Michael KreponThe Stability-Instability Paradox

Two questions were paramount after the destructive power of atomic weapons was revealed. Some, depending on temperament and instinct, asked whether the capability to make such weapons might possibly be placed under international control. Others, assuming failure in this pursuit, asked how the United States could most effectively prevent further battlefield use of nuclear weapons. Lesser wars facilitated by the Bomb did not figure very much in these early deliberations. As Bernard Brodie wrote in The Absolute Weapon (1946), “One does not shoot rabbits with elephant guns, especially if there are elephants available.”

When H-bombs entered the equation, some strategic analysts began to raise the question of whether scaling up nuclear deterrence might actually increase freedom of action at lower levels of violence. For example, in 1954, B.H. Liddell Hart surmised that, “To the extent that the H-bomb reduces the likelihood of full-scale war, it increases the possibility of limited war pursued by widespread local aggression.” U.S. plans and policies for massive retaliation quickly foundered on this conundrum.

Many academics who subsequently waded into the Big Muddy of deterrence theory arrived at much the same conclusion. Glenn Snyder’s book, Deterrence and Defense, published in 1961, characterized this dilemma as follows:

The Soviets probably feel, considering the massive retaliation threat alone, that there is a range of minor ventures which they can undertake with impunity, despite the objective existence of some probability of retaliation.

Snyder returned to this theme in a chapter he wrote for Paul Seabury’s edited volume, Balance of Power (1965), noting that,

The point is often made in the strategic literature that the greater the stability of the ‘strategic’ balance of terror, the lower the stability of the overall balance at lower levels of violence.

Just about everyone who writes on this paradox, myself included, pays homage to Snyder’s essay in the Seabury book. I belatedly went back to the original, only to discover that he actually hedged his bets by adding, “…but one could argue precisely the opposite.”

Robert Jervis offered a more generalized and yet succinct formula for this paradox in The Illogic of Nuclear Strategy (1984): “To the extent that the military balance is stable at the level of all-out nuclear war, it will become less stable at lower levels of violence.” In actuality, Washington and Moscow perceived greater dangers as their nuclear competition accelerated, so the stability part of this equation turned out to be deeply suspect. But Jervis’ larger point remained valid: Adversaries possessing nuclear weapons would exercise caution to avoid major wars and any crossing of the nuclear threshold. At the same time, their “insurance policy” of nuclear retaliation provided ample leeway to engage in crisis-provoking behavior, proxy wars, and mischief making.

This construct eventually became known in the trade as the stability-instability paradox. I’m not sure who first coined this phrase or when it became common usage. If anyone knows, please send a comment.

The stability-instability paradox is most harrowing at the onset of a nuclear competition for many reasons. Troubled relations between adversaries get worse when nuclear weapons are added to their disagreements. Stability is especially hard to achieve early on, when fear and ignorance are most pronounced because adversarial moves magnify security concerns when monitoring capabilities are rudimentary, at best. In addition, the jockeying to achieve advantage – or to avoid disadvantage – is greatest early on, before tacit rules of a nuclear-tinged competition are clarified.

The U.S.-Soviet experience – admittedly, an extreme and, thankfully, lonely case — suggests that big crises are most likely to occur in the first fifteen years of a nuclear competition. Misperception and spikes in nuclear danger can still occur later on (circa 1983), but the worst passages are typically front-loaded (e.g. Berlin, Cuba, and Korea) after the Bomb’s appearance. The U.S.-Soviet case also suggests that nuclear-armed adversaries can make it through very rough patches, especially if they tacitly agree not to play with fire in each other’s back yard.

The stability-instability paradox has now hit the road and traveled to South Asia. Initially, some very distinguished observers from the region, led by K. Subrahmanyam, K. Sundarji, and Abdul Sattar, believed that going public with nuclear capabilities would serve as a stabilizing factor. The 1999 Kargil War, the 2001-2 “Twin Peaks” crisis sparked by an attack on the Indian parliament building, and the 2008 Mumbai attacks suggest otherwise. A few western analysts, including Kenneth Waltz, Sumit Ganguly, and Devin Hagerty, have argued that, because these events did not cascade into full-blown wars or nuclear exchanges, deterrence optimism is in order. Perhaps, over time, this will be the case. But Cold War conceptualizers of the stability-instability paradox never made the acquaintance of the Lashkar-i-Taiba. Until Pakistan’s security managers tacitly accept the ground rule about not playing with fire, deterrence optimists for South Asia will remain in the minority.


  1. FSB (History)

    ‘In 1954, B.H. Liddell Hart surmised that, “To the extent that the H-bomb reduces the likelihood of full-scale war, it increases the possibility of limited war pursued by widespread local aggression.” U.S. plans and policies for massive retaliation quickly foundered on this conundrum. ‘

    This is the dirty truth.

    And I mentioned this in two of your earlier posts:


    “The threat of nuclear retaliation gives rational decision-makers great pause.”

    It gives them great pause to face-off each other with nukes, yes — but not to completely destroy a series of dirt-poor, peasant nations that act as their proxies for the occasional venting of their rage.

    Nukes have not ended mass-murder by the superpowers.

    and also your post:


    I do not see much evidence of nukes’ “pacifying effect on world politics” — this may be true from a narrow US/Soviet viewpoint, but not, say, from an Angolan or Vietnamese perspective.

    Rather than venting their rage on each other during the Cold War, the superpowers merely destroyed a series of dirt poor, sorry peasant nations in a long series of proxy wars (due to nuclear weapons). If you were a resident of Vietnam or Afghanistan (or even a US GI), nuclear weapons did not necessarily keep you safe.

  2. Sultan (History)

    “Until Pakistan’s security managers tacitly accept the ground rule about not playing with fire,deterrence optimists for South Asia will remain in the minority.”

    Isn’t this India-centric approach. Seems more concerned about – ‘What if another Mumbai happens.’ One could ask a similar question – What if another GHQ happens? – should Pakistan also resort to the same irrational behavior. Deterrence optimists would only be in majority once security managers from both South Asian neighbors accept the ground rule, and not only Pakistan. Or else, the NSAs would be the major beneficiaries.

  3. FSB (History)

    Regarding the reference to Lashkar-i-Taiba — it would be sensible to examine the root cause of why there is increased terrorism in Pakistan of late.

    The CIA station chief in Kabul has
    done this exercise for us:


    “The situation in Pakistan has gone from bad to worse as a direct consequence of the U.S. war raging on the Afghan border. U.S. policy has now carried the Afghan war over the border into Pakistan with its incursions, drone bombings and assassinations — the classic response to a failure to deal with insurgency in one country. Remember the invasion of Cambodia to save Vietnam?

    – The deeply entrenched Islamic and tribal character of Pashtun rule in the Northwest Frontier Province in Pakistan will not be transformed by invasion or war. The task requires probably several generations to start to change the deeply embedded social and psychological character of the area. War induces visceral and atavistic response.

    – Pakistan is indeed now beginning to crack under the relentless pressure directly exerted by the U.S. Anti-American impulses in Pakistan are at high pitch, strengthening Islamic radicalism and forcing reluctant acquiescence to it even by non-Islamists.

    Only the withdrawal of American and NATO boots on the ground will begin to allow the process of near-frantic emotions to subside within Pakistan, and for the region to start to cool down. Pakistan is experienced in governance and is well able to deal with its own Islamists and tribalists under normal circumstances; until recently, Pakistani Islamists had one of the lowest rates of electoral success in the Muslim world.

    But U.S. policies have now driven local nationalism, xenophobia and Islamism to combined fever pitch. As Washington demands that Pakistan redeem failed American policies in Afghanistan, Islamabad can no longer manage its domestic crisis.

    The Pakistani army is more than capable of maintaining state power against tribal militias and to defend its own nukes. Only a convulsive nationalist revolutionary spirit could change that — something most Pakistanis do not want. But Washington can still succeed in destabilizing Pakistan if it perpetuates its present hard-line strategies. ”


    So if you are truly worried about Pakistani terrorism I would urge you to fix the flawed US foreign policy — one way would be by going out and voting today.

  4. Jens (History)

    I read some chapters of Kapur’s “dangerous deterrent”…I think his analysis (quantitatively and qualitatively) is very plausible. When there is a high degree of stability in the nuclear realm, than Pakistan might not start “low intensity conflicts”…The point is made by Kapur through two very important questions: how is the status-quo power/revisionist power? and who is the conventional stronger/weaker power? his (in my eyes convincing) approach is the so called “instability/instability” paradox ( which is no longer a paradox). this is an interesting when it comes to the questions of arms control in South Asia. A thesis could be that Pakistan has “no interest in a stable nuclear situation.” I hope we can discuss some aspects here.

    • krepon (History)

      Thanks for your comment. Paul’s book is well worth reading. His interviews with Indian and Pakistani leaders involved in crisis management badly undermine the case made by deterrence optimists.
      The old Cold War paradigm of the stability-instability paradox was built around nuclear/conventional warfare. The South Asian paradigm is (as usual) more complex, because its about nuclear/conventional/unconventional warfare.

  5. archjr (History)

    Great post. While I think it increases the potential nuclear danger, the nuclear dyad in South Asia has actually increased stability there. In the early 80’s, the argument was made that the supply of F-16’s to the Pakistanis would decrease their paranoia about India’s substantial conventional dominance and their rush to nuclear weapons. This, of course, turned out not to be true. But absent the risk of some crazies taking over in Islamabad, which is unlikely IMHO, both India and Pakistan seem on a course to build relatively small nuclear forces, as opposed to the orgy of accumulation that characterized US-Soviet competition. Strategic thinkers in both countries doubtless wrestle with the question of “how much is enough,” but their activities since the last real war in 1971 certainly suggest their view is less rather than more, a sort of minimum deterrence posture. Of course, this is highly speculative, since we know so little about the thinking of the elites who make these decisions. What intrigues me is the possibility that escalation dominance will intrude on rational thinking among these folks. Developments in South Asia seem to indicate that this has not yet become a deciding factor. But I don’t believe the way this has borne out indicate a nuclear arms race, at least not yet. India has to consider China’s minimum deterrent, and all military planning there, if logical, must assume that the only purpose of their nuclear deterrent is to deter, and I think they are only viscerally focused on what Pakistan might do. Similarly, I believe Pakistan is only focused on India. There are no other third-party threats, unless Iran goes nuclear, that should lead them to build more than is necessary to deter India. Over time, this might provide a real opportunity for India and Pakistan to work out their own arms control agreement. In the meantime, the government structure in India, and the ability of Pakistan’s Army to take over when the “democrats” overstep their bounds, should provide for a rather stable environment. It of course would be better if neither had nuclear weapons, but it’s certainly too late to walk the dog back.

  6. Byron Skinner (History)

    Good Evening Folks,

    A good question that brings the discussion of this topic to the front of public discussion.

    A point that is always overlooked is what happened in the US and I’m sure the Soviet militaries after the 1950’s and the romance with the atomic bomb faded and a younger generation of pragmatic officers started to rise in ranks.

    The glaring fact gradually became apparent that nuclear weapons in the combat environment we useless. To be effective the enemy had to offer a target of such value as to warrant it total destruction, which neither side was offering or there had to be some geological feature the closing of witch would with denial of access stop the enemies advance, the problem here of course it also denied you access.

    As thinking went on and considerations of post war rebuilding, the human factors and the cost of maintaing a nuclear arsenal build the case that tactical nuc’s were not worth it, and by 1976 the US military was ready to get out of the tactical nuclear business. It took another ten years to convince the American politicians that this was reality. But by 1990 all nuclear cruise missiles. bombs and short range ballistic missiles were removed from western europe and from the US arsenal, with the exception of course of the B-61 bomb.

    In the strategic mode more or less the same problem became apparent how many nuc;s do you need to disable a country?

    A Rand study of the late 1950’s said to militarily disable the United States or the Soviet Union would need about 1,200 well placed hits on strategic targets. New York City would require at least 12 nuclear weapons to disable it. No way did the Soviets United States or anybody else have that first strike capability.

    Byron Skinner

    Those were 1950’s numbers. I’m sure it would take far more nuclear weapons today to reach the same goal of militarily disabling the United States.

    In short the use of military outside of be a terrorists weapon are basically useless.

  7. MarkoB (History)

    Interesting observations. Physicists like to test theories under the most extreme conditions. If they pass, they must be on to something. A Realist would argue that should an Islamist group take power in Pakistan then it would not necessarily follow that jihadism would go nuclear. The revolutionary state would be concerned more with internal consolidation and would be deterred just as any state would be from launching nuclear attacks, whether covertly or overtly. Revolutionary states are socialised by the anarchical state system; Iran, despite all the rhetoric, hasn’t been much of a revolutionary exporter. If that applies to Islamist states then this Realist conception would pass an extreme test, so to speak. The same sort of stuff we say in relation to jihadis was stated of marxist-leninists in the 1950s. I think the stability-instability paradox might lead to something different than what is commonly feared. A jihadi state would use the “stability” provided by nuclear deterrence to engage in more conventional forms of jihadi action, that’s the “instability” part of the equation. In jihadi ideology and grand strategy it is necessary to control some piece of land, preferably in the heart of the Arab world, from which to then launch a wider campaign to restore the “hakimiyya” or sovereignty of God, most likely through the application of conventional force and strategies . Looked at this way nuclear weapons might fit into jihadi grand strategy on “stability-instability” paradox grounds. That’s not classical nuclear terrorism as we conceive of it now for this is more suggestive of deterrence than terror, although the etymology of deterrence does relate to terror. Jihadis are nowhere near reaching this territory phase. They are engaged, following Abu Musab al Suri, in a leaderless small-cell strategy of attrition. For now.

    • BP (History)

      MarkoB, you’ve got the definition of a jihadi, and their intentions, totally wrong in the understanding of a Muslim, but I’m afraid this isn’t a forum to discuss philosophy so I’ll keep away from that discourse here. However your reading on the modus operandi of movements of muslim anti-neo-colonial groups is interesting and I’d like to know more.

      As for Haakimiyyah, it is already with God, we die at times not of our choosing, for instance.

  8. manoj joshi (History)

    Good post Michael. I agree with your basic argument with regard to India and Pakistan. I worry about one additional factor– Indian efforts in the direction of ballistic missile defence which seem to be driven by technology managers without any political guidance.

  9. FSB (History)

    RE. Radicalization in Pakistan, Imran Khan, a Pakistani politician, offers some useful insight:


    “There is only one solution to this chaos. This is to implement an immediate ceasefire and commence talks with all militant groups in Afghanistan. Either America leaves or Pakistan withdraws from this war.

    The US should not worry about Pakistan. Once the bombing stops, it will no longer be jihad and the suicide attacks will immediately subside. About 18 months ago the former head of the CIA’s Kabul station, Graham Fuller, wrote in the International Herald Tribune that once the US leaves the region Pakistan will be stable.

    Political leaders in the US and UK should realize that people in the streets of New York and London are not threatened by the people in the mountains of Afghanistan and Pakistan but by the growing radicalization of their own marginalised Muslim youth.”

    And Graham Fuller, the ex-CIA station chief said in the NY Times:


    “Meanwhile, Pakistan will never be willing or able to solve Washington’s Afghanistan dilemma. Pakistan’s own stability has been brought to the brink by U.S. demands that it solve America’s self-created problem in Afghanistan. Pakistan will eventually be forced to resolve Afghanistan itself — but only after the U.S. has gone, and only by making a pact with Taliban forces both inside Afghanistan and in Pakistan itself.

    Washington will not accept that for now, but it will be forced to fairly soon. Maybe the Pakistanis can root out bin Laden, but meanwhile, Al Qaeda has extended its autonomous franchises around the world, and terrorists can train and plan almost anywhere in the world; they do not need Afghanistan.

    By now, as in so many other elements of the Global War on Terror, the U.S. has become more part of the problem than part of the solution. We are sending troops to defend troops that themselves constitute an affront to Afghan nationalism. Only expeditious American withdrawal from Afghanistan will prevent exacerbation of the problem. “

  10. jeannick (History)

    I would suggest that nuclear weapons , once the novelty wore off , have no influence whatsoever on small scale conflicts ,
    the local players are usually bloody minded and well aware of their irrelevance for their situation,
    It is of concern for two situations only
    to defend one homeland from total defeat
    to be used as tacticals on a third party territory

  11. aditya (History)

    Possibility of nuclear exchange

    INDO-PAK Absence of MAD (+), Religious Nationalism(+)

    INDIA conventional superiority(-), globalised economy(-)

    PAKISTAN failed state(+)