best short internet satire on the Bomb. Can anything top this?

    • Ricki (History)
  3. Jonah Speaks (History)

    If one can stomach the genocidal implications, destroying the Soviet Union in 1954 made apparent sense, because the Soviet Union had no significant nuclear retaliatory capacity. The same policy pursued in 1964 would be mutually genocidal, and therefore both mad and MAD. However, we ignored the madness and pretended we were sane because the mad threats ostensibly provided “deterrence.”

    India has no good reason to threaten massive nuclear retaliation against even limited nuclear use by Pakistan. Similarly, Pakistan has no good reason to threaten limited nuclear use against conventional war by India. Unfortunately, both sides seem to think they are being rational.

    These attitudes could lead to a nuclear war. By luck, we escaped nuclear war during the Cold War. India and Pakistan may not be so lucky. If they are unlucky, millions will die directly in India and Pakistan.

    Additionally, a nuclear war between India and Pakistan would kick enough dust into the atmosphere to dim the sunlight and lead to global cooling for about ten years (similar to “nuclear winter”). Worldwide, an additional one or two billion people will die indirectly from starvation, because the dimmed sunlight will cut short the growing seasons and reduce crop output. These additional starvation deaths will be people living in India and Pakistan, as well as in China and third-world nations and perhaps include some first-world nations.

    • j_kies (History)

      While experimental testing of the hypothesis seems unadvisable in the extreme, the ‘Nuclear Winter’ scenarios have serious issues as to threshold and persistence as VE6 and VE7 volcanic eruptions eject vastly more mass of small particulate matter into the stratosphere.

      If a threshold exchange is postulated to result in a 10 year global cooling, where was the cold excursion in the 1990s from the VE6 Pinatubo eruption? The 1815 VE7 Tambora eruption included the ‘year without a summer’ but the cooling did not persist for the decade.

    • Jonah Speaks (History)

      Since I am an economist, not a climate scientist, I can only search for what others have said on this. Wikipedia gives a general overview, but does not answer your specific question: http://en.wikipedia.org/wiki/Nuclear_winter From this and other readings, I conclude the science appears sound.

      Your specific question is addressed here: http://www.wunderground.com/resources/climate/nuke.asp?MR=1 In their summary of a recent (December 2008) study:

      “The scientists used a sophisticated atmospheric/oceanic climate model that had a good track record simulating the cooling effects of past major volcanic eruptions, such as the Philippines’ Mt. Pinatubo in 1991. The scientists injected five terragrams (Tg) of soot particles into the model atmosphere over Pakistan in May of 2006.”

      “The intense heat generated by the burning cities in the models’ simulations lofted black smoke high into the stratosphere, where there is no rain to rain out the particles. The black smoke absorbed far more solar radiation than the brighter sulfuric acid aerosol particles emitted by volcanic eruptions. This caused the smoke to heat the surrounding stratospheric air by 30°C, resulting in stronger upward motion of the smoke particles higher into the stratosphere. As a result, the smoke stayed at significant levels for over a decade (by contrast, highly reflective volcanic aerosol particles do not absorb solar radiation and create such circulations, and only stay in the stratosphere 1-2 years). The black soot blocked sunlight, resulting in global cooling of over 1.2°C (2.2°F) at the surface for two years, and 0.5°C (0.9°F) for more than a decade (Figures 1 and 2). Precipitation fell up to 9% globally, and was reduced by 40% in the Asian monsoon regions.”

    • j_kies (History)

      Nuclear winter appears credible; I believe unwise human actions can replicate the stratospheric aerosol insertion masses of VE6 / VE7 class eruptions. I question assuming that black carbon from firestorms would persist substantially longer in the stratosphere than volcanic aerosols. Large scale rockets have deposited black carbon plumes which might be analyzed for agglomeration and aerosol removal mechanisms. We have very significant firestorm examples with persistence e.g. Hamburg, Dresden, and Tokyo; but I am unaware of studies of smoke transport efficiency to the Stratosphere (and the referenced studies show 1944/45 as a warm peak in the plots). Hiroshima and Nagasaki mushroom clouds clearly penetrated into the stratosphere; however the mass of the firestorm smoke transported to the stratosphere after the initial fireball convective cell is an open question.

      Lest the credibility of the message (to decision makers) be damaged by dire projections where the uncertainties in the data do not uniquely support the most extreme messages; I gently suggest that sensitivity excursions be exercised and the message be tailored to the most likely / demonstrated result. If the smoke generation and transport efficiency to the stratosphere from a limited nuclear exchange approach the optical effects of the 1815 Tambora eruption; a repeat of the 1816 world-wide effects are sufficiently terrible to inform choices without invoking unprecedented events.

    • Jonah Speaks (History)

      There are/were two waves of research on nuclear winter. The first wave occurred in the 1980s, occupied the attention of significant numbers of American and Russian scientists, and ended with the Cold War. The second wave started several years ago (maybe 2005+), has involved fewer scientists, and maybe still continues. The second wave uses faster, more powerful computers and better software. The software is standard for global warming models, but applied to nuclear cooling/winter.

      In the global warming models, there are at least six greenhouse gases of concern. The most famous is CO2, but water, methane, and others also have greenhouse impacts. Water has a residency of about 9 days, CO2 has a variable atmospheric lifetime of 35-90 years, N2O a mean lifetime of 114 years, and so forth. Hence, I cannot dismiss a residency time for black carbon in excess of two years as outlandish. Moreover, the scientists have pointed out plausible reasons for why the residency time is significantly longer than for SO2.

      Obviously, it will make a difference whether the residency time is two years or ten years. If the earth is subjected to sudden cooling from a regional nuclear war, I expect most of the 1-2 billion starvation deaths would occur in the first two years, even if the effect persists for ten years. For planning purposes, it would be useful to know how long the global cooling would last, in case international efforts can prevent some of those deaths.

      The fact that only a small number of scientists are involved in the second wave of studies does concern me. I wish more independent studies would be conducted, including analyses of possible uncertainties in what we know and don’t know about the nuclear cooling phenomenon.

  4. Carey Sublette (History)

    “But why has New Delhi adopted a posture of massive retaliation? ”

    The same reason any state does: to minimize the likelihood of suffering nuclear attack in the first place.

    And consider the statement of policy cited:
    “Nuclear retaliation to a first strike will be massive and designed to inflict unacceptable damage.”

    What is the precise definition of a “first strike”?

    There isn’t one of course, so India is free to interpret what that term might mean under the circumstances of a nuclear explosion occurring.

    There is no way an accidental nuclear explosion due to safety mechanism failure would be viewed that way by anyone. Neither would any demonstration shot. A single rogue weapon use can also easily be dismissed as “not a first strike”, unless India wanted a pretext for retaliation.

    A planned coordinated multiple weapon attack – now that it is a first strike without question, and that is what the policy is intended to minimize (but Pakistan would have to assume India would retaliate massively even without such a declaration).

    You hit the nail on the head with your passing aside:
    “They are also about as reliable as declaratory nuclear doctrine.”

    Exactly. These are words for peacetime political consumption only – once nukes explode they are meaningless.

  5. Rabs (History)

    MK..considering that you posted this on April 01st, India’s massive retaliation doctrine is the best practical joke ever (pun intended).

  6. RAJ47 (History)

    The trapdoor was opened in 2003, why talk about it now after more than a decade?
    Well, it is because of ‘conceivable supposition’ that ‘a very hawkish Indian government’ is likely to come to power soon.
    Anyway, NFU is the backbone of every Indian nuclear policy and it shall remain so, always & every time, forever.
    Indian leaders (political as well as military) are sane & matured in every sense. They would certainly uphold Indian national interests, may be more assertively this time around.
    Another peculiarity not discussed by you is India – PRC scenario. PRC enjoys conventional advantage over India. Although both India & PRC have NFU in place, India can’t lower its guard.
    First strike will certainly include first use.
    P:” In particular, it doesn’t seems evident to me that the use of a nuclear weapon against Indian forces located outside of pre-war Indian territory would qualify as a “strike”.”
    A nuke on Indian forces say in POK would qualify as a first strike/use.
    A nuke on Indian forces South of Fort Abbas would also qualify a first strike/use.
    Jonah Speaks: “India has no good reason to threaten massive nuclear retaliation against even limited nuclear use by Pakistan.”
    That is the most juvenile statement I read in a long time.

    • Jonah Speaks (History)

      If P and Carey argue that first strike requires massive nuclear use, while RAJ47 argues that first strike means any nuclear use against India or Indian troops, it appears that the Indian policy statement is ambiguous. Hence, I will expand on my question above: Does anyone know if the Indian policy statement on massive retaliation explicitly defines “first strike” as any use of nuclear weapons or only massive use of nuclear weapons? Does India otherwise make clear India’s planned (or declared) nuclear response to such contingencies?

      RAJ47, If statements such as mine are rarely or never heard in the Indian press, then I surmise that India does not yet have a healthy debate about nuclear weapons policy. Unlike the Cold War, where capitalism and communism appeared intent on overthrowing each other, Pakistan and India only have a territorial dispute in Kashmir. For what reason is Pakistan threatening tactical nukes, and India threatening a massive nuclear retaliatory response? From my perspective, this is a nuclear war waiting to happen, for causes too trivial to justify the risk.

  7. Bradley Laing (History)


    New Delhi, April 2: Prime Minister Manmohan Singh today proposed a global convention on ‘no-first use’ of nuclear weapons as it could lead to elimination of atomic arsenal. “If all states possessing nuclear weapons recognise that if this is so (nuclear weapons are only for deterrence) and are prepared to declare it, we can quickly move to the establishment of global no-first use norm. In many ways this can open the ways to gradual reduction and finally elimination through a nuclear weapon convention. Such a convention would require necessary verification. It would also require political measures to ensure the stability is maintained at the level as nuclear arsenal approaches zero,” he said.

    Read more at: http://news.oneindia.in/india/pm-manmohan-singh-proposes-no-first-use-nuclear-weapons-1423055.html

    • Arch Roberts (History)

      In the early eighties, no first use made no sense because of the Cold War. Now it does, even if it’s only declaratory policy. Israel’s policy is essentially no first use, as is China’s, as Jeffrey has written, and I can think of no reason why minimal deterrence in terms of numbers to be defined, coupled with no first use assurances, cannot be a stabilizing thing.

    • Jonah Speaks (History)


      The Bharatiya Janata Party (BJP), widely tipped to form the next government, pledged on Monday to revise India’s nuclear doctrine, whose central principle is that New Delhi would not be first to use atomic weapons in a conflict.

      Unveiling its election manifesto, the party gave no details, but sources involved in drafting the document said the “no-first-use” policy introduced after India conducted a series of nuclear tests in 1998 would be reconsidered.

  8. Bradley Laing (History)

    Pentagon leaders expect to soon give President Obama a plan for specific U.S. nuclear cuts to bring the arsenal in line with arms control caps.


  9. RAJ47 (History)

    Is truth so unpalatable?
    Why does the US government & its intelligentsia not want to forsake its threat of first use of nuclear weapons?
    Why does the US government & its intelligentsia feel arms control only mean arm twisting of all countries possessing or likely to possess nukes?

  10. Alexis (History)

    You write “Great Britain and France are postured to do far worse – one of the consequences of relying on MIRVed missiles aboard submarines”, which calls for several precisions:

    – The desirability to have more options than just massive retaliations is the rationale why France also maintains about 40 of the supersonic air-to-ground missile ASMP-A, whose role is to provide more flexible ways to retaliate than just to send out the city-busters. In French nuclear doctrine parlance this is known as the “ultimate warning” step, the first crossing of nuclear threshold which also is the ultimate opportunity for the aggressor to reconsider his action and avoid the most devastative consequences

    – Provided one or several of a British or French SSBN’ 16 ballistic missiles are fitted with a single warhead, a limited retaliation can also be effected by a SSBN. This of course would not be advisable against a superpower having the means to destroy a SSBN whose position has been revealed, however against lesser adversaries it would be a realistic option

  11. Russell Wellen (History)

    “One possibility is if, in a limited war, a weapon detonates when struck by conventional means because it lacks adequate safety mechanisms.”

    Thanks for pointing that out. Everyone forgets about it.

  12. Bradley Laing (History)


    “By Kris Osborn…

    The technical details, known as ship specification documents, will set the stage for engineering work and eventually construction of ships in the so-called Ohio Replacement Program being developed by General Dynamics Corp.’s Electric Boat unit.

    Consisting of three volumes, each with hundreds of pages, the documents detail the configuration, design and technical requirements for the next-generation boat, the first of which is slated to begin construction in 2021 and enter the fleet in 2031, Navy and company officials said.”

  13. Bradley Laing (History)


    WASHINGTON — The U.S. will keep its current force of 450 land-based nuclear missiles but remove 50 from their launch silos as part of a plan to bring the U.S. into compliance with a 2011 U.S.-Russia arms control treaty, the Pentagon said Tuesday.

    The resulting launch-ready total of 400 Minuteman 3 intercontinental ballistic missiles would be the lowest deployed ICBM total since the early 1960s.

    The decisions come after a strong push by members of Congress from the states that host missile bases — North Dakota, Wyoming and Montana — to not eliminate any of the silos from which the missiles would be launched.

  14. Bradley Laing (History)

    As part of the technical improvements to the missile, the Navy is upgrading what’s called the Mk-4 re-entry body, the part of the missile that houses a thermonuclear warhead. The life extension for the Mk-4 re-entry body includes efforts to replace components including the firing circuit, Benedict said.
    The Navy is also working with the Air Force on refurbishing the Mk-5 re-entry body which will be ready by 2019, Benedict said.

    Read more: http://defensetech.org/2014/04/11/navy-extends-trident-ii-d5-nuclear-missile-service-life/#ixzz2ye3oW44W

    The Trident II D5 also arms the United Kingdom’s Vanguard ballistic missile submarine. In fact, the U.S. and UK are collaboratively working on a common missile compartment for their next generation SSBNs, or ballistic missile submarines.

    Read more: http://defensetech.org/2014/04/11/navy-extends-trident-ii-d5-nuclear-missile-service-life/#ixzz2ye4196Z1

  15. shanks (History)


    > Does anyone know if the Indian policy statement on massive retaliation explicitly defines “first strike” as any use of nuclear weapons or only massive use of nuclear weapons?


    basically, even a small tactical one is enough for India

    “Two, whatever sophistry Pakistan may indulge in to justify its augmented arsenal and threatened recourse to tactical nuclear weapons, for India, the label on the weapon, tactical or strategic, is irrelevant since the use of either would constitute a nuclear attack against India. In terms of India’s stated nuclear doctrine, this would invite a massive retaliatory strike. “

    • Jonah Speaks (History)

      Shanks, Thank you for the reference. Since Shyam Saran is Chairman of the Indian National Security Advisory Board (NASB), his interpretation is close to being authoritative, albeit unofficial.

      In a more recent (August 13, 1013) speech, Saran offers this justification for massive retaliation: “A limited nuclear war is a contradiction in terms. Any nuclear exchange, once initiated, would swiftly and inexorably escalate to the strategic level.” http://www.irgamag.com/component/k2/item/2400

      Saran’s justification rests on a false premise. It cannot be said definitively that limited nuclear war necessarily leads to unlimited nuclear exchange between the combatants. There is a high risk that it could end this badly, but there is also a good chance that a series of limited nuclear exchanges may end, either with one side deciding not to proceed with further nuclear attacks, or with both sides coming to a mutual nuclear ceasefire.

      Because there is a high (but not inevitable) risk that a limited nuclear attack may end in unlimited nuclear war, I agree with Saran’s further statement: “Pakistan would be prudent not to assume otherwise as it sometimes appears to do, most recently by developing and perhaps deploying theatre nuclear weapons.” This high risk for Pakistan would remain, even if India would announce a policy of flexible, limited, or proportionate response to a limited nuclear attack.