Deterrence and the deal

The focus shifts to the delivery

Regardless of how the pie is sliced eventually, it is clear that for India, the cost of entering the international nuclear mainstream will be more constraints on the number of warheads it can produce. The essence of the compromise, which The Acorn supports in principle, is that this is a price worth paying considering the potential benefits — towards improving energy security, and more importantly, towards improving its troubled relations with the United States and other western powers. But given the reduction in the amount of weapons-grade fissile material and the consequent reduction in the number of warheads in its nuclear arsenal, does this mean that India’s nuclear detterent will necessarily weaken?

Not necessarily. Bear in mind that nuclear weapons, especially strategic ones, are not actually meant to be used. In the first instance, credible evidence of ownership and readiness for use serves as a detterent for potential attackers. Moreover, India’s no first strike doctrine carries with it an implicit signal of an arsenal and a delivery mechanism that can survive a first strike. With this much in place, it is reasonable to conclude that at the present time India does possess adequate nuclear deterrence vis-a-vis its strategic adversaries. Unlike in the cold war context, India shares borders with both China and Pakistan, making it harder to distinguish purely tactical (or battlefield) nuclear weapons. Partly for this reason, India’s no first strike policy (which is implictly an assured second strike policy) covers attacks on Indian troop formations. By making no distinction between attacks on Indian troops on non-Indian territory and attacks on its own territory, India has extended its nuclear deterrence to the tactical space.

But what if the balance of power changes? China and Pakistan have been among the most active proliferators in history and their strategic cooperation will continue. New nuclear powers, with yet unknown intentions towards India, may emerge. Or Pakistan may be further tempted to exploit the stability/instability paradox. Will the constraints on the size of the nuclear arsenal undermine India’s capability to safeguard its national interest in spite of potentially adverse changes in the international balance of power?

Again, not necessarily. Even if the number of nuclear warheads were kept constant, it is possible to increase deterrent capability by improving delivery mechanisms qualitatively and quantitatively. The number of nuclear-capable missiles on land, on aircraft, ships and submarines can be changed to create the uncertainty that underpins nuclear deterrence. It is not necessary for every nuclear-capable missile that is deployed to be fitted with a real nuclear warhead to achieve the deterrence. There are limits to this, but as long as you don’t call “five aces” while playing bluff with one deck of cards you will be taken seriously.

Similarly, qualitative improvements — cruise missiles that can achieve surgical strikes, or submarine-launched missiles that can extend the range — provide additional means to calibrate the posture of deterrence. These need to be accompanied by improvement in warhead designs. It is possible to achieve both without requiring huge amounts of additional fissile material.

The major implication for India’s strategic weapons programme is that it needs to increase its focus on delivery mechanisms. There is already some cooperation with Russia, Israel and the United States on missile-related programmes. A successful nuclear deal with the United States should make it easier for India to expand these programmes. Naval capabilities — including fleet expansion and the replacement of the navy’s aircraft carrier — are scheduled for the near term. India will need to accelerate the deployment of a nuclear submarine, and for this reason, should not accept any proposals to place facilities related to this project on the civilian list. Fortunately, India’s emerging security policy already recognises these imperatives.

Much of the separation anxiety has been driven by an assumption that India is about to accept a deal that will undermine its national security by weakening its nuclear deterrence. That is only true if the number of warheads is the sole determinant of deterrence. But even with a small number of warheads its the capacity to deliver them that makes a big difference. Indeed, if India imaginatively avails of the opportunities created by improved relations with the United States and the West as a result of the nuclear deal, it may well be able to improve its nuclear deterrence.

18 thoughts on “Deterrence and the deal”

  1. Nitin,

    I believe the whole deal may affect us in another, very bad way. If & when India & US sign the deal, India would become (or often seem to be) a member of the intra-Big 5 cartel (along with US & UK), that is increasingly becoming a grand alliance against islamic proliferation & terrorism. While there maybe nothing wrong in fighting proliferation & terrorism, the targets would more often be chosen as per the priorities (read: convenience) of US.

    The problem could arise because India is negotiatiing only with US for an acceptance into the Big 5 club. UK is behaving more & more as a sidekick of the US, so am sure US & UK will have no problems in accepting India as one of them. However, the pitfalls are obvious as China would most likely not accept India as an equal partner. Economic factors would gradually alienate India from Russia, as India starts to depend more on the developed countries (US & Western Europe) for economic & technology (both military & civilian) support, and as India itself becomes a much larger economy than Russia. India cannot even consider Russia as a partner for providing big-time energy security, as any pipeline to India from Russia would have to run across China, and bigger still, the Himalayas. As reasons for considering Russia as a natural ally diminish, India could be forced to side with or would be widely seen siding with the US & UK cartel.

    The silver lining, of course, is France, which has managed to keep an independent policy but it remains to be seen if India can do the same. I do not have any data on this, but I would guess Indian economy (especially as its services pie grows) is more strongly associated with & dependent on the US than the French economy is, and hence it might be tougher for India to disagree with US on major policy issues.

    The major issue with graduating to become a member of the US-UK-India cartel (or even seem to be a member of the cartel that is fighting islamic countries) could be internal, as India’s secular fabric could be challenged in a way that might make Gujarat riots look harmless. As long as we have the Bajrang Dals and people like the UP minister probability of this happening are too high for comfort.

  2. Nitin:

    Very well said w.r.t. nuclear deterrance.

    You talk about improving delivery mechanism, while the DRDO is waiting for ‘yes’ from political bosses to test Agni-3. The bosses don’t want to do it so as to not offend the US (similar episode circa late 80s when US applied pressure on Rajiv Gandhi not to test Agni-2). All this while Pakistan is freely resting Abdali (read M-11).

    The NP ayatollahs eyed on the RMP plant which is the enrichment facility for our nuclear submarine project.

    The US also wants ‘perpetual’ safeguards agreements – when there is nothing perpetual in international relations. India should at any point go back on the safeguards (under rare circumstances or changed geo-political situations) without inviting any sanctions from Congress.

    Al this despite India offering 14 of the 22 reactors and all future ones imported under IAEA safeguards.

    Do check out my posts here and here.


  3. Please learn how to spell deterrence if you are going to deal with such weighty matters in your website.

    Secondly, all of India’s nuclear weapons are “strategic”. There are no tactical nuclear weapons envisaged by the Indian Nuclear Doctrine.

    Please be clear on this point.

    Thank you very much.

  4. Chakrapani,

    Thanks for pointing out the error in spelling. The wonderful thing about blogging is that not only do people like you point such things out, it is also possible to correct the error rather easily. In any event, the error in spelling should not deter you from looking at the arguments in their own merit.

  5. I would ,unfortunately, tend to disagree with the arguments here. It would be nonsense to assume that the weapons are just a deterrent.

    If Indians are willing to give away their freedom Now, then why did they fought for independence at all. Indians probably would’ve done ‘not-so-bad’ even if the British Raj was still around, and would’ve been in the western club by default.

    Giving up the goal of increasing nuclear warheads for a short term might make sense, but getting locked in into a contract for a long tern would be nothing but foolish.

    Again, I see problem in the argument that ‘the nuclear weapons are not actually meant to be used’. May be not the big ‘Bomb’, but deviations of them. Taking out the options to use the even the big ‘bomb’ would again be nothing but suicide against psycho nations. When Indians limit their nuclear grade material quantity, Indians would be giving them a predictability factor, and no one ever won a predictable war.

    Also, we’re taking away the option to even conduct experiments with a ‘free-will’, which may or may not lead to more nuclear weapon grade material.

    This nuclear agreement would be good, no doubt, but should not be just favourable to one party of the agreement. Its not like Indians lost a war and are trying to sign a peace-treaty with the winners.

    Just a few points taht I could notice…

  6. Nitin,

    I have to disagree with the main arguments. While I agree with the premise that better delivery mechanisms and the completing of the triad have to be our primary thrust areas, I do not believe that capping our nukes as of date or even by the end of this decade is an answer.

    The arugment that strategic nukes are not meant to be used and are only needed to deter is a best-case scenario(which I hope prevails for eternity, but truth is stranger and more bitter than fiction), assuming capping our nukes as of now will not affect this deterrence assumes a second best case scenario – the one where we will not have to face the choice of second strike and/or where we will strike but not expect a retaliation.

    While one would not want such a situation to arise, being as we are, in a position of less uncertainity – owing to our avowed NFU and clearly stated terms of nuclear engagement, it would be foolhardy to assume that if, god forbid, an occasion arises for us to strike second, X number of warheads would do – and or that it would not spiral out of control.

    Our nuclear deterrence maxim while stating to be of minimum deterrence also envisages an “unacceptable” damage to an adversary should they, god forbid, strike us first. The definitions of unacceptable damage change over ther years – Kargil in 65 turned into a full scale war, in 99 was confined to one region of the border.
    We are nowhere close to having enough capability to inflict unacceptable damage to China. Neither in terms of reach, nor in terms of our nukes destructive/deterrent power.

    5 tests do not a nuclear power make. Much less one with two proliferating ones on its border. We might reach a stage at the end of this decade where we could reasonably think of capping our fissile material, but IMO now is not the time. We dont have enuf ammo to deter our adversaries just yet.

    Also, the perputuity clauses on our testing moratorium and the safeguards for our reactors is a cause for worry. Hope Dr MMS can get rid of those.

  7. Hi,

    You can’t put nuclear weapons on cruise missiles. It is banned under the INF treaty.

    I once ran a long thread about deterrence on BR. I think they still have it: link…it may be worth looking at.

  8. Also in the India-Pakistan context you cannot mate warheads to weapons. The no-mating clause is the foundation of deterrence stability.

    All this stability instability paradox is for idiots who don’t understand that there is no connection between conventional deterrence and nuclear deterrence and repeatedly link the two.

  9. Hi

    The first line of the last post should read, “you cannot mate warheads to delivery systems”.

  10. Prasanna and Sachin,

    The nuclear deal with the United States will not cap the number of warheads or even freeze India’s plutonium stockpile. It will constrain the amount of fissile material that India can divert from the civilian segment to the military segment. As PM Singh is to clarify today, India will be free to build new reactors for military purposes. So that argument about ‘capping’ can be set aside.

    Secondly, while constructing a nuclear deterrent is with a view to credibly threaten its use in certain circumstances, its objective is to use the threat to avoid the actual use of these weapons. In such a scenario, it is unlikely that apart from reasons of replacement and technological obsolescence, warheads are going to be ‘used up’.


    Thanks for those points, as indeed your threads at Bharat-Rakshak, which are among the most informative discussions on the subject in the public domain. My argument is India is not constrained from designing warheads that can be fitted onto cruise and submarine launched missiles. Developing and possessing the know-how to do this is a way of calibrating the nuclear deterrent. The fact that doing so will be seen as provocative and suggestive of a new arms race points, indirectly, to its utility as a way to enhance the deterrent. Of course, whether or if India should actually deploy such systems is a different matter.

  11. Nitin,

    On paper, as of now, the deal may not affect much the fissile stockpile/production, but current events point to it affecting the production – as your own piece points out, there’s to be an expected (significant) amount of loss in fissile material production.
    Now, it is upto our top dogs to ensure that the significant amount isnt life-threatening to deterrence – I am of the view that current concessions seem to be taking us in that direction, and would like the Govt. to correct course.

    On the deterrence issue, yes, the nukes are there to prevent an attack on us – by threatening massive and disproportionate retaliation. My point was that the rules of the game change (which was why I referred to Kargil) – so assuming that our current level of deterrence will be okay 20 years down the line is not correct. Not, especially when you dont have democracies on either side – and armies are required to buttress major policy.

  12. Hello Nitin,

    I cannot speak to the escalatory potential inherent in India developing nuclear warheads for cruise missiles (esp. the submarine launched variety). That kind of talk could only come from someone who has evaluated the nature of Pakistan’s response to such a gesture and the cost of actually developing such a system i.e. someone sufficiently senior in GoI.

    I do not see any escalatory potential if India decides to develop nuclear warheads for submarine launched ballistic missiles, but that is just my personal opinion and not that of anyone else.

    Deploying the system as you point out is a whole other kettle of fish.

    From a deterrence perspective, I would be keen at this point to invest in better air defence systems as they will provide a greater measure of early warning and countermeasures along the himalayan wall. A deeper interest in ballistic missile defence is also warranted, though I would be against a full scale investment until a better air defence shield is in place.

    I strongly caution against linking conventional and nuclear deterrence. For example the Pakistanis have told the non-proliferation mullahs that they have a “nuclear redline” on “India fomenting internal unrest in Pakistan”. Now some of their propagandists are publicly blaming India for the mess their army has made in Baluchistan. Perhaps they do not realize that this creates all sorts of questions…

    If indeed Musharraf is convinced that India is making problems in Baluchistan then his redline has been crossed, so where is the vaunted Pakistani first strike?


    1) The propagandists are lying to the Pakistani public and Musharraf knows India is not involved or

    2) Musharraf lied to the Non Proliferation Mullahs and this is not a redline at all or

    3) Some unknown faction inside the Pakistani government is generating an excuse for a first strike against India and Musharraf is unwittingly being pushed into it.

    Mind you I am in no way breaking with the narrative that “Musharraf is a good guy” and yet there are so many contradictions. This is mess that follows from a redline.

  13. Hi,

    I see there is some discussion on what is the right size of the deterrent. I fear I know of no way to exactly determine such a number but I suggest the following crude algorithm.

    If K Sub’s idea is correct then the credible threat of delivering a nuclear weapon to even one target inside Pakistan should be sufficient to deter a Pakistani first strike.

    Suppose I set that as the limit – then the question that has to be posed is which places must we target so as to ensure that the mere possibility of even one strike on any of these places is sufficient to deter Pakistani nuclear attack. This list most likely includes 10s of names of places like Tarbela, Mangla, Warsak, Islamabad, Rawalpindi, Karachi, Lahore, Quetta, Peshawar and so on. How long this list is determines how many weapons we need. For argument’s sake let m, be the list of places we have to demonstrate credibly the ability to hit, in order to deter a Pakistani first strike.

    Let us assume that our delivery systems are X percent accurate. This means that to guarentee even one delivery of a nuclear bomb we need 100/X weapons.

    Similarly lets say due to natural radioactive decay and other factors, we need to refurbish Y percent of our weapons. So in order to have 100 working weapons, we need 100+y weapons in stock. This determines the size of our stockpile.

    Thus the total number of weapons we stockpile for in order to guarentee one credible strike against a chosen target in Pakistan is given by the equation:

    no of weapons = 1 x m x (100/X) x (100+y)/100

    if we relax K Sub’s requirement and determine that we need to demonstrate the ability to credibly hit “n” targets in Pakistan to deter a Pakistani first strike – then the number of weapons above is simply multiplied by n.

    For example:

    If we want to hit 10 Pakistani targets, and our delivery mechanisms are 50 percent effective and we lose 10 percent of our weapons to stockpiling losses. then we need

    no of weapons = 1 x 10 x 2 x 1.1 = 22 weapons.

    Please note I have pulled the efficiency of the weapon and the stockpiling losses out of thin air. Someone who deos this in GoI will have the real numbers and they will not be shared.

    So per this calculation we will need about 22 weapons stockpiled for.

    The number of weapons we will actually deploy will be different as this number is related to a different calculation.

    The number deployed has to be calculated using the idea that deterrence has failed and despite our best attempts Pakistan carries out a nuclear first strike.

    Now the effort will focus on eliminating the Pakistani nuclear command and control chain so that it cannot launch another strike against India. Generally the nuclear command and control chain will be a much smaller number of targets than the number needed to deter a strike. Once this chain has been eliminated conventional forces could be used to mount a punitive retaliation against the aggressor. So lets assume that the number of targets needed to eliminate the command and control chain is k.

    The no of weapons needed for this is:

    1 x k x 100/X x ((100+y)/100)

    What I am trying to say is that the number of weapons we need for deterrence will exceed the number we actually have to use by a considerable margin.

    This calculation is more difficult in the case of China which has not spelt out what its strategy is vis-a-vis India. With regards our Chinese friends we only know that they have given us an assurance that they will not deploy their short range ballistic missile fleet on the TAR plateau. The aggreement does not apparently cover nuclear capable aircraft on the TAR airfields. This creates a whole new set of questions that cannot be easily answered.

  14. Nitin,

    As Prasanna points out in his comments, the rules keep changing.

    And if anyone thinks that India can get out of this contract later, even though contractually it can, it would be naive to think that there would not be any repurcussions (political\geo-strategic\economic\monetary\organisational\etc).
    Its going to be like a one way street, where if you want to go back, there is a long time consuming detour.

  15. Sunil,

    I’ve written about why the stability/instability paradox and Pakistan’s nuclear blackmail are bunk before. So I will summarily agree with you there; if there are differences, they are ones of nuance. Similarly on air defence.

    On your deterrence calculations (which I must say are a great piece of analytical work), I view the game slightly differently. Conceptually, I do not think that deterrence needs to be based on causing unacceptable damage to all ‘m’ targets. Even if ‘m’ is one or a small number, and even with relatively poor levels of accuracy, the threat of a nuclear attack (albeit less devastative) will remain a powerful deterrent until it occurs. As long as all the variables in the equation are sufficiently ambiguous (and there are signals of improving reach and accuracy) it is possible to retain deterrence, and also compel the adversary to incur costs of defending its targets.

  16. Hello Nitin,

    My crude calculation allows us to keep the Pakistanis guessing as to which targets we may choose to hit while we retain the actual capacity to target a large number of places inside Pakistan at will. That is why that factor of “m” is in that calculation, otherwise I would have dropped the m term altogether and stuck a narrower interpretation of what K Sub says. Done in this fashion, the Pakistanis know that we have atleast enough fissile material stockpiled for “m” guarenteed hits where exactly those hits will fall is left to their imagination.

    What you are describing is the Pakistani calculation.

    The Pakistanis routinely talk about hitting 4-6 major population centers (Mumbai, Delhi, Calcutta, Chennai, Bangalore and Hyderabad) and when pushed to the wall, they talk about hitting Hargobind Sagar and some other dams. Unlike our weapons (40-60 kT), the Pakistani weapons are in the 5-15 kT range. When absolutely pushed to the wall, the Pakistanis say they will use their nuke on an advancing Indian military column on Pakistani soil. So the Pakistanis rely on the thing you are talking about – ambiguity. It is for this reason that some analysts find the idea of Pakistani deterrence too fake. The Pakistanis try to ambiguate the size of their stockpile and their non-proliferation buddies prop up the lies they tell about the Pakistani stockpile. I am not saying that ambiguity isn’t a part of the game, but the Pakistanis carry that too far to the point that it undermines the credibility of their deterrent. To this day I am hard placed to tell you exactly how much U-235 the Pakistanis have and exactly how many working bombs they could make out of it. The NP mullahs parrot out the 50-95 weapons line but frankly I have no idea how they assess the proliferation risks with such wide estimates. David Albright tells us that Pakistan’s nuclear weapons are “under lock and key” but how many locks and how many keys? 50 or 95? Can you perform any real counterproliferation with such a wide error bar on the estimates?

    I feel this lack of credibility can breed an escalation so beyond a point the Pakistanis have to come clean and say “this is what we do…” .. the problem I find in talking to the Pakistanis about this is that they tend to see any discussion on nuclear issues as an excuse to carry out rhetorical escalation. Their social programming makes it easy for them to talk absolute rubbish even on the most critical of fora and you actually have to knock them on the head with a few cold facts before they shut up and actually listen to you.

Comments are closed.