India and Pakistan remain locked in a nuclear-armed rivalry in which each side fears aggression in multiple forms from the other. In this tense environment, the maintenance of deterrence stability is key to avoiding conflict and potential escalation to nuclear war. But classical models of deterrence may be insufficient to understand and manage security challenges in today’s South Asia.
Pakistan and India compete sharply in Kashmir and, now, in Afghanistan.1 Each believes with varying intensity and evidence that the other projects agents of violence to subvert its domestic order. India cites the terrorist attacks on Mumbai in 2008 and on the Lok Sabha in New Delhi in 2001; Pakistan alleges that India abets the insurgency in Balochistan. These causes of insecurity stimulate conventional military preparations and an unregulated build-up of fissile material stockpiles, nuclear weapons, and delivery vehicles.
In this environment, the objective of strengthening deterrence stability is highly advisable. The ideal goal of inducing these states to abandon their nuclear arsenals and join the NPT as non-nuclear-weapon states is impractical, even if the US and others cannot publicly abandon it due to global considerations. The priority now should be preventing war between India and Pakistan which could too easily escalate to nuclear exchanges. Indeed, deterrence stability is a better framework for conceptualizing and redressing the nuclear challenge in South Asia than is focusing on preventing “loose nukes” and nuclear terrorism. The threat of India-Pakistan war is more immediate than that of nuclear terrorism. In any case, deterrence stability would reduce the risks of terrorists acquiring nuclear weapons. Moreover, India and Pakistan would be more inclined to engage in dialogue and Confidence-Building Measures framed around deterrence stability than they are when the agenda seems to reflect other US priorities such as countering nuclear terrorism or strengthening the nuclear nonproliferation regime.Theories of nuclear deterrence rely on the assumption that nuclear competitors are “unitary rational actors.” Bruce Bueno de Mesquita and William H. Riker concisely express this assumption: “national decisions of such magnitude as acquiring a nuclear capability or using such a capability in a war are made by a single, dominant leader who is an expected utility maximizer.”2 Among other things, this assumption extrapolates to the nuclear domain the key feature of a modern state as defined by Max Weber, that a state exercises a monopoly on the legitimate use of force in and from its territory.
To the extent that analysts and policy-makers have worried in recent years about instances when the unitary rational actor model might not obtain, they have tended to focus on the problem of irrationality. Terrorists are presumed to be undeterrable because they do not conform to the prevalent model of rationality. States led by mad men, religious zealots, or neo-Hitlerites also are deemed to violate the model. More broadly, as Lawrence Freedman has noted, the unitary rational actor model does not account for “a whole range of psychological and sociological factors – such as mental quirks, lack of awareness, domestic political pressures, value-conflicts or sheer errors of judgment.”3 As will be discussed below, irrationality is a real problem and it may not be limited to terrorists and militant states, as historical examples from the Cold War suggest. But, the greater problem may relate to the first adjective in the “unitary rational actor” model: unitariness.
Social scientists have observed, of course, that complex systems such as states are often riven by competing organizations, interest groups and personalities. The internal dynamics within “normal” states render the unitary rational actor model an ideal type rather than a descriptive reality. Nevertheless, when it comes to functions as portentous and centrally controlled as initiating and managing warfare between nuclear-armed states, it is generally assumed that a tight, coherent line of authority operates approximately in ways consistent with the unitary model. If a state is not functioning as a unitary actor, or claims not to be when it is convenient, or is not perceived to be by those who seek to deter it, the implications for deterrence stability are profound.
When India is attacked by actors emanating from Pakistan and with ties to Pakistani intelligence services, it naturally infers that such actions represent the intentions and policies of Pakistani authorities. The projection of violence from Pakistan into India means that deterrence (through non-nuclear means as well as nuclear) has failed to prevent aggression. The task then remains for India to threaten or undertake punishment to compel Pakistan to redress the offense and to deter Pakistan from repeating it and from escalating the conflict.4 If Pakistan does not redress the original instigation of violence – for example, by genuinely seeking to detain and prosecute the perpetrators – pressure mounts for India to demonstrate through force that it will not be deterred from escalating the conflict in self-defense. But this intertwined process of deterrence and compellence is dangerously complicated by uncertainties over the unitariness of Pakistani authority which arises when terrorism or sub-conventional aggression occur.
For example, while India could perceive that the terrorist attacks it attributes to Pakistan signal Pakistani aggressiveness, Pakistani leaders (and the public) could perceive the initial terrorist attacks as a signal that the Pakistani state does not seek a wider conflict but is merely signaling resolve to press India to make political accommodations, in Kashmir or more broadly. This signaling process becomes all the more difficult and precarious if the Pakistani leaders who are presumed to be the authors of Pakistan’s signals and actions deny that the perpetrators of the conflict-triggering violence actually do manifest the policies of the state. Indian leaders then face a highly unstable dilemma. They could act as if the initial violence reflects the intentions of Pakistan’s chain of command, and send countervailing signals of retaliatory action according to normal models of deterrence, in which greater credibility and righteousness tend to reside with the defender.5 But if Pakistani leaders believe or claim that the perpetrators were not carrying out state policies, and India does escalate, Pakistani leaders will feel that India is the aggressor, significantly changing the dynamics of crisis and deterrence stability. “Normal” models of deterrence do not hold in such a situation.
This paper explores the challenge of deterrence stability in the face of real or perceived disunity in the chain of command between top Pakistani authorities and actors who may commit violence against India (or others) of a scale that could lead to inter-state war with potential to escalate to potential use of nuclear weapons. It concludes by suggesting policy approaches that the US and other states might consider toward Pakistan to redress this set of problems.
The 9/11/2001 attacks on the United States elevated the threat of nuclear terrorism to the top of American national security priorities. The fact that Al Qaeda and other extremely violent terrorist organizations trained and operated from Pakistan, and that the most notorious leader of Pakistan’s nuclear weapon program, A.Q. Khan had sold nuclear-weapons know-how and fissile material production capabilities to North Korea, Iran, Libya (and perhaps others), made Pakistan the clearest focus of concern over nuclear terrorism. The priority of preventing Pakistan from being a source of nuclear terrorism soon became reflected even in congressional debates over providing civilian development aid to Pakistan.
This emphasis on nuclear dangers that Pakistan may pose to the US is natural and perhaps unavoidable. Yet, it also may be counter-productive. It reinforces a Pakistani narrative that the US does not care about Pakistan or Pakistanis, but is only interested in Pakistan as a battleground for efforts to counter terrorist threats to the US. The perception that the US only cares about nuclear risks to itself, and not the well-being of Pakistanis, is joined with the perception that the US favors India, and has demonstrated the capability and will to stealthily enter Pakistani territory and carry out daring military operations. Added up, these perceptions make the Pakistani military and media-consuming public feel that the greatest threat to them is a potential US effort to steal or neutralize their nuclear weapons (perhaps in cooperation with India and Israel). Public opinion polls in Pakistan now rank the US as a greater threat than India.
To be sure, the stewards of Pakistan’s nuclear arsenal in the Army’s Strategic Plans Division at the Joint Staff Headquarters, along with other high-level Pakistani policy-makers, also worry about the security of their nuclear arsenal from insiders and/or anti-government groups. They tend to express these concerns privately, however. Their greater and more public priority is to prevent the US and India from knowing enough about the locations and security of this arsenal to be able to target it. This interest can have a benign or even positive effect: some measures the Pakistani military takes to hide and secure the nuclear arsenal from the US are also useful to prevent unauthorized elements within Pakistan from gaining control of fissile materials or weapons. On the other hand, measures to disperse nuclear assets and to prepare for their rapid mobilization in a crisis can make fissile materials and weapons more susceptible to theft or unauthorized use.
When assessing problems and possible solutions, it is useful to ask what are the motivations and capabilities of the relevant individuals and agencies in the state of concern, in this case Pakistan. Nuclear weapons are Pakistan’s crown jewels, the ultimate guarantor of national sovereignty in the view of Pakistani security professionals. The military’s control over the nuclear arsenal also serves its parochial interest of being able to negotiate terms of a future evolution to genuinely civilian-led democracy: the civilians who are nominally at the apex of Pakistan’s National Command Authority would not be able to gain effective command and control over this arsenal without the military’s willing cooperation. (According to the National Command Authority Ordinance of 2007, promulgated by then-President (and General) Pervez Musharraf, the president is the chairman, and the prime minister the vice chairman of the Authority. In 2009, President Zardari ceded the chairmanship to the Prime Minister, a position that has not been held by military leaders in the past. The chiefs of the military services and the director general of the Strategic Plans Division are also members of the Authority. Notwithstanding the law, Army generals have been the head of state throughout half of Pakistan’s history, and in a severe military crisis there would be reason to doubt the actual authority of non-military leaders to determine nuclear moves that could differ from the Army’s preferences).6
In terms of capabilities, the Pakistani military has been more successful securing its nuclear arsenal than it has been in quelling the ongoing insurgency in Balochistan, imposing a peaceful outcome in Afghanistan, or pacifying the tribal areas. Nuclear security can always be improved everywhere, and it is impossible for outsiders to know what are the shortcomings in the Strategic Plans Division’s effectiveness. Still, combining the military’s security capabilities with its clear motive to maintain control over the nuclear arsenal leads to the conclusion that the security of this arsenal is not the most pressing problem that needs to be redressed in Pakistan. To put it another way, the general level of violence, insurrection, and political instability in Pakistan pose the greatest long-run threats to the security of Pakistan’s nuclear arsenal, either from insiders or outsiders, than do specific nuclear security practices of the Pakistani security establishment. Focusing, as many US officials and commentators do, on the presumed insecurity of Pakistan’s nuclear assets distracts attention and effort from the more fundamental objective of redressing militancy and disorder within Pakistan.
The most immediate nuclear risk is that war will occur between Pakistan and India and could escalate to nuclear use. Deterrence stability requires control over the nuclear arsenal and preventing conflict that can escalate – perhaps unintentionally – to major war between India and Pakistan. These two objectives are related, of course. Some of the same groups that threaten Pakistan’s internal security also are most likely to instigate conflict with India. And preventing India-Pakistan nuclear crisis and unintended escalation is related to the objective of securing the arsenal from unauthorized actors: if and when Pakistan were to deploy nuclear weapons for possible use in conflict, these weapons would become most vulnerable to unauthorized use, theft or accident.
Prioritizing prevention of India-Pakistan conflict and augmenting deterrence stability has several implications. It invites a US discourse that is less about saving American lives, and more about preventing massive Pakistani casualties. It acknowledges the India-Pakistan dynamic, which is important to the Pakistani security establishment. The India-Pakistan context points to the need to address the causes of conflict, including Kashmir and Afghanistan. The deterrence stability narrative also does not question the effectiveness of the Strategic Plans Division’s efforts to secure Pakistan’s nuclear deterrent. Rather, it focuses more on the ISI’s need to stop cultivating and tolerating actors whom they perceive to be useful in bleeding India and protecting Islamabad’s interests in Afghanistan, but are increasingly seen as a threat to the unity of Pakistan itself. Concerted efforts to regain the state’s monopoly on the legitimate use of force would not only enhance Pakistan’s overall future as a modern sovereign state; they also would reinforce deterrence stability.
Highlighting the need to prevent India-Pakistan conflict and stabilize deterrence between the two countries acknowledges that Pakistan like India possesses a nuclear deterrent and that the US is not seeking to eliminate it, but rather to encourage the two states to manage it stably.7 The potential political-psychological value of this reframing could be significant. It would replace the two counter-productive frames that have shaped US policy, and Pakistanis’ perceptions of it: first, that Pakistan (and India) should abandon their nuclear weapon programs and join the NPT, and more recently, that Pakistan poses a grave risk of nuclear weapons getting into the hands of terrorists, and that US policy should concentrate on preventing that from happening perhaps by stealing Pakistan’s nuclear weapons.
Another advantage of this frame is that the US can seek to promote India-Pakistan deterrence stability without reference to US operations in Afghanistan and without pressing the Pakistani Army to fight the Haqqani network or conduct military operations in the FATA. Whatever happens in Afghanistan or vis-à-vis the Haqqani network, if Washington prioritizes the stabilization of India-Pakistan deterrence and the rolling up of actors that could instigate India-Pakistan conflict, it would signal an abiding interest in staying constructively involved in the region. The character of this involvement would be much less militaristic and therefore less threatening to Pakistanis than is the case now when the US is seen to concentrate on military operations to achieve its objectives in Afghanistan and perhaps to steal Pakistan’s nuclear assets to prevent anti-American terrorism.
If the US (and anyone else) is to prioritize stabilization of deterrence between India and Pakistan, it will have to confront directly the challenges to the unitary rational actor model on which established theories and policies of deterrence rest.
Deterrence depends on nuclear-armed states approximating unitary rational actors. Indeed, as noted above, the unity of command and authority over action emanating from a state assumed by this model extends to the nuclear domain the classic definition of a modern state as the entity that enjoys the monopoly on the legitimate use of force within a given territory.8 This means that the highest authority of the state must be able to communicate and act on its intentions through a chain of command, so that the state that is to be deterred knows that the signals it is receiving are intended by the signaling state. The sending of deterrent threats and the management of potential crisis escalation, war, and war termination depend on the linear connection between instruments of violence and the state authorities that order them.
The classical texts on nuclear deterrence, including those that welcome the stabilizing effects of gradual proliferation, treat states as the unit of account and assume that state decisions and actions are made by unitary authorities who rationally maximize values. The passage quoted earlier from Bueno de Mesquita and Riker is emblematic; Kenneth Waltz has echoed it in his writings on the stabilizing effects of gradual nuclear proliferation. For example, Waltz argues that “states are not likely to run major risks for minor gains,” and that “nuclear weapons make military miscalculation difficult and politically pertinent prediction easy.”9 The same cannot be said for jihadis, irregular forces, or terrorists who act under varying degrees of control by high state authorities.
The rationality requirement in deterrence is obviously challenged if and when terrorists, nihilists, or ecstatically violent actors are the opponent. A core assumption of rationality in nuclear deterrence is that actors seek self-preservation. Terrorists do not generally fit this model. Nor do state leaders who are unwilling or unable to control terrorists or who are willing to “bring down their own house” if in the process they can destroy their enemies.
To be sure, even the classical model of rationality in nuclear deterrence is more problematic than commonly recognized. Under the classical model of nuclear deterrence leaders of a “rational” state make threats to initiate the use of nuclear war even if this is likely to trigger retaliation that can destroy them. The potential of mutual suicide is posited.10 Indeed, it is precisely the suicidal implication of “going nuclear” and the presumed rational imperative of self-preservation that creates the mutual deterrent effect. That is, each competitor acts as if it is willing to risk suicide, but the combination or interaction of these threats leads to a stable stand off and aversion to war that removes the “need” for them actually to carry out the threat. Yet, if there were not a willingness to risk actually engaging in suicide, the deterrent effect would be attenuated or lost. Deterrence depends on the risk of its failure; that is, that the other side may initiate armed attack and nuclear weapons will then be used, perhaps in mutually devastating exchanges. Whether it is rational to run this risk can be debated, but states that engage in nuclear deterrence owe it to their citizens and the world community to exert the greatest possible efforts to manage their own affairs and their external relations in ways that minimize the probabilities of armed conflict and its escalation to nuclear use. Indeed, this is a norm that internal and international politics seek to hold states responsible for fulfilling.
Carrying the discussion of rationality further, the willingness to risk a break down in nuclear deterrence would only be rational if the threat that is being countered or deterred is of an existential scale. To risk suicide to redress a threat that is not itself mortal would be irrational, because if escalation to nuclear exchanges resulted, it would leave the state taking this risk worse off than it would be if it did not trigger such a nuclear exchange.11 Retaliatory (as distinct from initial) nuclear use would be less irrational insofar as the initiator of nuclear use already would have demonstrated that the existence of the attacked state was unmistakably in jeopardy. Then the dilemma would be whether to forbear from retaliation in order to keep the level of destruction from being total, or to retaliate in order to demonstrate one’s own credibility and retain leverage for war termination.
Irrationality cannot be excluded from a state conforming to the unitary actor model, but irrationality is much more probable in non-unitary state. Rationality aside, disunity produces dangerous confusion and ambiguity that interfere in the management of deterrence. Who is sending signals through violence that is perceived to be emanating from the state and/or its territory? What is being signaled? To whom does the victim of aggression address counter-signals and actions? How does one calculate the interests of the attacker if the putative leaders of the state are not the authors of the perpetrated violence, or are pretending not to be? If states and their leaders are presumed to be rational and to highly value the preservation of their state, but disunity exists and other actors who may not place a high value on preservation of the state are conducting aggression and may be able to gain control over nuclear assets, how does one manage deterrence and escalation processes in such a situation? In this latter scenario, disunity erodes the rationality on which deterrence is predicated.
The outcomes of confusion prompted by violence whose authorization is ambiguous could be either stabilizing or destabilizing. If, for example, authorities in India believe that authority in Pakistan is splintered, New Delhi may choose to refrain from counter-attack, which could reinforce stability and escalation control. Or they could respond with force, risking unintended escalation. However, because leaders of adversarial states must put any immediate conflict in the context of longer-term relations, they must think how their action or inaction today could raise or lower the risk of inviting more violence from the adversary in the future. Thus, forbearing counter-attack in one crisis, can be seen to weaken deterrence of future violence. With each additional crisis that is not met with counter-attack, the pressure mounts to act more forcefully the next time in order to restore deterrence.
Pakistan illustrates the unity problem and its implications for rationality more acutely than any other nuclear-armed state. Early in its history, Pakistan’s military leadership saw the potential value in mobilizing “irregular” forces to augment the regular Army in pursuing the state’s objectives vis-à-vis India. These objectives most famously have included wresting the Kashmir Valley from Indian control, and/or raising the costs of India’s ongoing occupation of the Valley and diverting India forces to the occupation. In the 1947 war initiated by Pakistan authorities falsely claimed that the violence was perpetrated by irregulars outside of state control. In 1971, the Pakistan Army mobilized proxy forces in East Pakistan to combat the Bengali separatist insurgency, which was backed by India.12 These proxies, who were mostly Urdu-speaking Biharis and activists from the Jamaat-e-Islami party, massacred large numbers of Bengalis. In the 1980s, Pakistan and the United States (with help from Saudi Arabia and others) famously mobilized irregular “freedom fighters” to drive the Soviet Army from Afghanistan. After that successful mission, the Pakistani military and intelligence services shifted these forces to Kashmir, exploiting the opening created by India’s rigging of state elections there in 1989. Pakistan nurtured and abetted the growth of jihadi organizations to carry on the struggle with India. Over time, these groups proliferated. There is room to debate the degree to which Pakistani authorities controlled the growth and operations of these groups, but through the present era Pakistani authorities have not stifled them decisively.
The United States and the Soviet Union during the Cold War used proxies in violent struggles in the Third World, but, importantly, did not extend violence directly into each other’s homeland. There are several reasons for this, but one was recognition that attacks in the Soviet or US homeland could escalate to nuclear war. To preserve nuclear deterrence and avoid nuclear war, the two antagonists eschewed the projection of violence into each other’s territory. This does not imply that the US or the Soviet Union was acting from virtue; their proxy wars killed many more people than would have been killed if they had held back. Rather, the point is that the projection of violent actors by Pakistan into India or vice versa poses unprecedented risks of deterrence instability.
On October 1, 2001, terrorists rammed a truck filled with explosives into the gate of the Jammu and Kashmir legislative assembly complex in Srinagar, killing thirty-eight people. The Pakistani-based militant group, Jaish-e-Mohammed at first claimed responsibility for this attack, but later denied it. Jaish-e-Mohammed activists also live in India. Two months later, on December 13, a handful of militants attacked the Indian parliament building in New Delhi, resulting in the deaths of seven guards and five attackers. Indian authorities quickly blamed the Jaish-e-Mohammed and the Lashkar-e-Toiba for the assault, and within days arrested four Indian Kashmiris as conspirators.13 Pakistan condemned the attack immediately. Subsequently, the government of Pakistan contested Indian allegations that the attackers were linked to Pakistan. As recriminations flew across the border, India mounted a massive military mobilization toward the India-Pakistan border. This force eventually totaled 500,000 troops, backed by naval forces shifted from the Bay of Bengal to the northern Arabian Sea to join the Western Fleet for a blockade of Pakistan.
In explaining this mobilization, which itself resulted in nearly 800 Indian fatalities, India’s Ministry of Defense declared that “Pakistani provocation reached a dangerous point with the December 13 attack on the Parliament. A more forceful response became necessary.”14 By raising the threat of major conventional war, which clearly would have been of a scale that could put the burden of nuclear escalation on Pakistan, India sought to compel Pakistan to take clear measures to curtail the operations of violent actors with ties to Pakistan. This reflected the escalatory continuum between subconventional, conventional, and nuclear competition in South Asia. The crisis also exposed the related difficulty of simultaneously pursuing deterrence and compellence: India sought to compel Pakistan to crack down on terrorist groups and to deter it from facilitating terrorist attacks in the future.
Pakistan’s leadership responded by denying complicity in the initial attack and by portraying India as the aggressor for having mobilized its conventional forces in such an escalatory way. At the same time, Pakistan’s president and chief of army staff, Pervez Musharraf pledged that no organization would be allowed to indulge in terrorism in the name of Kashmir, and that organizations would be barred from using militant names such as Jaish (army), Lashkar (volunteer force), or Sipah (soldier).15 Musharraf announced 10 further that he had banned five Pakistan-based militant organizations, arrested their leaders, frozen their assets, and locked up their offices. He called on India to end its own “state terrorism” and to undertake dialogue on Kashmir.
As the risk of war mounted, American diplomats exerted great effort behind the scenes and publicly to encourage both states to find a diplomatic resolution. To complicate matters further, and to demonstrate the complexity of both sides’ grievances, devastating riots erupted in the Indian state of Gujarat in April 2002, leaving more than 1,000 Muslims dead. There was evidence that the BJP government of Gujarat and in New Delhi were complicit in failing to prevent or limit the carnage. Musharraf’s January pledge to disband militant organizations and arrest their leaders was now called into question as prominent jihadis were released and supposedly-disbanded organizations re-emerged under different names. After US remonstrations, Musharraf in late May reasserted pressure on militant organizations. However, cross-border terrorism continued in the prelude to Kashmir elections. This violence paradoxically was made easier by the deployment of Indian forces away from Kashmir to the central India-Pakistan border.
The on-again off-again crack down on jihadi organizations in Pakistan, and the ongoing cross-border infiltrations of militants from Pakistan into Kashmir, raised questions about the unity of authority in Pakistan. Was President Musharraf being duplicitous? Was he changing his mind? Was he capable of ensuring that his decisions to disband terrorist groups were enacted down through the ranks of the Army’s intelligence service and the state’s security apparatus which had tolerated if not nurtured the jihadi organizations that continued to operate in Pakistan despite his pledges? These questions indicate the real or perceived, or purposely projected, erosion of top-down control by Pakistani leaders over actors capable of starting war.
By the time US Deputy Secretary of State Richard Armitage visited the region in June 2002, it was clear that India was not going to unleash its force. The crisis formally ended in October when India announced a demobilization.
The 2001-02 crisis, which followed only two years after the Kargil conflict that had been initiated by Pakistan, motivated the Indian military to seek a new military doctrine and capabilities to deter Pakistan from undertaking similar low-intensity aggression in the future. Indian security officials and experts conceptualized the challenge in terms of deterrence, especially nuclear deterrence. They perceived that Pakistan was emboldened by the possession of nuclear weapons to believe that India would not risk escalation in response to terrorist attacks or irregular-force incursions in India. Therefore nuclear deterrence was providing Pakistan a shield behind which it could stab India at will. India thus needed to deploy capabilities that could allow it to retaliate to low-level attacks emanating from Pakistan with conventional military operations that would be limited enough to deter Pakistan from running the risk of counter-escalation that could rise to the nuclear level. The existence of such capabilities and related operational doctrine could also compel Pakistan to resolutely curtail the recruitment, training and projection of militants into India.
Of course, developing such capabilities and using them precisely enough not to trigger nuclear war is easier said than done, and is inherently risky. The calibration of escalation is exceedingly tricky and requires unattainable confidence in one’s knowledge of the opponent’s red lines – what level of conventional retaliation would the adversary be willing to withstand without countering with greater escalation, including the use of nuclear weapons?
The calibration and management of such deterrence signaling and military operations is all the more difficult when the competitor, Pakistan, denies that the perpetrators of attacks are carrying out orders of the state. The situation is complicated further in Pakistan when the nominal leaders of the state – the president and prime minister – are civilians, but the dominant actors are military. This troublesome dynamic could be seen early in the Zardari government’s tenure when it tried to impose civilian control over the ISI, but was compelled within days to reverse itself. In such cases where formal authority is contradicted by the realities of power, it is difficult to ascertain at any given time what the civilians actually know, and therefore whether what they say is true when they deny state authorship of an attack. With whom should India communicate? New Delhi’s leadership may think it most relevant to address the military directly, but India cannot choose whom Pakistani authorities designate to manage communication and signaling. And if there is long-term value in promoting civilian authority in Pakistan, seeking to communicate directly with the military may be counter-productive. An additional asymmetry further complicates communication and potential confidence-building: military-to-military dialogue between India and Pakistan is difficult because the civilian leadership in Delhi will not cede this authority, and even if it did, the Indian military could not communicate on a peer basis with their Pakistani counterparts because they have less information and authority.
Unfortunately, the 2001-02 crisis was not the last of its kind. On November 26, 2008, terrorists launched eleven coordinated shooting and bombing attacks in Mumbai. By the end of the assault three days later, 164 people had been killed and more than 300 others wounded. The attacks were conducted by ten young men trained and orchestrated by Lashkar-e-Taiba from Pakistan, despite the organization’s nominal ban announced by then-President Pervez Musharraf in 2002. Pakistan, the United States, India, and Italy subsequently arrested other suspected participants in the conspiracy.
Repeating earlier patterns from the 1999 Kargil Conflict and the 2001-02 attacks in Srinagar and New Delhi, Pakistani authorities at first denied that the attackers were Pakistani and that Pakistani territory had been used to train and orchestrate them. Investigations by the US and India, including interrogations of the Pakistani-American David Headley, indicated that elements of the Pakistani ISI had facilitated the planning and operation of the terrorists. In slightly more than a month, Pakistani authorities acknowledged that its nationals had been involved in the attacks and that training and planning had been conducted in Pakistan. In November 2009, a Pakistani court charged seven citizens for their role in the terrorist assault, including the operations commander of Laskkar-e-Taiba. Prosecution has been halting and the leader of LeT has been released from a deferential form of detention.
Perhaps mindful of the earlier experience of the 2001-02 crisis, Pakistan and India, with encouragement from the United States, took pains to avoid repeating the cycle of recriminations and mobilization of large conventional forces. The horrendous gravity of the attacks, and the clearer evidence of the perpetrators’ origins in Pakistan, chastened Pakistanis, including their recently elected civilian leadership. Still, there was at least one episode within the crisis that demonstrated the precariousness of its management, as documented by a recent Stimson Center report. Shortly after the attacks in Mumbai, President Zardari received a phone call from a man claiming to be Indian Foreign Minister Pranab Mukherjee, warning that India would launch a war on Pakistan the next day. Such a threat would naturally prompt the Pakistani military to increase its readiness, if not to launch pre-emptive operations. Fortunately, American officials learned of the alleged call and quickly arranged for Secretary of State Condoleeza Rice to phone Mukherjee to ascertain what was happening. Meanwhile, the Deputy Chief of the US mission in Islamabad, Jerry Feierstein, who was in charge in Ambassador Anne Patterson’s absence, was alerted that Washington knew about the call and was investigating it. Rice reached Mukherjee in Calcutta and he denied having called Zardari. Further investigation concluded that it had been a prank call.
While Pakistan and India managed to prevent the 2001 and 2008 subconventional attacks on high-value Indian targets from escalating into inter-state war, these episodes and the ongoing operation of jihadi organizations in Pakistan highlight the grave instability of India-Pakistan relations. Doubts about the monopoly of the Pakistani state’s control of forceful actors and actions emanating from the state challenge the unitary rational actor assumptions on which deterrence rests. The precariousness of deterrence stability grows with each crisis. Each presents a pressure on India to counter-attack, and then on Pakistan to respond in an escalatory action. And each time India refrains, the pressure grows on it not to do so the next time, in order to demonstrate that Pakistan cannot continue to get away either with instigating attack or not acting decisively against those who carry out such attacks.
The implications are profound and have not been adequately addressed by Pakistan, by India, nor by international strategic analysts. In some ways, the operation by ostensibly uncontrolled violent groups acting from Pakistani territory is a violation of Pakistani sovereignty, much like Osama bin Laden’s long presence in Pakistan was. In the latter case, the Pakistani security establishment chose to focus on the United States’ violation of Pakistani sovereignty in raiding Osama’s compound, but in subsequent debates increasing numbers of Pakistanis acknowledge that bin Laden’s presence also was an embarrassment. When Pakistani authorities deny that terrorists known to be based in Pakistan have anything to do with state agencies or objectives, they implicitly are acknowledging that state sovereignty is woefully incomplete, or they are lying. When they then do not act systematically to curtail the actions of such organizations after they have attacked India (or the US) they reinforce the perception of duplicity or of ineffective sovereignty.
Pakistani security officials naturally are loathe to admit these implications out of pride and fear that the military’s standing and competence would be called into question. Yet, the risks that subconventional uses of force could escalate to conventional and perhaps nuclear war creates a clear interest for Pakistanis, Indians, and the international community to treat the uncertain quality of Pakistani state sovereignty as a fundamental strategic problem. The lack of coherence of authority in Pakistan – and of the state’s monopoly on the legitimate use of force – calls into question the integrity of Pakistan’s chain of command. Or, if the military does in fact retain control over organized perpetrators of violence, its ongoing support or toleration of them poses a grave threat to deterrence stability.
If in fact Pakistani authorities do exert influence or control over organizations that have conducted terrorist operations in India (and elsewhere) and are merely denying it for tactical reasons, Pakistani authorities can fairly be treated as the authors of the signals that are sent by these actors. India can then seek to manage pre-conflict and intra-conflict deterrence according to the traditional model, while still facing severe challenges of escalation control. The challenge certainly is complicated by the ambiguity that Pakistan would be seeking to create by denying responsibility for future sub-conventional operations; deterrence stability would be weaker than in “normal” scenarios. But if India were correct in acting on the belief that Pakistani authorities were culpable for such operations, the burden of having stepped first on the escalation ladder would be on Pakistan, and the defender’s advantage in deterrence would accrue to India.
The graver problem arises if and when Pakistani leaders are not in control of the perpetrators of violence emanating from Pakistani territory. In that case, when faced with an attack, India would conclude that deterrence had failed or was inapplicable, but that if India did not retaliate, it would encourage further attacks and do nothing to compel Pakistani leaders to assert control over violent actors. But if India did retaliate, Pakistani leaders, feeling that they had not authorized aggression against India, would feel that India was initiating war. It is widely recognized that victims of aggression – defenders – are more highly motivated to retaliate because they have suffered an injustice. Knowing this, Pakistani defenders would feel that their threats to escalate in response to an Indian attack would be more credible than if they had been the initiators of the conflict. Indians, of course, would feel that this logic rewards Pakistani authorities for not exercising a monopoly on the legitimate use of force emanating from their territory, precisely the situation they want Pakistan to correct.
Perhaps because the classical model of nuclear deterrence was developed in the context of the bipolar US-Soviet competition in which neither state deployed proxies to commit terrorist or sub-conventional attacks on the other’s homeland, the model is relatively silent on how such low-intensity aggression could or should be deterred. The literature’s silence on this problem, and Pakistan’s self-interested belief that nuclear deterrence should be applied only to the conventional-nuclear threat spectrum, further distract analysts and policy-makers from wrestling with the problems that sub-conventional violence pose to deterrence stability. Yet, as the psychologist Steven Pinker has elaborated in his masterly book, The Better Angels of Our Nature, “the necessity of vengeful punishment as a deterrent…has been demonstrated repeatedly in mathematical and computer models of the evolution of cooperation.” 16 Retaliation “is necessary for cooperation, [for] preventing a nice guy from being exploited.” 17 If one does not retaliate, it may invite further aggression.
Of course, if competitors merely engage in a cycle of action and retaliation, both end up losing over time. Survival and betterment require someone to break the cycle. The victim of aggression may retaliate and at the same time signal an interest in returning to cooperation in which both sides refrain from further violence. Psychological experiments and computer modeling indicate that the cycle can be broken by “random” grants of forgiveness of an aggression and invitation to cooperate. To make this strategy work over time, of course, the actor who has been forgiven a transgression should display contrition for the transgressive act. Otherwise, the forgiver can be perceived embarrassingly as a sucker, which can trigger powerful emotions of vengefulness.
It is possible that Indian leaders – at least the two most recent, Atal Bihari Vajpayee and Manmohan Singh – have concluded that India’s greatest strategic priority is steady economic growth and development, and that avoiding provocations to engage Pakistan in war is necessary to achieve this higher good. Moreover, such forbearance of war could even serve strategic purposes vis-à-vis Pakistan, much as Gandhi’s strategy of nonviolence ultimately caused the United Kingdom to leave India. That is, each act of terrorist violence projected from Pakistan into India, matched by Indian forbearance, further weakens Pakistan’s international standing and increases India’s moral-political leverage in pressing Pakistani authorities to act decisively against the perpetrators of cross-border terrorism. The more Pakistan looks like an outlaw to the international community and to its own citizens, and the more stark Pakistan’s failure to fulfill international norms against terrorism appears, the greater the chance that Pakistani citizens will eventually put pressure on the security establishment to change its behavior.
This dynamic has been on display in the post-Mumbai relationship between India and Pakistan. India suffered the attack, but Prime Minister Singh resisted pressures to retaliate and, in essence, “forgave” Pakistan in order to avoid an escalatory cycle of retaliation. Pakistani leaders, half-heartedly and ambiguously, signaled regret for the terrorist attacks (while denying direct responsibility for them). The opportunity has been created for cooperation to stabilize the situation. No matter how justified Pakistanis might feel for the acts of jihadis operating in India, and how much they feel India should “get over it,” they would be wise to recognize that India’s interests in reinforcing deterrence, in not looking like suckers, and in satisfying emotional impulses for revenge make the situation quite precarious in the event of another terrorist attack emanating from Pakistan. Today’s Indian leaders may have strategic reasons for forbearance, but this would not guarantee the persistence of such a policy into the future.
The safest way to defuse this unstable competition and reinforce deterrence stability, of course, is for Pakistan to make unambiguous efforts to restore the monopoly on the legitimate use of force that is central to modern statehood. Achieving this objective will be difficult at two levels. In terms of motivation, if Pakistani military and intelligence authorities see some advantage in claiming that they do not control jihadi organizations, they will be reluctant to exert themselves fully to exercise control. In terms of capabilities, it is no doubt difficult for Pakistani authorities to demobilize and contain all of the militant organizations that have evolved in the country since the 1980s. Moreover, if they decided to do so without careful preparations and public education, they would prompt attacks against the security establishment itself, as happened after the operation against the Red Mosque in Islamabad in 2007. Fortunately, Indian (and American) officials recognize the latter problem. What they most want is for Pakistani leaders to demonstrate not only in words but also in constant deeds a determination to delegitimize violence against India and arrest and prosecute actors who violate the law. Perfection in accomplishing this objective would not be expected, but clear and uncompromising effort would be.
There is a political logic for holding the Pakistani state accountable for acting to prevent the projection of terrorism. Political change requires time and a confluence of drivers. It tends to occur when existing practices violate norms whose salience are increasing and groups within a society mobilize to demand the abandonment of delegitimized practices. In simplistic terms this is how witch-burning and slavery were ended in the West and how the Muslim Brotherhood and Fatah abandoned terrorism in Egypt and Palestine. The international norm against the export of terrorism has grown since 2001. It is now reflected in legally binding international law. In the words of a recent Cornell International Law Journal article:
A state is required to prevent extra-state forces, which engage in hostile acts towards other states, from operating within its borders. In particular, the United Nations Security Council has said that ‘every state has the duty to refrain from…acquiescing in activities within its territory directed towards the commission of such acts, when such acts involve a threat or use of force.’ 18
UN Security Council Resolution 1373, of 2001, makes this obligation legally binding under Chapter VII of the UN Charter. Of course such laws are not self-enforcing and there is not a world government to enforce them. But the laws and norms against terrorism may provide a basis for civil society and political leaders in Pakistan to press the security establishment to act more decisively against those whose external violence could lead to nuclear war.
There are some recent indications that key Pakistani authorities are joining modern elements in civil society in recognizing the intolerable effects of violent militancy within Pakistan and directed at India. Whereas five years ago visitors to Pakistan would hear even well-educated counterparts deflect attention from the internal and external militancy problem by blaming the US or India for it, today the same circles volunteer that distinctions between “good” jihadis and “bad” jihadis are self-defeating. Militants are destroying Pakistan from within, and terrorism against India has harmed Pakistan severely. This view also is now inculcated in National Defense University courses for Pakistani officers.
Most remarkably, perhaps, Chief of Army Staff General Pervez Kayani devoted his August 14, 2012 speech on Independence Day to the imperative of combating “extremism and terrorism.”19 It is extremism when one tries to impose one’s opinion on others, he continued, and “if one tries to enforce his opinion through use of gun, it becomes terrorism.” “If this is the correct definition of extremism and terrorism,” Kayani said, “then the war against it is our own war, and a just war too. Any misgivings in this regards can divide us internally, leading to a civil war situation.” This clear expression of national interest and priority was all the more remarkable insofar as General Kayani did not “balance” it with excuses or bellicosity relating to India. The only oblique reference to outside powers was that “the forces hostile to our motherland are benefitting from internal weaknesses.” And here, Kayani said, “Blaming such anti-Pakistan elements aside, our efforts must be directed towards stabilizing the internal front.”
Stabilizing India-Pakistan deterrence and encouraging Pakistani security agencies to reassert the state’s monopoly on legitimate instruments of violence is a necessary condition for saving Pakistani lives and creating conditions for peaceful democratic politics as well as for deterrence stability. Indeed, it is necessary if Pakistan is to regain this central attribute of a modern sovereign state. By framing its objectives this way, the US could begin a long-term process of rebuilding public support for its role in Pakistan.20
Framing is vitally important. The troubled political-psychology of US relations with Pakistan has enormous material implications for both states. To restore their political will to pursue shared interests it is necessary to build a shared understanding of each other’s intentions to cooperate in at least some key areas, recognizing that interests diverge in others. Framing can be dismissed as too easy or too conceptual to be a serious policy recommendation. Yet, “merely” changing the frame and discourse of US-Pakistan relations is extremely difficult. This is clear to participants in the interagency process in the executive branch and even more so with Congress in recent years as US officials have struggled to reconceptualize US policy in recognition that Pakistan does not share US interests regarding Afghanistan. The difficulty of changing frames is also evident in civil-military dynamics in Pakistan, where no consensus exists on the goals and scope of US-Pakistan relations. This can be seen when Pakistani military and civilian leaders diverge over when and how to blame unpopular events or policy choices on the US, even in cases where the central cause of the difficulty is internal. Sometimes the military blames the US for speaking and acting in ways urged by Pakistani civilians, while at other times civilians blame the US for cooperating too closely with the military.
Within the large frame of deterrence stability, a vital subsidiary concept is that a state cannot be a responsible possessor of nuclear weapons if it does not have sovereign control over organized perpetrators of international violence operating from its territory. The absence of such sovereign control impedes efforts by state authorities to ensure national preservation and minimize risks of escalatory conflict that risk annihilation. To put it colloquially, US officials could say to Pakistanis, “We do not challenge your possession of nuclear weapons. Our objective is to promote in any way we can the responsible management of nuclear forces. First and foremost, this means sovereign control over all organizations that can project violence from your territory which is also an obligation under international law. Second, and relatedly, it means you should not tolerate acts that could start wars with other nuclear-armed states, because that would be suicidal and therefore irrational. Given the global implications of nuclear war and the breaking of the nuclear taboo, all states have a shared interest in Pakistan’s coherence, sovereignty, and responsible nuclear stewardship.”
With shared interests defined this way, it follows that the US would naturally offer Pakistan, as requested, assistance to help responsible state agencies to control actors that could challenge the state’s monopoly on the legitimate use of force, both internally and outside Pakistan’s borders. To the extent that the police in Pakistan can and should play a more effective role in this mission, the US could re-emphasize willingness to provide training and equipment to them, if this would be welcomed.
More problematic is counter-terrorism cooperation. Such cooperation has been sharply curtailed since the raid on Osama bin Laden and the November US/NATO killing of 24 Pakistani servicemen on the Afghanistan-Pakistan border. Both of these trust-destroying episodes highlighted the divergence in American and Pakistani priorities regarding which “terrorists” should be priority targets, and by what means. Any future restoration of cooperation would depend on clearer agreement on who is to be targeted and why. The proposal here is to explore potential agreement in prioritizing cooperation in arresting or, if that is impossible, killing actors who conduct violence within Pakistan and also against India.
The issue of US drone strikes on violent militants in the Federally Administered Tribal Areas (FATA) of Pakistan severely complicates US-Pakistan relations. A full treatment of this issue is beyond the scope of this essay, but a few relevant points can be made. The US military, and more, the CIA focus narrowly on their objectives of finding and killing “terrorists” who threaten allied forces and objectives in Pakistan, and occasionally, those who wage violence in Pakistan. They tend to leave it to others to manage the backlash this creates in Pakistan against the US more broadly. Pakistani authorities have at various times secretly consented to and cooperated in targeting drones against Al Qaida and anti-Pakistan elements, while publicly blaming the US for such attacks. Both sides – the US and Pakistan – have failed to explain and conduct drone 18 operations in ways that could challenge reasonable Pakistanis to confront the horrible dilemmas involved.
Violent extremists from Central Asia, China, and the Arab world have found refuge in the FATA for planning, preparing and conducting attacks in Afghanistan, as have groups associated with the Pakistani Taleban and groups such as Lashkar-e-Taiba who attack the Pakistani state and its innocent civilians. The Pakistani state does not exercise effective sovereignty over this territory. It does not effectively curtail the operations of these violent actors, even those who clearly threaten Pakistan. This poses the central dilemma: if these foreign and Pakistani actors will continue to act violently against Afghanistan and international forces working to defend it, and also against Pakistan unless they are physically stopped from doing so, and if the Pakistani state cannot stop them, then US drone-strikes may offer the least-damaging means to do so. Clearly it would be better if these groups would desist from violence and if Pakistan could impose its sovereign control over them and the territory where they operate. If ceasing drone strikes would help achieve these results, then it would be wise to do so. But in the absence of evidence that these violent actors would stand down if drone strikes stopped, continuing strikes appears to be the only course available. In this case, the question is whether and how the US and Pakistan could alter the conduct and public presentation of drone operations to minimize the public backlash against them. It is worth debating whether framing these dilemmas in terms of achieving Pakistani state sovereignty over its full territory would help.
The Pakistani security establishment’s wariness towards offers of assistance in counter-terrorism and policing should not be under-estimated. The torturously complex motivations and perceptions surrounding US and Pakistani policies towards Al Qaeda, the Haqqani Network, the Pakistan Taleban, and Lashkar-e-Taiba will not be reconciled easily or completely. In recent months, Pakistan seems to have prevailed in holding the Haqqani network off-limits for attack, while still tolerating US strikes on Al-Qaeda leaders and other foreign fighters. Lashkar-e-Taiba is perhaps the most important organization from the standpoint of instigating conflict with India, and it is difficult to imagine Pakistani authorities cooperating with the US to target it. The group enjoys some public support and is dispersed throughout the Punjab. LeT has the will and the capacity to retaliate violently against the Pakistani security establishment. It may be too strong to challenge frontally today. Nevertheless, the concept of focusing on the restoration of the state’s sovereign monopoly on the legitimate use of force offers a basis for beginning a process of reconciliation between the US and Pakistan. The US could say to Pakistani counterparts, “if you share this fundamental objective, then you tell us what would be a feasible strategy over time to pursue it, and what, if anything, we can do to assist, and what would be reasonable metrics to evaluate progress. And if there are actions we think you should take but you think are unnecessary, premature, or infeasible for the overall objective defined this way, explain why.”
In any such discussion, Pakistani counterparts will ask, “what about India? India is a threat; you are helping to make it stronger; the Indians are simply trying to wish away the Kashmir problem without meeting our fundamental requirements.” The US cannot pretend to be able to compel India to meet Pakistan’s demands vis-à-vis Kashmir or to 19 desist from building up its military capabilities as its economic growth allows and as China’s power increases. But there are some steps the US can take.
India-Pakistan deterrence will not be stabilized if Pakistan’s relationship with India is not normalized and pacified. Nor is Pakistan likely to get on track to reform itself if it is perpetually at loggerheads with India. More normal civil-military relations within Pakistan – which are essential for state coherence – depend on more normal relations between Pakistan and India. Otherwise the Army will resist initiatives to reduce its role in shaping not only national security and foreign policy, but also in the political economy. The sense of threat from India justifies Army predominance within Pakistan, and the Army’s obsession on India keeps the sense of threat high. But history and institutional interests suggest that greater civilian ascendency in Pakistan will be necessary to transform the India-Pakistan relationship. Improved economic prospects within Pakistan, which also are necessary to augment state coherence, will be inherently limited if India-Pakistan trade and investment do not grow.
The challenge is enormous, obviously, but it is not impossible due to the vital fact that India does not harbor offensive intentions toward Pakistan. India does not covet territory that Pakistan controls. India does not wish for Pakistan to be dismembered. Indian leaders recognize that it is in their country’s interest for Pakistan to develop economically, to democratize politically, and to live in peace. India does not want Pakistan’s problems to spill over into its territory or restive Muslim populations. The two countries diverge in their visions for an ideal political outcome in Afghanistan, but could settle for an Afghan state that does not allow itself to be a base for hostile actions against Pakistan and India. The fundamental point is that India will not be a military or security threat to Pakistan if Pakistan will cease pursuing offensive strategies (albeit of a low-intensity nature) against it.
Of course, Pakistani military leaders continue to view India as a mortal threat and to justify the Army’s financial and political privileges on its role as protector of the nation against this threat. Authorities in Rawalpindi and Islamabad bristle at Washington’s contentions that India does not harbor offensive intentions. It will take much time and effort for India and the US to shift the Pakistani security establishment’s mindset. Much will depend on how Afghanistan evolves and the roles that the US and India play in this. Yet, Washington and New Delhi can repeatedly remind Pakistani counterparts that India did demonstrate its defensive intentions and remarkable restraint in the aftermath of the 2001-02 crisis and 2008 Mumbai attack. In each of these instances, Washington exerted its influence to encourage Indian restraint. These cases demonstrated that India’s and America’s interests in India’s ongoing development and accretions of power should not come at Pakistan’s expense.
The US has a clear interest in continuing to demonstrate that India does not and will not pose an offensive threat to Pakistan. India’s plans and wherewithal to augment its conventional military power through arms imports, including from the United States, can alarm Pakistan. Knowing this, Washington and India would be wise to minimize Pakistani alarm by taking pains to demonstrate that their defense cooperation will not challenge Pakistan’s capacity to deter and defend against Indian forces. All three states have an interest in deterrence stability. Shaping and communicating US-Indian cooperation in terms of deterrence stability can reinforce this. To make their intentions credible, the US and India could describe what sorts of Indian capabilities would undermine deterrence stability, and vow not to undertake them if Pakistan demonstrates its commitment to deterrence stability by acting to control organizations that project force from Pakistan against India. In other words, if Pakistani authorities make fulsome efforts to impose a monopoly on the legitimate use of force from Pakistani territory, India and the US will be more forthcoming in policies that could reinforce deterrence stability. India would not then need capabilities and plans (which Pakistanis view as threatening) to compel Pakistan to disable terrorists groups. Indeed, deterrence is most stable when competitors pursue defensive strategies and postures, which should be a clearer objective of US interactions with India and Pakistan, and of India’s and Pakistan’s relations with each other. This could include US arms sales to India.
India-US cooperation in developing ballistic missile defenses poses a particular challenge. There is considerable evidence that offensive forces can always be expanded and adapted to defeat defenses,21 and that developing even theoretically effective defenses requires a web of super-sophisticated and expensive technologies beyond India’s technical and budgetary reach. Recent reports by the US Defense Science Board and the National Academy of Sciences indicate that the US after decades of effort and expense still does not possess anything like an effective national missile defense system. Nevertheless, India’s defense science establishment claims to be on the way to developing missile defenses, and some within the US would like to contribute (for a price) to this effort.
If and when the US (and Israel) were to join forces with India in developing missile defenses with a theoretical capability to intercept Pakistani missiles, Pakistanis will naturally feel that the deterrent relationship is being destabilized. The most knowledgeable leaders of the Pakistani nuclear establishment volunteer in private that such defenses will not work and can be readily circumvented by Pakistani cruise missiles and other means.22 Moreover, India could limit the technological ambition and cost of missile defenses by seeking to defend only its national command and control apparatus in New Delhi. Were such limited defenses effective, they could reinforce India’s confidence in its No First Use policy, which in turn enhances deterrence stability. Indians could then more readily believe in their capacity to retaliate after a Pakistani first strike, obviating the need for India to reverse its No First Use policy. The limited scope of such a defense could, in turn, reassure Pakistani leaders that India is not seeking to negate Pakistan’s nuclear deterrent, but rather to protect the Indian leadership against a very limited decapitation strike or an unauthorized nuclear attack by a rogue military unit.
Leaving aside whether even a limited missile defense would be effective and affordable, deploying any form of missile defense that would be highly touted in India (which would be necessary to motivate spending on it) would reanimate a broader Pakistani narrative that the US and India are colluding to take away Pakistan’s deterrent. Such a development would not augment deterrence stability. (Unless there was unprecedented improvement in India-Pakistan relations at the highest levels, which itself would be made less likely by Indian-American collusion on missile defense). Most likely missile defenses in India would exacerbate arms race instability.
To further the general objective of deterrence stability, the US can encourage and, where welcome, facilitate official or unofficial India-Pakistan dialogue on the subject. The Indian and Pakistani strategic communities of retired diplomats, military officers, and long-time defense analysts have met for years in countless Track II and Track I.5 dialogues. The foreign secretaries of the two countries have met occasionally, as have government-appointed back channels. But deterrence stability has not been a central topic of such discussions. Nor have active-duty military officers participated in these forums. It is worth considering whether and how the two governments could be persuaded to task relevant military officers and security professionals to conduct private workshops on deterrence stability. It would behoove the US government to task its own strategic and military experts who work on South Asia to analyze the deterrence stability challenge and offer to share this analysis if Pakistanis and Indians would find it helpful.
Deterrence stability in South Asia presents unique analytic difficulties as well as political-diplomatic obstacles. The “classical” models of deterrence on which most experts have been trained derive from the bi-polar competitions between the US and the Soviet Union and more recently the US and Russia and the US and China. Modeling explores the interaction between conventional balances and operations on the initial spectrum of the escalation dynamic, transitioning to nuclear use and intra-war deterrence of escalation. But the classical models do not account for the dynamic between India and Pakistan where conflict may start through subconventional use of force and then transition up through conventional to nuclear war. The subconventional-conventional-nuclear escalation problem is significantly more complicated for reasons suggested above. It is especially problematic when subconventional actors are not operating – in fact, or according to Pakistani claims – within the unitary state apparatus on which deterrence models depend. This is primarily a problem for Pakistanis and Indians to address, but given the global interests and responsibilities of the US, there is reason for American officials to mobilize talented, regionally knowledgeable personnel to analyze the subconventional-conventional-nuclear deterrence challenge as well.
To the extent that arms control can be useful to manage deterrence stability, another new and difficult challenge emerges. Nuclear arms control is a well-developed field due to the decades of US and Soviet/Russian experience. Conventional arms control is a less-understood and practiced field. The interactions between conventional and nuclear arms control is even less-analyzed or experienced. When the challenge of verifying efforts to constrain subconventional violent actors is added to the problem, there is no historical or analytic work to draw upon. Even if Pakistan were prepared to commit itself genuinely to make all feasible efforts to curtail the actions of groups that seek to export violence, what are reasonable ways to verify this commitment? This is extremely important insofar as it is possible that even in the midst of concerted state efforts to “disarm” militant actors some may persist and carry out attacks. To maintain stability, the recipient of such attacks (and the international community) would need some basis for judging that the attacks did not reflect the intention of the state whence the attackers originated. Indeed, Pakistani efforts to curtail the operations of violent actors could prompt these actors to undertake retaliatory attacks in Pakistan or against India that could challenge deterrence stability.
To undertake the policies suggested here, the US government would need to address the problematic separation of AfPak from India in the NSC and the State Department, and the division of responsibility between United States Central Command (CENTCOM) and United States Pacific Command (PACOM) whereby the former deals with Pakistan and the latter with India. These bureaucratic divisions make it nearly impossible to develop and sustain a strategy and diplomatic program to address the organic nature of the India-Pakistan deterrence problem and US policies to attenuate it.
Finally, the concept and practice of deterrence stability can be reinforced further by declarations and policies of the other nuclear-armed states. The US, Russia, China, the U.K., France, and Israel have reinforced the taboo against the use of nuclear weapons which emerged after the nuclear attacks on Hiroshima and Nagasaki in 1945.23 In varying degrees, these states over time also have eschewed making threats to use nuclear weapons. In recent years they have accepted political obligations pursuant to the indefinite extension of the NPT to reduce the role of nuclear weapons in their national security policies. These states still fall well short of meeting global demands to eliminate all nuclear weapons, but this need not prevent them and other states from making further efforts to strengthen the taboo against using nuclear weapons. If the highest leaders of these states were to make regular declarations that they recognize the disastrous consequences that would follow any use of nuclear weapons, and the shared imperative to resolve conflicts without resort to nuclear threats or use, they would strengthen moral and political incentives for Indian and Pakistani leaders and publics to explicitly uphold the nuclear taboo in South Asia. The younger generations of urban Pakistanis and Indians aspire to live in societies and states that embody the positive values of more peaceful, prosperous and democratic societies. In multiple ways they express desires to overcome the post-partition legacy of conflict between India and Pakistan. The futures of India and Pakistan depend on the mobilization of these younger generations to demand and support enlightened national policies. The idea that nuclear war should never be fought and can not be won would bolster the centrality of deterrence stability as a strategic objective for Pakistan and India.
1 The author thanks Michael Krepon, Toby Dalton, Sadia Tasleem and Marvin Miller for their critical comments on this paper, while accepting all responsibility for its flaws.
2 Bruce Bueno de Mesquita and William H. Riker, “An Assessment of the Merits of Selective Nuclear Proliferation,” Journal of Conflict Resolution, Vol. 26, No. 2 (June 1982), p. 292.
3 Lawrence Freedman, The Evolution of Nuclear Strategy (Palgrave Macmillan, New York, 3rd edition, 2003), p. 173.
4 Deterrence and compellance can be difficult to distinguish. Broadly, deterrence seeks to mobilize threats of punishment to induce an adversary to forego aggression or, if conflict has already begun, to forego escalation. Deterrence essentially seeks to preserve a status quo. Compellence seeks to motivate an actor to change the status quo, for example by withdrawing forces from occupied territory, or to end the projection of militants into one’s territory. In a conflict where deterrence has failed and an opponent has used force, a state may also act to deter future repetitions of aggression.
5 To follow the logic of proportionality and avoidance of dramatic escalation, India conceivably could retaliate against terrorist attacks attributed to Pakistan by instigating similar attacks against Pakistani targets. Yet, if India’s larger strategic and normative purpose is to delegitimize terrorism, then using similar means would be both wrong and counter-productive.
6 It is difficult to imagine in Pakistan, for example, a scenario like the Cuban missile crisis where President Kennedy rejected the military leadership’s recommendations to attack Cuba and the president’s decision was carried out.
7 Critics of this approach argue that recognizing that India and Pakistan will retain nuclear weapons as long as China, the US and others possess them will undermine the goal of making the NPT universal. However, a strong case can be made that dealing with South Asian realities the way that Pakistan and India perceive them is more likely to lead eventually to conditions that would motivate them to eliminate their nuclear arsenals than will policies of simply demanding their nuclear disarmament.
8 A number of states, of course, fail to impose a perfect monopoly on the projection of violence from their territory, in the sense that some of their citizens may be members of terrorist groups and leave home to commit acts of terrorism abroad. This has occurred in the US, Germany and many other states. The point here, however, is that states are obligated to prevent such actions and to vigorously prosecute those citizens who conduct terrorist acts. It is this standard to which Pakistan can and should be held to account.
9 Kenneth Waltz and Scott D. Sagan, The Spread of Nuclear Weapons: A Debate Renewed (New York: W.W. Norton & Company, 2003), p. 6-7.
10 States have in the past and may in the future seek to escape from the deterrent effect of mutual nuclear suicide by seeking through a combination of offensive and defensive capabilities to destroy an adversary’s capacity to effect nuclear retaliation. Such attempts at escalation dominance are meant to remove or attenuate the mutuality of deterrence and instead enable the advantaged state to deter an adversary (or adversaries) without being reciprocally deterred. To date, however, in the US-Soviet/Russia experience and the India-Pakistan experience the condition of mutual vulnerability has prevailed.
11 “Better dead than red” was a catchy slogan during the Cold War, but it was not a sound basis for state policy.
12 Ahmed Rashid, Pakistan on the Brink (New York: Viking, 2012), p. 47.
13 The Indian Supreme Court subsequently acquitted two of the Kashmiris and reduced the sentence of a third, while upholding the death penalty of one, Mohammad Afzal.
14 Quoted in P.R. Chari, Pervaiz Iqbal Cheema, Stephen P. Cohen, Four Crises and a Peace Process (Washington: Brookings Institution Press, 2007), p. 154.
15 Ibid., p. 157.
16 Steven Pinker, The Better Angels of Our Nature (New York: Viking, 2011), p. 532.
17 Ibid., p. 534.
18 Andrew C. Orr, “Unmanned, Unprecedented, and Unresolved: The Status of American Drone Strikes in Pakistan Under International Law.” Cornell International Law Journal, 44, no. 3 (Fall 2011), p. 736.
19 COAS Speech, August 14, 2012, http://www.ispr.gov.pk/front/main.asp?o=t-press_release&id=2123.
20 Former President Pervez Musharraf told CNN May 26, 2011, that “if Pakistan disintegrates, then it can be dangerous. Otherwise, if Pakistan’s integrity is there, and which I’m sure it will be there as long as the armed forces of Pakistan are there, there is no danger of the nuclear assets or strategic assets falling in any terrorist hands.” Piers Morgan Tonight, CNN, May 26, 2011.
21 Interview with former high-ranking leader of US Strategic Command.
22 Interview with high-ranking leader of Pakistan’s Strategic Plans Division, September 13, 2012.
23 North Korea is also nuclear-armed. While prospects of its contributing to global norms of responsible nuclear stewardship are dubious, the implications of its role in this regard are less important because the nature and practices of the North Korean regime are not models that dignified states seek to follow.
The Carnegie Nuclear Policy Program is an internationally acclaimed source of expertise and policy thinking on nuclear industry, nonproliferation, security, and disarmament. Its multinational staff stays at the forefront of nuclear policy issues in the United States, Russia, China, Northeast Asia, South Asia, and the Middle East.
You are leaving the website for the Carnegie-Tsinghua Center for Global Policy and entering a website for another of Carnegie's global centers.