You are here


Reviewing India’s Nuclear Doctrine

April 24, 2009

A long standing observation on India’s strategic culture is that national strategy remains unarticulated. A significant departure from this characteristic was made by India following a review of the nuclear doctrine in Jan 2003. It is now more than six years since the event. There is a need to review doctrine periodically in any case. In this specific case the need is more acute given changes in strategic circumstances. The present juncture is an apposite one in that a new government would be coming into power soon. Therefore initiating a case for a review of India’s nuclear doctrine is in order. This policy brief proposes a direction of review by interrogating a principal pillar of the doctrine – that of ‘massive punitive retaliation’.

There are other contending directions of review. These include whether India should continue to include ‘minimal’ in its formulation ‘credible minimum deterrent’ in light of ‘minimum’ seemingly contradicting the important dimension of the two i.e., ‘credible’. There has even been a recommendation by a departing National Security Advisory Board on jettisoning ‘No First Use’ – perhaps the most salient pillar of the doctrine. The votaries of the Triad would prefer a mention of a Triad based second strike capability in the doctrine. These possible directions indicate that there is a need for review. It is another matter that in doing so, some of the proposals would be accommodated and some disregarded.

In this regard, the proposal requires a shift away from ‘massive punitive retaliation’ in favour of ‘flexible punitive retaliation’. The policy brief first establishes the need to do so by discussing three conflict scenarios highlighting the dangers of the formulation and the advantages from the proposed shift. It concludes that a strategic dialogue with both China and Pakistan is necessary for clarity in communication. This would enhance deterrence and dispel possible misperceptions and apprehensions. This is particularly necessary with respect to Pakistan, given that the state is perpetually poised on ‘failed state’ status with implications for India.

The current doctrinal precept

The sub-paragraph of interest of the press release subsequent to the Cabinet Committee on Security endorsing the nuclear doctrine of 04 Jan 03 reads: “(ii) A posture of “No First Use”: Nuclear weapons will only be used in retaliation against a nuclear attack on Indian territory or on Indian forces anywhere; (iii) Nuclear retaliation to a first strike will be massive and designed to inflict unacceptable damage.”

The inclusion of the term ‘massive’ was a discernible change from the earlier formulation of the Draft Nuclear Doctrine in which the term had not found mention. Instead the Draft had used the term ‘sufficient’ implying a degree of choice on the nature of the response being available to the political decision maker. The specific sentence in the sub-paragraph on Credibility in the Draft reads: ‘Any adversary must know that India can and will retaliate with sufficient nuclear weapons to inflict destruction and punishment that the aggressor will find unacceptable if nuclear weapons are used against India and its forces.’ Though the Draft was just that - a ‘draft’ to compel the government’s attention, the critique stands. The principal problem with the change is that it restricts the choice of the decision maker by excluding the set of less expansive responses.

‘Massive’, not defined explicitly, can be taken as a product of throw weight and target set that produces the promised ‘unacceptable damage’. There are three implications: one is in terms of ‘pain’ implying counter value targeting; second, is reducing the ability of the enemy to mount a counterstrike, which would be counter force; and third is a mix of both. Since in all three options ‘unacceptable damage’ is inflicted, it is worth questioning whether only ‘massive’ nuclear counter strike would cause ‘unacceptable damage’. It is well understood that even a single warhead through a counter value strike can be ‘catastrophic’. Therefore, the term ‘massive’, in its emphasis on throw weight or numbers, is superfluous. It has even been averred that the inclusion of ‘massive’ was likely an ‘unconsidered formulation’. On this count there is a need for review.

Massive nuclear retaliation is definitely a possibility and would be credible in case the enemy’s nuclear first use is in an expansive (‘massive’) form such as resort to first strike, decapitating strike or counter value targeting. However, should ‘first use’ be of a restricted nature such as at the tactical level, for India to up the ante by going ‘massive’ to counter it would be irrational. This was an observation true in the Cold War era as pointed out by Thomas Schelling in his landmark, The Strategy of Conflict: ‘The threat of massive retaliation, if ‘massive’ is interpreted to mean unlimited retaliation, does indeed lose credibility with the loss of our hope that a skillfully conducted all out strike might succeed in precluding counter retaliation.’ Since precluding counter retaliation is not possible in India’s case with respect to Pakistan, leave aside China, it would be prudent for India to go down a route traversed by the US during the McNamara years. The logic that persuaded McNamara in his own words was:

‘One cannot fashion a credible deterrent out of an incredible action…What we are proposing is a capability to strike back after absorbing a first blow. This means we have to build and maintain a second strike force. Such a force should have sufficient flexibility to permit a choice of strategies… Such a prospect would give the Soviets no incentive to withhold attack against our cities in a first strike. We want to give them a better alternative…the strongest possible incentive to refrain from attacking our cities.’

India’s promise of massive counter strikes to first use against its territory or its forces is wanting in credibility, particularly if the strike were of a tactical nature but with a strategic purpose of nuclear signaling for war termination. This is particularly important since both the likely adversaries are unlikely to resort to nuclear weapons in a massive mode in the first salvo.

Consider the case of China. Though bound by an NFU, it is reportedly a qualified NFU in not being applicable to territory it claims. In a border conflict with India it could resort to nuclear first use on its claimed territory of Arunachal Pradesh. Such use would likely involve the use of tactical nuclear weapons. Since India’s is an Assured Retaliation doctrine, India would only be complicating the aftermath of the nuclear exchange for itself should its counter strike be ‘massive’.

The same is the case with Pakistan. Pakistan, emulating NATO in the Cold War era does not profess NFU. In case it were to resort to nuclear first use, it is quite apparent that this would not be of an order of a debilitating ‘first strike’ given the imbalance in numbers and the security of information surrounding locations of Indian nuclear assets. Even if it were to attempt to do so, it could not preclude assured Indian counter value retaliation. Having fired off a major proportion of its arsenal in attempting a first strike, it would not have the numbers to mount a counter strike. In effect, it would ab initio be deterred from attempting a first strike. Therefore Islamabad’s most likely first use is a tactical strike with a strategic purpose of forestalling Indian conventional military advances or to bring about conflict termination by focusing the efforts of the international community. Counter retaliation in a ‘massive’ mode to such a symbolic strike would be to India’s disadvantage since there is no guarantee that some Pakistani weapons would not survive. These would inevitably be directed at counter value targets to maximize vengeance. To open itself to such a threat would be irrational.

The problem has been pointed out earlier following the release of the Draft nuclear doctrine in the following manner:

‘….Our intent of causing ‘unacceptable damage’ is credible only in case our population centers and nuclear-industrial concentrations are hit, inclusion of military forces as targets that will invite such a response makes it less credible…the point is having caused ‘unacceptable damage’ is no consolation for ending up a recipient of it…Thus there is a need to move beyond the avatar of ‘massive retaliation’…in favour of ‘flexible response’…’ (Ali Ahmed, ‘Doctrinal Challenge’, USI Journal, Jan 2000)

It is seen that the term ‘massive’ is not only tying down India’s options but dangerously so. This is elaborated through scenarios in the next section with respect to Pakistan as the nuclear adversary. In the case of China as an adversary in similar scenarios, there is no way India could survive the eventual nuclear exchange.

A consideration of scenarios

Scenario One – ‘Massive’ defined as infliction of ‘pain’

In a conflict, if Pakistan was to resort to limited nuclear first use, such as targeting advancing military forces on its territory with a single or limited number of warheads, then India, in accordance with its doctrine, would have to launch a massive nuclear retaliatory attack. In case inflicting pain is the aim, counter value targeting may result. In case Pakistan’s ability to retaliate is not taken out simultaneously, Islamabad is likely to counter as massively on population centers. This could result in India finally attempting to take out Pakistan’s ability to retaliate by launching attacks similar to a first strike with a mix of counter force, decapitation and counter value strikes. This would render Pakistan with a scattered nuclear capability that would intermittently target Indian population centers indiscriminately, leading to repetitive attrition attacks by India. This would of course ‘finish’ Pakistan, but at grievous cost to India, in that India would itself have suffered ‘unacceptable damage’. Therefore it can be seen that ‘massive’ punitive response does not serve India’s national interest.

Scenario Two – ‘Massive’ defined as taking out the adversary’s ability to retaliate

Non extensive first use by Pakistan in this case would require that India respond with ‘first strike’ levels of retaliation with a mixed counter force, decapitation and counter value targets. This would leave Pakistan with little to counter strike with other than a limited capability. It would likely therefore respond with vengeance attacks to the extent possible. These would be suppressed by India with nuclear or conventional means but only after the fact. Thus it would eventually suffer considerable counter value damage amounting to ‘unacceptable’ levels. Even if India would have won the ‘nuclear war’, it would have lost the war politically. Therefore even if Pakistan does not exist as a coherent nation state thereafter, it would be of little consolation to India and certainly not in Indian national interest.

Analysis Alpha

Clearly, Indian national interest is not served by a doctrine that posits ‘massive’ punitive retaliation. A doctrine should be applicable sui generis. A differentiated doctrine in the form of one response type with respect to China and another with respect to the other adversary, Pakistan, would not be tenable. Therefore there is a case to consider the alternative. This would be in the form of ‘flexible’ punitive retaliation with the nature of the first use determining the likely Indian response. A scenario on how this would play out with respect to Pakistan is below.

Scenario Three – ‘Flexible’ punitive retaliation

In case of the enemy’s non-extensive resort to first use, India has the option of retaliating in kind or at a higher level for punishing the enemy and conveying resolve. This would either lead to the enemy countering, escalating or terminating the exchange. In case of an equivalent counter on India’s part, there is no incentive for the enemy to either counter or escalate, both having suffered equally. In case of higher level counter strikes by India, the enemy may get even but may be deterred from doing so in the face of Indian resolve evident in its retaliatory response. Therefore, a lower level response by India, particularly to counter tactical strikes (even if a strategic purpose of nuclear signaling or war termination) is non-escalatory. This would be preferable to exposing value targets to heightened danger through counter value targeting.

Analysis Bravo

In case of receipt of enemy first use in a non-extensive mode, India can draw on a menu of options ranging from quid pro quo through quid pro quo plus to spasm responses. Responses commensurate to the provocation are given in the table below:

Type of first use Commensurate response
Nuclear tests No nuclear counter strike
Demonstration ‘No Target’ strike
‘First use’ on tactical battlefield target in his own territory in defensive mode Quid pro quo
‘First use’ on tactical battlefield target in offensive mode on his own territory
‘First use’ on tactical battlefield target in Indian territory Quid pro quo plus
‘First use’ on counter force targets
First strike ‘Massive’ punitive retaliation
Decapitating strike
Counter value strike(s)

Answering ‘What deters?’

Invariably such consideration as being done here has to contend with this question. The logic apparently informing the ‘massive’ punitive retaliation formulation in the words of Gurmeet Kanwal is:

“However there can be no doubt that for India’s No First Use to be credible, India’s strategy should be to target high value population and industrial centers in adversary countries with a high level of assurance after absorbing the full weight of what in all probability will be a disarming first strike. Only then would the adversary be sufficiently deterred to avoid launching a nuclear strike against India.”(‘India’s Nuclear Force Structure’, Strategic Analysis, Sep 2000)

Firstly, a ‘disarming first strike’ is not the most likely form of nuclear first use by an enemy. Pakistan does not have the capability and China would not have any reason to do so. Secondly, India only requires the capability spelt out by Kanwal to be able to deter such a strike by developing the required second strike capability. Kanwal is right in believing that this ability would deter first strike levels of attack. For deterrence, Kanwal notes capability, intent and resolve are required to take out up to ten cities in retaliation. Since this would expose Indian cities to similar retaliation from surviving an enemy nuclear capability, particularly if his nuclear retaliatory capability is not also addressed simultaneously, it is not credible. Further, such ability may not deter lower levels of nuclear attack. First strike levels of counter for lower order nuclear provocations are incredible on account of being disproportionate and, as seen earlier, irrational. Having this capability is no guarantee of deterrence of lower order nuclear first use. Therefore, there is a need to be prepared for the enemy’s non extensive first use. This means moving towards ‘flexible punitive retaliation’.

Popularly the answer to the question is that an enemy be convinced of ‘massive’ retaliation and not break the nuclear taboo. This choice being with the enemy leadership can only be influenced and not guaranteed. Therefore the reaction requires thought. In case of ‘flexible’ punitive retaliation, deterrence is not disadvantaged because what deters is not so much certainty of retaliation but the mere possibility. Therefore given that the nature of the response is with the target state, there is no guarantee that it would not be exercised in a spasmic. This is enough to deter. Since deterrence cannot be guaranteed, it would be prudent not to be held to an unenforceable promise. Compromising on the promise on receipt of a nuclear strike would bring credibility further into question. Therefore it would be better ab initio to move to a flexible punitive retaliation doctrine, one that does not in any case rule out ‘massive’ punitive retaliation. The element of ambiguity is thereby better preserved, even as the clarity necessary for deterrence is communicated with potential adversaries.

The strategic circumstance

It would appear Indian patience is wearing thin in light of continuing terrorist outrages originating in Pakistan. The Deputy Director General of the Institute for Defence Studies and Analyses observes:

‘The nation should also shed its reluctance to use force that is ingrained in its national psyche, cemented by centuries of occupation by foreign powers and failure to be an occupying power. The state owes to its citizens its primary duty of being a protector. Indian pride is already hurt as was clear in the aftermath of the Mumbai attack. Another major attack by terrorists in India would invariably trigger an avalanche of public opinion in the country for direct action that no government in power would be able to resist, and equally, render powerless any US effort to stop it.’ (

The possibility of more robust measures being taken in response to such future attacks not only from punishing the originators but also to deter them cannot be ruled out. In case the Pakistani establishment is handled in a measured manner in such a circumstance, then escalation into conventional conflict can be precluded. However, the possibility exists and so does the escalatory potential into the nuclear domain. The Indian Army has adopted a Cold Start doctrine that envisages swift administration of retribution. Though the doctrine of the Air Force is classified, it no doubt has a strategic bombing component. Naval posturing off Pakistan’s most important city and port would further impact Islamabad’s calculus. Internal instability generated by the Islamist wild card can be apprehended. Concerted in-conflict pressures could impact the nuclear threshold, unhinging it from an initially ‘high’ threshold to a lower level. This may eventuate in early first use, making Indian responses critical to the future course of the conflict. Therefore rethinking nuclear doctrine at this juncture would be timely and makes eminent sense.


Taking on the counter arguments against the proposed shift is only fair. The logical question would be ‘Would Pakistan be emboldened if the threat of going ‘massive’ is withdrawn?’ Pakistan would in any case be reckoning that India would not reflexively resort to ‘massive nuclear retaliation’ since it is not rational to do so for the reasons already noted. Therefore withdrawing the threat of the same would not impact their calculations.

The second question is of a higher political order: ‘Why not cultivate and demonstrate political resolve?’ If as indicated the threat of ‘massive’ punitive retaliation lacks credibility for lower order nuclear provocations, for votaries of ‘massive’ punitive retaliation India needs to demonstrate that it would not be self-deterred. This would imply changing Indian political and strategic culture to accommodate a particular form of nuclear use, which would mean the ‘tail wagging the dog’. It is logical that it is not the political and strategic system that is to adapt to doctrine but quite the other way round. The doctrine has to be in accord with the prevailing political and strategic culture of the state. Tying down the political decision maker to only one option is both irrational and undemocratic.

There is a possibility that given the advantages of ‘flexible punitive response’ and the disadvantages of ‘massive punitive response’, Indian doctrine may indeed be one of ‘flexible punitive response’. In this circumstance, the declaratory doctrine may be departing from the actual one in the belief that promising the worst would stay the nuclear card more appropriately. As has been shown this may not be the case. Therefore, for the actual doctrine to be different from the declaratory doctrine should be for a very strong reason. A departure from the declaratory doctrine militates against clarity that should attend nuclear doctrine as posited in nuclear deterrence theory. Therefore, there is no reason why this should be so. If this be the case, then there is all the more reason for review and the two should be brought to converge as the outcome of the review.

Lastly, the nature of the ‘flexible punitive retaliation’ needs to be considered briefly. Since it would emerge from the review, it is not intended to anticipate the outcome here. Suffice it to mention that a nuclear war cannot be ‘won’. Inflicting graver punishment on the other side is no consolation for receiving a nuclear strike, even of a lower level. Therefore, there is a case for terminating any nuclear exchange at the lowest level possible. To this end extension of negotiating terms to the adversary that are mutually acceptable needs to be done and must be articulated as such in the doctrine. This concept was the original contribution of General Sundarji (Vision 2010: A Strategy for the Twenty First Century, New Delhi: Konark Publication, 2003) and could inform the proposed review.

The promising departure from Indian strategic culture that the nuclear doctrine articulates, first in the form of the Draft and next in the public adoption of the nuclear doctrine in question, should be taken further. A review of the nuclear doctrine a decade after its contours were first expounded is in order. The incoming political dispensation would do well to establish this as a refreshing trend in Indian strategic culture. The national interest would be best served in this case by a movement towards a ‘flexible’ doctrine.

PDF icon PDF Version171.17 KB