Andrew C. Winner Toshi Yoshihara A Publication by The Institute for Foreign Policy Analysis, Inc. In Association with The Fletcher School of Law and Diplomacy, Tufts University

CONTENTS

Executive Summary iii

Acknowledgements xvi

Chapter One 1 The Tests and Their Aftermath

Chapter Two 25 Nuclear Force Structures and Doctrines

Chapter Three 53 Stability Issues from Past Wars and Crises

Chapter Four 82 Future Crises and Escalation Potential

Chapter Five 103 Conclusions and Recommendations

The Authors 122

Nuclear Stability in South Asia i

EXECUTIVE SUMMARY

THE SEPTEMBER 11 TERRORIST ATTACKS on the United States and the sub- sequent war in Afghanistan have focused world attention on South Asia. These events represent only the latest in a series of crises in that region of the world, and Washington, and much of the rest of the international commu- nity, tend only to pay attention to it when explosions occur. This tendency is unfortunate because it means that policies are rarely comprehensive or long lasting. In the wake of this latest crisis, Washington has been admonished not to repeat the mistakes of the past by leaving the region to its own devices once the smoke has cleared. This is sound advice, but it begs the question of what any long-term engagement strategy should seek to accomplish. It seems clear that one objective will be rooting out what remains of the al-Qaeda terrorist network in the region and ensuring that such a group can never use the region again as a base of operations. Another objective, however, should be improving the nuclear stability of the region – that is, ensuring that states or non-state actors in this newest of openly declared nuclear hot spots do not use nuclear weapons. Regardless of the outcome of the war on terrorist groups and their supporters in the region, the fact that two states with a history of wars and crises possess overt and evolving nuclear arsenals should be a cause for U.S. concern.

Nuclear Stability in South Asia iii The situation in Pakistan and between ventional forces across the line of control into India and Pakistan in the aftermath of Septem- Pakistan against terrorist training camps. Pak- ber 11 has only heightened the need for careful istan immediately responded by saying that it study of and changes to U.S. and internation- would defend its territory, raising the possi- al policy with regard to nuclear stability on bility of war and escalation between the two the subcontinent. Two examples in the after- nuclear-armed powers. math of September 11 serve to illustrate this Seeing the issue of nuclear stability only point. First, the United States and other states through the lens of the latest crisis, however, became more concerned about the security of risks the development of overly narrow policy Pakistan’s nuclear weapons as fi ghting began prescriptions that will not help to ameliorate in Afghanistan and opposition to Pakistani the problem. The problem is not merely that President Musharraf’s decision to support the the region is prone to crises or that any politi- United States began to mount. Generalized cal-military crisis carries with it the potential worries about Pakistan’s weapons falling into for escalation to the use of nuclear weapons. the hands of extremists or terrorist groups It is that the types of crises endemic to the existed before September 11, but the situation region interact with the ongoing developments in Afghanistan heightened the perceived, and in nuclear doctrine and force structure by both possibly the actual, risk. In October, the Pakistan and India to create a range of pos- Pakistani government arrested two of its nucle- sible situations in which nuclear weapons may ar scientists and questioned them regarding be used. To address that full range of interac- contacts they had with Osama bin Laden’s tive situations, it is necessary to have a better al-Qaeda network. The fear was that they understanding of exactly how crisis scenarios might have passed on nuclear weapons secrets, interact with nuclear doctrine and force struc- technology, or material to terrorist groups. A ture decisions that are being made, and will retired Pakistani naval offi cer was also inves- continue to be made, in both Islamabad and tigated for having possibly assisted Osama bin New Delhi. Laden and his al-Qaeda network in obtaining This study seeks to do just that. It pro- nuclear technology. In a related worry, the Pak- vides an analytic framework for reviewing and istani intelligence services’ ties to al-Qaeda developing policies regarding nuclear stability was highlighted, with the added concern on the subcontinent. It develops proposals that elements of the two organizations might for specifi c, prioritized stability measures that have plotted to gain control of some of Paki- would most directly ameliorate the most like- stan’s nuclear weapons. Second, as the United ly problems of nuclear stability between India States initiated its bombing campaign against and Pakistan in future crises. The goal is for Afghanistan under the rubric of a war on the United States to induce India and Paki- terrorism, new attacks took place in Indi- stan to adopt these stabilizing measures, or to an-controlled Kashmir. India, believing that eschew certain destabilizing moves, in order to Pakistan has long supported such attacks and reduce the chances for a nuclear exchange on was encouraging these to occur as a way of the subcontinent. defusing internal tension over the situation in The study is based on three key assump- Afghanistan, threatened to retaliate with con- tions. The fi rst is that both India and Pakistan

iv Nuclear Stability in South Asia will continue to develop nuclear weapons and tary in nature. In addition, the ruling Hindu will likely deploy them. The second is that nationalist Bharatiya Janata Party (BJP) has as arms control measures sought by the Clinton a central component of its ideology the fulfi ll- administration, such as Indian and Pakistani ing of India’s quest for nuclear-power status. As adherence to the Comprehensive Test Ban long as the BJP remains a key political party in Treaty, address only part of the problem and India, it is likely to be a strong supporter of a are actually peripheral to questions of stabil- robust posture and doctrine. Finally, key politi- ity in crises. The third is that, despite recent cal fi gures in India have been arguing of late efforts by the two states to address bilateral that the country needs to adopt a new strategic issues and negotiate confi dence-building mea- doctrine that better refl ects its position in the sures, political-military crises between them world. Such a new strategic outlook could infl u- will continue and have the potential to spiral ence decisions on future nuclear force postures into a nuclear exchange. and doctrines. In Pakistan, the long-standing This study examines the overall goals infl uence of the military on strategic matters, sought by Washington and the international coupled with the current formal rule of the community from 1998 through 2000, and the military on all governmental matters, will have balance between concerns of nonproliferation signifi cant consequences for future nuclear and crisis stability. Based on this assessment, structures and doctrines. a new analytical framework is developed to Other key drivers in future force posture guide the reshaping of policies, and specifi c and doctrine decisions include the historical recommendations will be offered for consider- strategic competition between the two states, ation by India and Pakistan, the international including the intractable dispute over Kashmir. community, and particularly by Washington. Fundamental power asymmetries between the Recommendations for Washington are put into two (such as landmass, population, military the context of broader relations and interests power, and economic health) will also deter- that the United States has with each country mine future paths of their nuclear postures. In bilaterally and in the region more generally. addition to these tangible measures of power, The fi rst part of this study covers the inter- both sides continue to suffer from severe nal and external forces that could drive the misperceptions about the intentions and capa- nuclear postures in South Asia. It examines bilities of the other that could in turn infl uence emerging Indian and Pakistani force structures postures and, more critically, responses dur- and doctrines and then postulates several plau- ing times of crisis. In addition to the bilateral sible nuclear futures and the implications of dynamic between India and Pakistan, broader these futures for strategic stability. The key strategic factors could infl uence nuclear devel- drivers are as follows. For both sides, domes- opments in South Asia, including particularly tic politics can have a critical infl uence on the direction China takes in its military build- how they each pursue their nuclear postures. up and in its ongoing provision of military In India, the deeply embedded political tradi- and strategic assistance to Pakistan. Several tion of absolute civilian supremacy over the technological considerations will infl uence the military has led to tight civilian control over scale and the sophistication of nuclear weapons weapons development that is inherently mili- and associated delivery systems. For India, its

Executive Summary v policy of self-reliance in these areas has been doctrine is applicable to Pakistan and China, a double-edged sword, shielding it from out- the two likely short- and medium-term state side embargoes but also slowing development threats to India’s security. Both contributing to the rate of India’s domestic industrial capac- to, and a result of, India’s doctrine is its ity, which has proven less than world class. command and control structure. The strong For Pakistan, reliance on outside sources for desire to have civilian dominance and control its nuclear program – particularly North Korea over the decision-making process has led to and China – could render it vulnerable to joint custody of nuclear capabilities, with future limitations or cutoffs in response to the military controlling the delivery systems U.S. pressure or other considerations. Finally, and civilian authorities controlling the actual the level of economic growth in each country nuclear bombs or warheads. Should the need will frame the resources devoted to develop- for nuclear use arise, the civilians and military ing nuclear weapons, although the historical would convene to integrate these separated record indicates that economic under-perfor- elements and the civilians would then incre- mance has not deterred advances in either mentally decentralize the authority for the country. military commanders to carry out the mis- sions. India’s Current and Future Nuclear The size and shape of India’s current nucle- Force Postures ar forces are not certain, but various estimates While the secrecy surrounding the nuclear provide a reasonable picture. Based on publicly programs in both India and Pakistan makes available data, it is estimated that India has analysis diffi cult, it is possible to sketch the the capability to manufacture fi ssion, boosted outlines of their current nuclear doctrines fi ssion, and fusion (thermonuclear) weapons and capabilities based on open sources. India’s although some analysts doubt that the last nuclear doctrine remains in a state of transi- category is fully developed. Through a long- tion, and a comprehensive nuclear strategy has standing nuclear energy and nuclear weapons yet to emerge. The only semi-offi cial document, development program, India has a stockpile the August 1999 National Security Advisory of fi ssile material (plutonium) of somewhere Board statement, was widely seen as more of a upwards of three hundred kilograms, enough wish list than a roadmap, and the Indian gov- to produce between sixty and seventy-fi ve ernment distanced itself from the report after it early-generation devices. Despite this rather was published. Based on various other govern- large stockpile of fi ssile material, India may so ment statements and India’s strategic situation far have only assembled around a half-dozen (large landmass and conventional superiority bombs, most likely gravity bombs, for delivery to Pakistan), the beginnings of a doctrine can by India’s Mirage, Jaguar, or MiG-27 Flogger be derived. India has a declared no-fi rst use aircraft. In addition to aircraft, India has had of nuclear weapons policy. This second-strike an indigenous ballistic missile program for posture enables India to retaliate at a time many years that has produced systems capable (days, weeks, or even months) and scale of its of delivering nuclear warheads. India currently own choosing. The strike would be a counter- operates three variants of the nuclear-capa- value one, essentially targeting cities. Such a ble, liquid-fueled, short-range Prithvi missiles

vi Nuclear Stability in South Asia (ranging from 150 to 350 kilometers). In teen years is something between the low and addition it has developed the medium- to inter- medium option, but all three must be examined mediate-range Agni series. The Agni-I was as one considers the potential for future crisis successfully tested in 1989 after long delays instability. and has a range of 1,500 kilometers with a 1,000-kilogram payload. The Agni-II, which Pakistan’s Current and Future Nuclear succeeded the Agni-I, has a 2,500-kilometer Force Postures range with a 1,000-kilgram payload. This mis- Not surprisingly, Pakistan’s nuclear force pos- sile is fully operational and is being deployed ture and doctrine differ signifi cantly from in the 2001-02 timeframe. The Agni-III, which those of India. Pakistan’s nuclear strategy is is in development, will have a 3,500-kilometer driven exclusively by what Islamabad sees as range, which would enable New Delhi to hold the threat from India. Unlike India, where at risk targets throughout China. civilians dominate strategic thinking, in Paki- Based on how the various drivers interact, stan the military controls both the development and how the strategic situation around India of systems and the strategy and doctrines changes over the next ten to fi fteen years, three adopted for their use. Finally, Pakistan has notional future force structures and associated relied extensively and more recently on out- doctrines could emerge. A low nuclear posture side sources for both its nuclear and ballistic would look much like today’s, with a small missile programs. While not stated as such, number of warheads and bombs and only a few Islamabad’s nuclear doctrine is essentially fi rst overtly deployed ballistic missiles. India would use, largely because of Pakistan’s convention- retain its no-fi rst use stance, and its targeting al inferiority to India. What this use would would remain countervalue in nature. A medi- be is ambiguous, but three policy objectives um nuclear posture would see more solid-fuel can be inferred: 1) Islamabad hopes to deter Agni missiles being deployed and slowly replac- fi rst nuclear use by India; 2) it wants to have ing aircraft as the primary nuclear delivery a capacity to deter or blunt an Indian conven- vehicle. Warheads and delivery systems would tional attack; and 3) it wants to use nuclear remain de-mated, but they would be in close weapons to demonstrate intent and, if possible, proximity to one another in order to ensure internationalize any future crisis. Pakistan has more timely response options. A high nuclear announced a national command authority and posture could not happen until closer to 2015, provided a few details about its structure, but simply because of development bottlenecks and in essence the command authority lies with the funding limitations. Such a posture would military up to the head of government – cur- see the adoption of a triad, including subma- rently General Musharraf. rine-launched ballistic or cruise missiles with Pakistan began its nuclear program in ear- nuclear warheads. Warheads would be assem- nest in the early 1970s, in response to its defeat bled with missiles, and command and control in the 1971 war with India. China is suspected would be more decentralized, with the mili- of having provided signifi cant support for the tary having signifi cantly more autonomy and program up to and including possible designs involvement in the decision-making process. for warheads, fi ssile material, and technology The most likely outcome in the next ten to fi f- to produce highly enriched uranium. Pakistan’s

Executive Summary vii 1998 nuclear tests used highly enriched urani- what Pakistan appears to be edging toward um and were fi ssion devices, not crossing the today. This involves a limited number of nucle- line into thermonuclear explosions. The nuclear ar weapons to be delivered by a mix of aircraft infrastructure centers on uranium enrichment and ballistic missiles. Over time, the ballistic technologies, and best estimates put Pakistani missile component would replace the aircraft. stockpiles of fi ssionable material at more than Most of the weapons would be targeted on Indi- six hundred kilograms of highly enriched ura- an cities, but a few may be kept in reserve nium and over fi ve kilograms of plutonium for for demonstration purposes or actual use on approximately thirty-fi ve to forty warheads. a conventional battlefi eld. A medium posture With Chinese and North Korean assis- for Pakistan would consist of a larger force tance, Pakistan has amassed an array of mounted on solid-fueled missiles such as nuclear-capable ballistic missiles based on both the Shaheen-II. Like the low force posture, liquid- and solid-fueled technologies. Pakistan the targets would primarily be cities, but a has apparently acquired from North Korea warfi ghting option would be more heavily technologies for the liquid-fueled Ghauri-I and weighted. Weapons would still be de-mated –II with 1,500- and 2,000-kilometer ranges from missiles, but command and control would respectively. Pakistan successfully tested the be more robust and sophisticated to track and Ghauri-II in April 1999. Islamabad may also control a larger, mobile missile force. Finally, be developing the Ghauri-III based on the a high posture for Pakistan would consist of 3,000-kilometer range Taepodong-I ballistic a very large solid-fueled, road-mobile ballistic missile, which would enable Pakistan to hit missile force that would be targeted both on any target in India. The Chinese have also sup- Indian cities and on its nuclear and command plied Islamabad with technologies and fully and control facilities. Research and develop- assembled missiles for the solid-fueled Shaheen ment on a naval option for nuclear delivery series. The Shaheen-I, -II, and –III – have rang- would be accelerated and an inventory of short- es of 300 kilometers, 800 kilometers, and 2,500 er-range ballistic missiles would be stockpiled kilometers, respectively. Pakistan has indicated for use with nuclear weapons in a warfi ghting that the Shaheen-II would satisfy the nation’s mode against Indian conventional forces. The long-range strike requirements against India. most likely outcome is a low to medium pos- The indigenously produced Hatf series mis- ture, but in the ten- to fi fteen-year timeframe, siles have been deployed only on a limited basis Pakistan could opt for the high posture if because of their modest ranges (100 and 280 it believes the Indian threat is severe and it kilometers respectively). In addition to ballis- continues to receive substantial support from tic missiles, its main delivery vehicle, Pakistan China. also has several aircraft types that may have been modifi ed for nuclear delivery and now Historic Crisis Instability serve as backups: A-5 Fantans, Mirage-III/5s, Regardless of which force postures and doc- and F-16s. trines are adopted over time, a key condition Three future Pakistani nuclear force pos- to ensure strategic stability is mutual vulner- tures and associated doctrines are posited for ability. As long as both sides can credibly infl ict analytical purposes. The fi rst, low posture is unacceptable damage on each other, mutual

viii Nuclear Stability in South Asia deterrence should work if there are no short- one or both sides moving ground forces to areas comings in the perceptions or decision-making near the border where the other has geograph- processes of either of the two sides. The his- ic vulnerabilities. The moves were, in most tory of crises between the two states, however, cases, defensive, but perceived as just the oppo- suggests that perception problems and deci- site and helped to escalate the crises. Second, sion-making shortcomings are all too common, in most cases one or both sides – more often potentially undermining nuclear deterrence. Pakistan than India – tried to bring in outside Moreover, nuclear deterrence may work at powers in an attempt to change the balance of certain levels of confl ict, but it might, paradoxi- forces and prevail in the confrontation. This cally, make confl ict more likely at levels below will also likely be repeated in the future, either those where nuclear forces might actually be for purposes of deterrence, escalation, or de- used. escalation. An analysis of past crises and wars The wars and crises between Pakistan and between India and Pakistan has revealed two India until the run-up to the 1990 confl ict over not surprising themes that have implications Kashmir all had some effect on the issue of for future crisis stability, particularly the pos- nuclear weapons. Each war and crisis led to sibility of the use of nuclear weapons in those consideration and reconsideration of the nucle- crises. The fi rst is that many of the crises were ar option, or to exercising it openly, in New started over the still-unresolved Kashmir ques- Delhi and Islamabad. The 1990 crisis may have tion. While not the sole cause of all crises or been the fi rst actual crisis in which nuclear wars, it remains a point of contention that has weapons played a role, either through both international and domestic political implica- sides actually weaponizing their latent nuclear tions. The latter point, its relation to domestic capability during the crisis or through Paki- politics, makes it particularly problematic in stani’s signaling of its intent to use nuclear crisis terms because of the volatility of domes- weapons under certain circumstances. While tic considerations in both states. The second nuclear use was not as likely as some journal- theme underlying causes of past wars and istic accounts have claimed, the possibility for crises is the constantly changing military equa- eventual escalation to the nuclear level was tion, on both the conventional and nuclear clear to both sides and to the United States. sides. This has led to perceptions of vulnerabil- Similarly in Kargil in 1999, direct threats of ity and opportunity that in turn have triggered nuclear use were not apparent, but the possi- actions and reactions. Both of these issues bility for escalation to the nuclear level was can be expected to continue to trigger crises present. Moreover, Pakistan’s possession of an and possibly wars over the study timeframe overt nuclear capability may have both embold- (through 2015). ened Islamabad to undertake the operation to In addition to recurring causes of wars and begin with and deterred India from escalating crises, other patterns arise when examining early in the confl ict. past crises between India and Pakistan. Over several wars and crises, two patterns of esca- Future Crises and Crisis Stability lation were repeated and can be expected to Taking into consideration the possible future be factors in future crises. The fi rst involved force structures and the deterrence and other

Executive Summary ix failures evident in past wars and crises between engagement. Second, India could escalate hor- India and Pakistan, it is possible to project izontally by mobilizing and moving armored future crisis scenarios. In these scenarios, we forces forward in the Rajasthan Desert near can identify the potential triggers for escala- Lahore, threatening Pakistan at a vulnerable tion to potential nuclear weapons use. This point, similar to the wars and crises of 1971, in turn can serve as a useful input to design- 1987, and 1990. ing stabilizing measures. While the types of Six escalation problems present themselves future crises that could occur are almost limit- in this scenario, assuming the low force pos- less, three scenarios are expanded upon that tures and associated doctrines. First is the both are within the realm of the plausible over problem that Pakistan may initially believe that the next fi ve to fi fteen years and encapsulate nuclear deterrence will work as a shield under some of the most destabilizing elements of each which it can conduct low-intensity operations alternative doctrine and force structure that that would force India to come to terms over could push either side to escalate to nuclear Kashmir. Second is the uncertainty on the part use. of India as to what constitutes a Pakistani red line in terms of Indian use of convention- Scenario 1 al force to respond to incursions in Kashmir. The fi rst scenario centers on yet another Third, the small size of Pakistan’s force puts clash over Kashmir when both sides have low Islamabad in a “use or lose” situation almost nuclear postures. Such a scenario could occur regardless of India’s actions if Pakistan believes because the Pakistani leadership over time the worst about New Delhi’s intentions and believes that the 1999 Kargil incursion was not capabilities. Fourth, India’s conventional esca- such a serious setback and, in fact, provided a lation options may not in and of themselves useful lesson in the utility of nuclear weapons. be suffi cient to threaten Pakistan’s national Specifi cally, Pakistani leaders believe that they survival and trigger a nuclear response, but can engage India in a ramped-up low-intensi- they may incrementally threaten the surviv- ty confl ict in Kashmir under the umbrella of ability of Pakistan’s small nuclear force and its their nuclear capabilities, and New Delhi will command and control structure. Fifth, even if be deterred from escalating with its conven- Pakistan were to choose a demonstration use tional superiority. If indeed India were unable of its nuclear capability, India would face both to push out a renewed, Kargil-like incursion political and military pressure to respond with using the same methods it used in 1999, it nuclear weapons, particularly considering the would be faced with an acute policy dilemma. It small size of its own force and the potential would either have to escalate the war or capit- vulnerability of its civilian-dominated com- ulate to Pakistan or negotiate on Islamabad’s mand and control system. Finally, of the three terms, which would be equally unacceptable. options available to Islamabad, its limited force India’s options would be limited, and each of size and the history of ineffective international the two major ones holds risks for eventual intervention on its side point toward an all-out escalation to the nuclear level. First, Indian use of its nuclear capability. ground and air forces could be allowed to cross the line of control with more relaxed rules of

x Nuclear Stability in South Asia Scenario 2 ventional forces of each side to the border, The second scenario is one in which fi eld train- raises worries about tactical use opening corri- ing exercises involving both conventional and dors for deep conventional attacks. Even if one nuclear forces spiral into a crisis and introduce assumes that neither side was at least initially the risk of nuclear escalation. The timeframe interested in conducting a nuclear or conven- for this scenario is 2010 and each side has con- tional fi rst strike, force structures, doctrines, tinued the development of its nuclear doctrine and ongoing geographic vulnerabilities create and force postures to the point where they pressures to escalate. Finally, India’s and approximate the medium option. As each side Pakistan’s ability to manage such a crisis modernizes and expands its force structure to and eventually de-escalate remain suspect for the medium option outlined above, they realize the following reasons: 1) absent signifi cant the need for expanding the level and sophis- improvements in general relations, commu- tication of command post and fi eld training nications during such a crisis will remain exercises to improve joint operations on the inadequate; 2) intelligence shortfalls will likely conventional and nuclear levels. In something remain, leading to worst-case assumptions; and of a repeat of the Brasstacks exercise series in 3) inability to differentiate between a conven- 1986-87, the Indian leadership holds a series of tional attack and a nuclear one at the border exercises for both its conventional and nuclear could lead to escalatory moves. forces near the Chinese and Pakistani borders. Given Islamabad’s uncertainty about India’s Scenario 3 capabilities and worry about its intentions, The third scenario is set in approximately 2015 the exercise proves particularly disturbing. when both Pakistan and India are on the verge In response, Pakistan forward deploys its of obtaining the high-posture option in their conventional forces and its short-range, nucle- respective nuclear force structures. The sce- ar-capable ballistic missiles. As in 1987, nario is based on the sudden opening of a attempts at communication to lower tensions strategic capabilities gap between the two, fail to reassure either side. fueling instability and heightening the poten- The elements of potential escalation in tial for crisis. Specifi cally, India purchases this scenario derive in part from the relative a conventional submarine with submarine- newness of the larger and more ready force launched ballistic missiles from Russia, with structures assumed in the medium posture. an option for future purchases. This sends Just as the Brasstacks exercise was a fi rst shock waves through Pakistan, which fears that attempt by India to test new theories and oper- India would use such a system to achieve fi rst ational doctrines for conventional forces, the strike and escalation dominance over Pakistan. exercises in this scenario are the fi rst, or among Fearing that once the submarine is operational, the fi rst, of ready nuclear forces. In addition, in the gap in capabilities would become perma- the medium force posture, Pakistan will begin nent, Pakistan considers responses that could to have a large enough force to begin at least be highly destabilizing and lead to the use of thinking about options for tactical use and nuclear weapons. These options include pur- warfi ghting with its nuclear weapons. This, chasing countervailing capabilities from China combined with the forward deployment of con- and/or North Korea, striking again in Kashmir

Executive Summary xi before the capabilities gap is permanent, and provide an analytical framework for consider- conducting a conventional or nuclear pre-emp- ing stabilizing measures. From a U.S. policy tive strike on the submarine before it achieves perspective, the issue is not to attempt to get full operational capability. India or Pakistan to adopt one or other force The potential for escalation to the nuclear posture. Rather, U.S. policy should focus on level in this scenario has less to do with the making sure that regardless of the force posture actual force structures in place at the time it and doctrines that India and Pakistan adopt, takes place (assuming medium moving to high chances of nuclear use are as low as possible. postures for both India and Pakistan) than the fact that the levels of capability are changing Conclusions and Recommendations rapidly and unequally. Because of the belief The United States has a vital interest in keep- by one side or the other (depending on who ing nuclear weapons from being detonated, acquires a signifi cant new capability fi rst) that either accidentally or intentionally, in South it has been trumped strategically, it may be Asia. It also has a vital interest in preventing willing to take high-risk actions in an attempt further proliferation of nuclear capabilities to recover. In this scenario this may involve around the world – to states and non-state attacks that could invite nuclear retaliation actors alike. The latter concern has only esca- or begin a confl ict spiral that could end in a lated in the aftermath of the September 11, nuclear exchange. This proclivity for risk tak- 2001, terrorist attacks on the United States. ing is particularly true for Pakistan if India India and Pakistan have been a potential continues to pull away from Pakistan in both source of nuclear proliferation since they fi rst conventional and nuclear capability over the acquired nuclear weapons technologies, but next decade and longer. It should be noted that their recent change to overt nuclear weapons the assured second-strike capability that is part states now means that the issue of further hori- of an SLBM force is stabilizing theoretically in zontal proliferation can be addressed openly terms of deterrence, but submarine-launched for the fi rst time. These two vital U.S. interests capabilities raise other possibilities for acci- – keeping nuclear weapons from being used dent and incident. Sea-based assets raise new and continuing to staunch their further hor- and different command and control issues, izontal proliferation – are not incompatible. particularly if India wishes to keep tight Pursuing improved safety, security, and crisis civilian control over its nuclear assets. Simi- stability in South Asia need not undermine larly, potential accidents at sea create different U.S. global nonproliferation goals. Therefore opportunities for misunderstanding and pos- the United States should craft new policies sible miscalculation, particularly if Pakistan designed to enhance stability, particularly cri- attempts to begin regular surveillance of an sis stability, in the region and work to advance Indian SLBM capability. those policies in a systematic manner. Specifi c In some ways, these three scenarios out- policy recommendations are tailored to the sce- line the worst-case combinations of crises narios for purposes of analysis, but they should and future force postures. However, they are obviously be adjusted according to what comes designed to draw out the most troubling insta- to pass on the ground in terms of development bilities in those combinations in order to of actual force structures and doctrines.

xii Nuclear Stability in South Asia · Kashmir and low-level force postures. To Pakistan’s military through renewed and address the potential escalation issues increased funding of the International Mil- when both sides have low level force pos- itary Education and Training (IMET) pro- tures, the United States should under- gram and other exchanges. With India, the take the following policy initiatives. First, United States should engage in discussion it should begin a series of separate bilat- and technology transfer to improve Indian eral political-military dialogues with each ability to detect and analyze weapons of country covering some key issue areas. mass destruction use. Finally, despite pos- With Pakistan, these areas should include sible resistance on the part of India, the debunking myths about the outcome of United States should increase its diplo- the Kargil confl ict; adamant refusal to sup- matic involvement in pursuing a peaceful port any further Kargil-like attempts; reg- solution to the issue of Kashmir. ular discussion of the Pakistani-Indian · Nuclear exercises and medium force conventional military balance; command postures. In addition to the regular, sepa- and control (including safety and security rate political-military dialogues suggested issues); reassurance on the survivability above, the United States should promote of Pakistan’s nuclear arsenal; and specifi c regular, bilateral discussions among the implications and analysis of the risks of protagonists over their specifi c concerns nuclear weapons use. After the September about exercises, either conventional or 11 bombings, the United States has already nuclear. Working with both India and Pak- begun a limited dialogue with Pakistan on istan, the United States should help the two the safety and security of its nuclear arse- sides craft enhanced exercise notifi cations nal, but that dialogue should be expanded. and information exchanges on exercises. With India, the dialogue should include It should also offer to demonstrate U.S. ongoing discussion of the uncertainty of monitoring technologies, including the use Pakistani red lines in terms of nuclear of Open Skies aircraft, for exercises with weapons use; the broader strategic coop- the possibility of follow-on sales of certain eration possibilities with the United States; items being open for discussion. In addi- and the specifi c implications and risks for tion, the United States should broach with India of nuclear use against Pakistan. In both sides the possible remote monitor- addition, the United States should engage ing of nuclear weapons sites and offer to in separate, bilateral simulations and exer- sell the technology that would make such cises of political-military crises designed to monitoring possible. Finally, to ensure fol- highlight escalation issues. With Pakistan, low-up and ongoing implementation (a the United States should engage in regu- chronic problem with confi dence-building lar, joint defense assessments and provide measures in the region), the United States technical assistance to bolster nuclear com- should push for, and offer to help fund, the mand and control capabilities as long as establishment of a permanent, joint risk- they are consistent with safety and secu- reduction center. rity goals. The United States should also · Strategic surprise and high force postures. review and deepen its understanding of Should India and Pakistan ever reach the

Executive Summary xiii point where they are approaching high agency community, with Congress, with force postures as outlined above, then key international partners, and with India it is to be hoped that the United States and Pakistan on how to address the ques- is already well engaged with each side tion. Such a consensus may be easier in about avoiding destabilizing changes in some ways after the events of September their force postures and doctrines. In other 11 because of the national and interna- words, the recommendations for this sce- tional focus on the region. However, in nario are similar to those above, with the other ways it may become more diffi cult as added recommendation that the United issues such as nuclear stability are, rightly, States offer to provide strategic early warn- subordinated to the more urgent task of ing of major changes in the military bal- eradicating the terrorist groups responsi- ance – either conventional or nuclear – in ble for the attacks on the United States. order to minimize the destabilizing reac- Second, the United States must tailor its tions that strategic surprise can produce. approaches to India and Pakistan and rec- Finally, if India and Pakistan eventually ognize that because of history, geogra- construct larger force postures, it will be phy, political structure, and doctrine and time for them to engage in regular strategic force structure issues each country must discussion about these arsenals, their doc- be engaged differently. Third, the United trines, and ways to ensure that stability is States must work to create and sustain enhanced, much like the United States and political, military, and economic leverage the Soviet Union eventually got around to on these issues with each country. Fourth, after years of competition without conver- the U.S. interagency community must be sation. organized around a new comprehensive · Implementation strategy. Many of the rec- plan for engagement with these two coun- ommendations above center on dialogue tries. Finally, the executive branch must and risk reduction measures that have construct a public diplomacy strategy that either been proposed but never adopted or will help build broad support for its new have been formally adopted but have failed policies in Congress and with key interna- in their implementation. Talk is cheap and tional partners. confi dence-building measures are more Nuclear weapons are not going to disappear often than not used as confi dence-busting from the subcontinent any time soon. India, measures on the subcontinent. Therefore Pakistan, China, the United States, and many how the United States goes about pursuing other states will have to adjust to this reality these various initiatives is critical to their and adapt their national security policies to chances for success. Five specifi c issues it. For the United States this means focusing have to be addressed in implementing such on what it can do, given this situation, to best a strategy to improve chances for success. advance its national security interests. Such First, despite the urgency of the issue of a focus requires a re-balancing of priorities nuclear crisis stability, the United States to favor enhancing crisis stability over global is going to have to move ahead slowly, nonproliferation. Actively pursuing policies building a consensus within the U.S. inter- to enhance crisis stability in the region will

xiv Nuclear Stability in South Asia require a sustained and greatly deepened engagement with both India and Pakistan on a range of issues. It should be remembered, however, that despite a willingness, or even an eagerness, on the part of the United States to re-engage with these two countries and to offer all of the assistance it can to enhance stability on the subcontinent, ultimately it is up to Paki- stan and India to establish a modus vivendi that lowers the chances for a nuclear confron- tation. The United States, as the fi rst nuclear power, and as the world’s sole superpower, however, has both the interest and responsibil- ity to do all that it can to assist them in reaching this goal.

Executive Summary xv ACKNOWLEDGEMENTS

A project of this magnitude benefi ted from the support of numerous organizations and individu- als. The authors would like to thank the Smith Richardson Foundation for its generous grant to the Institute for Foreign Policy Analysis (IFPA), which made this study possible. They would also like to thank Robert L. Pfaltgraff, president of IFPA, for his guidance and comments on the draft as well as for his drive to get the fi nal product out the door. The authors would like to thank Adelaide Ketchum for editing the manuscript. Finally, the authors would like to thank Christian Hoffman for his graphic and cover design and layout work on this study. In conducting this study, the authors found that the policy and analytic community covering South Asia to be a collegial and helpful one. Those with whom we spoke were eager to help, offer opinions, ideas, further contacts, and facilitate follow-on interviews and travels. In particular, one of the deans of South Asian studies in the United States, Stephen P. Cohen, provided advice, opened his rolodex of countless colleagues and students, facilitated meetings, and read a draft of the study on short timeline and provided invaluable comments. Noted South Asian scholars and policy practioners Walter Anderson, George Perkovich, Ashley Tellis, Robert B. Oakley, and Neil Joeck were also very generous with their time, providing contacts and dispensing advice about how to proceed with the study and travels to the region. Samina Ahmed also provided advice, opinions, contacts, and generously commented on a draft of the study, again with a very short deadline. In India, Commodore C. Uday Bhaskar generously took it upon himself to act as host and set up innumerable useful meetings with a range of high-level offi cials and senior scholars and journalists. P.R. Chari arranged several meetings and provided his own extremely useful insights into the issues we explored. In addition, the U.S. Embassy staff, including Robert K. Boggs, was generous with their time and views. In Pakistan, with the able assistance of the Pakistani Embassy to the United States, the Joint Staff Headquarters and the Ministry of Foreign Affairs hosted numerous meetings with a range of knowledgeable and helpful offi cials. The Institute for Regional Studies and the Islamabad Policy Research Institute each arranged and hosted tre- mendously useful roundtable discussions. Finally, the U.S. Embassy staff in Islamabad, including John Schmidt, was extremely generous with its time and insights.

xvi Nuclear Stability in South Asia Nuclear Stability in South Asia

CHAPTER ONE

The Tests & Their Aftermath

THE SEPTEMBER 11 TERRORIST ATTACKS on the United States and the sub- sequent war in Afghanistan have focused world attention on South Asia. These events represent only the latest in a series of crises in that region of the world, and Washington, and much of the rest of the international commu- nity, tend only to pay attention to it when explosions occur. This tendency is unfortunate because it means that policies are rarely comprehensive or long lasting. In the wake of this latest crisis, Washington has been admonished not to repeat the mistakes of the past by leaving the region to its own devices once the smoke has cleared. This is sound advice, but it begs the question of what any long-term engagement strategy should seek to accomplish. It seems clear that one objective will be rooting out what remains of the al-Qaeda terrorist network in the region and ensuring that such a group can never use the region again as a base of operations. Another objective, however, should be improving the nuclear stability of the region – that is, ensuring that states or non-state actors in this newest of openly declared nuclear hot spots do not use nuclear weapons. Regardless of the outcome of the war on terrorist groups and their supporters in the region, the fact that two states with a history of wars and crises possess overt and evolving nuclear arsenals should be a cause for U.S. concern.

Nuclear Stability in South Asia 1 The situation in Pakistan and between ventional forces across the line of control into India and Pakistan in the aftermath of Septem- Pakistan against terrorist training camps. Pak- ber 11 has only heightened the need for careful istan immediately responded by saying that it study of and changes to U.S. and internation- would defend its territory, raising the possi- al policy with regard to nuclear stability on bility of war and escalation between the two the subcontinent. Two examples in the after- nuclear-armed powers.4 math of September 11 serve to illustrate this Seeing the issue of nuclear stability only point. First, the United States and other states through the lens of the latest crisis, however, became more concerned about the security of risks the development of overly narrow policy Pakistan’s nuclear weapons as fi ghting began prescriptions that will not help to ameliorate in Afghanistan and opposition to Pakistani the problem. The problem is not merely that President Musharraf’s decision to support the the region is prone to crises or that any politi- United States began to mount. Generalized cal-military crisis carries with it the potential worries about Pakistan’s weapons falling into for escalation to the use of nuclear weapons. the hands of extremists or terrorist groups It is that the types of crises endemic to the existed before September 11, but the situation region interact with the ongoing developments in Afghanistan heightened the perceived, and in nuclear doctrine and force structure by both possibly the actual, risk. In October, the Pakistan and India to create a range of pos- Pakistani government arrested two of its nucle- sible situations in which nuclear weapons may ar scientists and questioned them regarding be used. To address that full range of interac- contacts they had with Osama bin Laden’s al- tive situations, it is necessary to have a better Qaeda network. The fear was that they might understanding of exactly how crisis scenarios have passed on nuclear weapons secrets, tech- interact with nuclear doctrine and force struc- nology, or material to terrorist groups.1 A ture decisions that are being made, and will retired Pakistani naval offi cer was also inves- continue to be made, in both Islamabad and tigated for having possibly assisted Osama bin New Delhi. Laden and his al-Qaeda network in obtaining This study seeks to do just that. It pro- nuclear technology.2 In a related worry, the vides an analytic framework for reviewing and Pakistani intelligence services’ ties to al-Qae- developing policies regarding nuclear stability da was highlighted, with the added concern on the subcontinent. It develops proposals that elements of the two organizations might for specifi c, prioritized stability measures that have plotted to gain control of some of Paki- most directly ameliorate the most likely prob- stan’s nuclear weapons.3 Second, as the United lems of nuclear stability between India and States initiated its bombing campaign against Pakistan in times of crisis. The prioritization Afghanistan under the rubric of a war on of these measures is critical because getting terrorism, new attacks took place in Indi- the two states to adopt any stability mea- an-controlled Kashmir. India, believing that sures, whether they are unilateral, reciprocal Pakistan has long supported such attacks and gestures, confi dence-building measures, or was encouraging these to occur as a way of operational or structural arms limitations will defusing internal tension over the situation in be very diffi cult. The goal is for the United Afghanistan, threatened to retaliate with con- States to induce India and Pakistan to adopt

2 Nuclear Stability in South Asia these stabilizing measures, or to eschew cer- that, despite recent efforts by the two states tain destabilizing moves, in order to reduce the to address bilateral issues and even despite chances for a nuclear exchange on the subcon- increased U.S. engagement in the region since tinent. September 11, political-military crises between The United States has a vital interest, one India and Pakistan will continue and have the that could directly affect the security of the potential to spiral into a nuclear exchange. U.S., in seeing that crises and confl icts in the This study will examine the overall goals region do not escalate to the nuclear level. related to nuclear weapons in South Asia Washington should therefore develop policies sought by Washington and the international that are designed to improve nuclear stability community from 1998 through 2000, and the on the subcontinent and integrate them with balance between concerns of nonproliferation policies that it has for India, Pakistan, and Asia and those of crisis stability. Based on this as a whole. The Clinton administration was assessment, the study will develop a new ana- reluctant to pursue stability measures because lytical framework to guide the reshaping of it was believed that they clashed with non- policies and specifi c recommendations will proliferation goals. The Bush administration, be offered for consideration by India and when it came into offi ce, signaled that it was Pakistan, the international community, and reconsidering that stance as part of its overall particularly by Washington. review of policies toward India, Pakistan, and Recommendations for Washington will be the region. That review obviously was overtak- put into the context of broader relations and en by events in September 2001. As the Bush interests that the United States has with each administration now seeks to put in place its country bilaterally and in the region more gen- new, post-September 11, policy for the region, erally. Obviously, the events of September 11 it needs to put the issue of nuclear stability at and after will have a signifi cant effect on U.S. the top of its list along with the issue of how policy priorities and interests in the region, to ensure that terrorism against the United given Washington’s renewed focus on Afghan- States no longer emanates from this part of the istan, Pakistan, and the al-Qaeda terrorist world. network operating out of Afghanistan. The The study seeks to provide the analytical analysis contained in this study and the recom- framework and policy recommendations nec- mendations that result from it take these events essary to address nuclear stability in a sound into account to the degree possible consider- manner. It is based on three key assump- ing the fl uid nature of the situation in South tions. The fi rst is that both India and Pakistan Asia. Regardless of some of the tactical changes will continue to develop nuclear weapons and that will take place on the ground in the region will likely deploy them. The second is that over the next several years, certain fundamen- arms control measures sought by the Clinton tal realities about the nuclear forces, doctrines, administration, such as Indian and Pakistani and history of confl ict between India and Pak- adherence to the Comprehensive Test Ban istan will drive the nuclear stability issue. Treaty (CTBT), addressed only part of the Initial worries by the international community problem and were actually peripheral to ques- about nuclear safety and stability in the region, tions of stability in crises. The third is briefl y outlined above, need to be grounded

The Tests and Their Aftermath 3 in a fi rmer understanding of these realities in to the next crisis or set of interactions. From order to design relevant policies that have last- this examination, patterns of interaction, esca- ing impact. That fi rmer understanding is what lation, and de-escalation will be analyzed for this study seeks to provide. use as a guide for how India and Pakistan may The fi rst chapter of the study briefl y interact in future crises. reviews the events surrounding the 1998 nucle- Chapter 4 integrates the fi ndings of chap- ar tests carried out by India and Pakistan and ters 2 and 3, developing notional future crises the U.S. and international reaction to those using the postulated future doctrines and force tests. It outlines the U.S. goals toward the two structures. These notional crises are examined states after the tests under the Clinton admin- for the critical points at which nuclear safety, istration and assesses whether any of these security, and control could break down or the goals were met. Finally, it outlines the con- greatest potential for planned or inadvertent cerns that nonproliferation advocates had with escalation exist. Those instances will be exam- pursuing stability measures and makes the ined with an eye to developing measures that case that such concerns were overwrought and might lower the chance that escalation to the detracted from the equally important goal of nuclear level may occur. improving nuclear stability in South Asia. Chapter 5 proposes appropriate stability Chapter 2 of the study provides a basis for measures to address the critical points of insta- the analysis by detailing the current state of bility identifi ed in chapter 4. Stability measures India’s and Pakistan’s nuclear doctrines and will not be limited to formal arms control- force structures. It describes the strategic, cul- like measures or confi dence-building measures tural, political, economic, and technological but will span the range of unilateral changes drivers in their current and future develop- in doctrine or force structure, to formal ment. Building on this base, three alternative agreements, to improvements in intelligence doctrines and force structures are posited for analysis and decision making. In this chapter, each state through 2015. These notional doc- we will select the most important from among trines and corresponding force structures will the measures developed and put them into the provide some clues about potential critical form of specifi c policy recommendations for issues of safety and stability that exist with Washington to pursue both with the inter- various confi gurations. While not predictions national community and with New Delhi of exactly what the situation will look like on and Islamabad. The recommendations will be the subcontinent over the next fourteen years, incorporated into broader strategies for Wash- such projections provide a basis for examining ington’s engagement with each state and for the how nuclear weapons might be considered and region after September 11, taking into account used in future crises. other U.S. interests and goals. These policy rec- Chapter 3 examines past wars and crises ommendations for stability measures will be between Pakistan and India. It lays out the ori- developed with an eye to regional needs, his- gins of the crises, how they escalated, and how tory, and, most importantly, to the future, to they were resolved. It then enumerates how where India and Pakistan are going in develop- each side viewed the crisis and the outcome, ing their nuclear arsenals and doctrines and and the lessons each learned from it to take where relations are likely to go with these two

4 Nuclear Stability in South Asia countries. That way, policy recommendations responded to the India actions with nuclear will target the realm of the probable rather tests of its own. Pakistan conducted fi ve tests than the fanciful. on May 28, 1998, and one more on May 30, 1998, showing that it could match its tradition- The Nuclear Tests and their Aftermath al rival, the fi nal test being a counterpart to On May 11, 1998, India shocked the world by India’s 1974 “peaceful nuclear explosion.” Like conducting three nuclear tests, one of which India, Pakistan wasted no time in declaring it claimed was a thermonuclear device. India that it was a nuclear-weapons state, declaring then went on to conduct two more tests on May that it had the capability to fi t nuclear warheads 13, 1998. Both sets of tests caught the world, to missiles that would be able to strike targets including the U.S. intelligence and policy com- across “most of north and central India.”6 munities, by surprise. The ability of New Delhi The combination of the tests and the sub- to prepare for the tests and have them occur sequent declarations that each state would without being detected was important because proceed to develop nuclear weapons arsenals of the pressure that would have been brought re-ignited the interest of the international com- to bear by the United States and others if they munity in the region and in the long-standing had suspected what the Indian government Indian-Pakistani dispute over Kashmir. The was planning for such tests. In fact, in Decem- international community’s concerns after the ber 1995 the United States detected Indian tests were twofold. First, there were worries preparations for a nuclear test and warned the that the fi rst nuclear tests by a non-nuclear Congress Party-led Indian government against power since 1974 and the fi rst declaration of a it, leading to their cancellation. Then in 1996, new nuclear state since 1964 would break the Prime Minister Vajpayee, who had been head taboo on new, overt nuclear weapons states and of a coalition government for less than two undermine the international nonproliferation weeks, authorized tests and then retracted the regime. Second, the international community order because the government fell. However, worried about not only the existence of these the 1995 incident was useful in showing New weapons but about the heightened potential Delhi the kind of pressure it could expect for their use in what was seen as a relatively from Washington if preparations for testing unstable region with a history of confl ict and were detected. It also provided New Delhi unresolved territorial questions. This concern with useful information on the capabilities about the use of nuclear weapons, or crisis sta- of U.S. national intelligence-gathering systems bility for short, was heightened in the spring of that enabled India to conceal its test prepa- 1999 with the outbreak of fi ghting in the Kargil rations in 1998. After the tests, the Indian sector of Kashmir. government declared that India was a nuclear- The international community’s responses weapons state, leaving none of the ambiguity to these two crises – the testing of nuclear that existed after India’s 1974 peaceful nuclear weapons and the fi ghting in Kargil – echoed explosion.5 previous involvement in the affairs of the Despite signifi cant international pressure region: only crises seemed to provoke sustained in the form of both carrots and sticks, par- interest and attention. Neither the fact that the ticularly from the United States, Pakistan political and military situation in Kashmir was

The Tests and Their Aftermath 5 unstable and unresolved nor the fact that there overt nuclear capabilities. Compounding this were nuclear weapons in the hands of both policy preference was a series of international India and Pakistan was new. The dispute over and domestic laws and obligations forcing fair- Kashmir is as old as the creation of indepen- ly swift and severe sanctions to be imposed and dent India and Pakistan in 1947. In terms limiting maneuver room on the side of address- of a nuclear confrontation, India has been ing the crisis stability issue. presumed to possess at least a rudimentary Second, the United States and other lead- nuclear weapons capability since 1974. Indeed, ing states believed that if they focused on the its nuclear development program, with at least crisis stability aspects of the problem on the some weapons potential, was begun almost subcontinent, they would be tacitly acquiesc- simultaneously with the founding of the Indian ing to India and Pakistan becoming declared state.7 Pakistan was openly acknowledged to nuclear-weapons states. Finally, those who had have nuclear weapons by the United States in followed South Asian issues closely understood October 1990, when the U.S. president could that the underlying tensions between India and no longer certify that Pakistan did not pos- Pakistan were deep and broad and would not sess them, thereby triggering sanctions under be resolved easily or swiftly. The issue of nucle- U.S. legislation known as the Pressler Amend- ar proliferation was seen as a more discrete ment.8 In fact, it is likely that Pakistan had at problem that could be handled relatively quick- least a limited nuclear weapons capability sev- ly and neatly, unlike the lingering problems of eral years before that, but the U.S. government, Kashmir and other historic issues between the because of the Soviet occupation of Afghani- two states. stan and Islamabad’s central role in helping the United States counter it, chose to downplay its The International Response to the Tests knowledge of this fact.9 Led by the United States and other keepers Of the two areas of concern – nonprolif- of the international nuclear nonproliferation eration and crisis stability – the former took regime, the international community reacted precedence in the responses to the testing to the May 1998 tests swiftly and harshly and follow-on declarations of nuclear-weapons- but with little apparent effect on either India state status by India and Pakistan. The reasons or Pakistan. The most concrete response was for this emphasis were threefold. First, the United Nations Security Council (UNSC) res- United States led the response to the nuclear olution 1172, passed on June 6, 1998.10 This tests, and its goal was to cap and eventually resolution built on principles coming out of roll back the development, thereby keeping the an unprecedented meeting of the foreign min- nuclear nonproliferation regime intact. During isters of the fi ve permanent members of the the 1990s, the United States was involved in Security Council just two days earlier.11 The a multi-faceted push to strengthen the global UNSC resolution condemned the tests by both nuclear nonproliferation regime. The tests in India and Pakistan, called on them to halt 1998 threatened to unravel those efforts, and their nuclear weapons development programs, Washington hoped at the outset that a hard-line and urged them to adhere to international response would force New Delhi and Islam- agreements, such as the Treaty on the Non- abad to cap, reduce, and possibly eliminate any Proliferation of Nuclear Weapons (NPT), that

6 Nuclear Stability in South Asia would put a legally binding cap on, if not elim- states supported their statements of condemna- inate, any nascent weapons capabilities. In tion with unilateral sanctions of varying size addition, the resolution called on India and and scope. Interestingly, the UNSC resolution Pakistan to exercise restraint in relations with did not contain provisions for sanctions nor one another and reinvigorate their dialogue to was the resolution adopted under chapter 7 of resolve outstanding issues, including Kashmir. the United Nations charter, which would have While the resolution covered both nonprolifer- deemed the events a threat to international ation and stability, the stronger emphasis was peace and security. The two most important on nonproliferation, and more concrete steps countries to levy sanctions were Japan and the were urged on that front. United States. In the case of Japan, the impo- Following up on the action of the Securi- sition of sanctions against India and Pakistan ty Council, the foreign ministers of the G-8 was signifi cant not only because of the moral nations gathered in London on June 11-12, status claimed by Tokyo as a non-nuclear weap- 1998, and invited a number of other interested ons state and a nuclear weapons victim, but states to the meeting to discuss the nuclear tests also because of the size of its trade and overseas and to take further actions to address what development aid programs. The United States’ was seen as a serious threat to both the global action was signifi cant because of its superpow- nonproliferation regime and stability in South er status and, for Pakistan in particular, the Asia. The additional states included China, ability of Washington to cut off key loans from Argentina, Brazil, South Africa, Ukraine, and international fi nancial institutions. the Philippines. China was there because of its Perm-5 status and its obvious interest in The U.S. Policy Response and its Evolution – a region on its border, and the Philippines 1998-2001 was representing the Association of Southeast The most comprehensive statement of U.S. pol- Asian Nations. The other states were present icy goals during the Clinton administration because they had all voluntarily foresworn vis-à-vis India and Pakistan in the nuclear nuclear weapons capability even though they realm and the rationale underlying the U.S. either had nuclear weapons or had the capabil- approach to achieving those goals can be found ity to produce them. Out of this meeting, the in a speech made by Deputy Secretary of State Japanese foreign minister proposed a task force Strobe Talbott at the Brookings Institution of senior offi cials, led by Japan, to follow up on November 12, 1998. In that speech, Tal- on the issue. This Senior Offi cials Task Force bott noted that the U.S. response had been (SOTF) met three times, most recently on Feb- couched in an international context, and he ruary 11, 1999, when the Republic of Korea cited the UN and G-8 statements as well as joined the group. The role of the SOTF has other multilateral condemnations of the tests. been to coordinate the international commu- He underlined the two principles that had nity’s expressions of concern over the nuclear so far guided U.S. efforts in response to the issue in South Asia. tests. First, the United States wanted universal In addition to multilateral undertakings, adherence to the NPT. In other words, Wash- most countries (152 by one count) voiced oppo- ington was not going to accept, either legally or sition to the nuclear tests. A number of the implicitly, that India or Pakistan was a nuclear-

The Tests and Their Aftermath 7 weapons state under the treaty and entitled to pursuit of these fi ve interim steps, was played join that very elite club of fi ve nations. The out from 1998 to 2000 in a set of bilateral second principle he cited was the practice of dialogues between the United States and each diplomacy as the art of the possible. This meant country at a very senior level, and with India that the United States recognized that for India in the context of reciprocal visits by the heads and Pakistan to adhere to the NPT as non- of state. Relations with Pakistan, and the dis- nuclear weapons states was a long-term goal cussion of nuclear issues, were also wrapped and that the United States should attempt to up in the politics of visits as President Clinton induce them to take fi ve particular steps that and the administration debated, and ultimately would “…help avoid a destabilizing nuclear and decided, to have a brief meeting with Pakistan’s missile competition and more generally reduce chief executive General Pervez Musharraf tensions on the subcontinent and bolster our when the president traveled to South Asia in global non-proliferation goals.”12 2000. Those fi ve near-term steps formed the basis During the Clinton administration, Strobe for ongoing discussions with India and Paki- Talbott led the nuclear dialogues with India stan throughout the remainder of the Clinton and Pakistan. He met ten times with his administration. The fi rst step was to get both Indian counterpart, External Affairs Minister India and Pakistan to sign and ratify the Com- Jaswant Singh, and nine times with Paki- prehensive Test Ban Treaty. The second was to stani Foreign Secretaries Shamshad Ahmad have both nations undertake a moratorium on and Abdul Sattar. These dialogues primarily the production of fi ssile material – both pluto- covered the fi ve U.S. near-term objectives nium and highly enriched uranium – pending described above and were included in dis- the conclusion of a global treaty being nego- cussions of broader bilateral relationships. tiated in the Conference on Disarmament in Despite the good personal relationships that Geneva that would cut off production of such were formed in these meetings and the breadth material on a worldwide basis. The third step of the discussions, the results in terms of the was to limit the development and deployment United States’ achieving its near-term objec- of missiles and aircraft capable of carrying tives were limited. Similarly, the reciprocal weapons of mass destruction (WMD). The visits of the U.S. president and Indian prime fourth was for both India and Pakistan to minister produced a tremendous amount of tighten their export controls on materials and general goodwill but little progress on the set technologies that could be used to manufac- of nuclear issues. ture nuclear weapons or other weapons of mass In addition to taking the lead rhetorically destruction or missile delivery systems. The to condemn the nuclear tests and setting out fi fth step was a call for direct, high-level, and interim steps it wanted India and Pakistan frequent dialogue between India and Pakistan to take on the way to the longer-term goals over core issues in their relationship, including outlined in the UNSC resolution, the United Kashmir, confi dence-building measures, and States imposed relatively stiff economic sanc- better communications in general.13 tions on both countries. The sanctions were U.S. policy related to nuclear issues and required under legislation popularly known as stability on the subcontinent, particularly the the Glenn Amendment. This amendment to

8 Nuclear Stability in South Asia the U.S. Arms Export Control Act requires templating a nuclear option and that they were that if the president determines that a non- also a show of support by the United States for nuclear weapons state, as defi ned by the NPT, states that had renounced a nuclear capability. detonates a nuclear explosive device, certain The Glenn Amendment, however, was partic- sanctions apply.14 At the time, the administra- ularly rigid in that it did not allow for any tion stated that the sanctions had the following waiver mechanism by the administration, a objectives: to send a strong message to would- common characteristic of sanctions legislation be nuclear testers; to have maximum infl uence and other congressionally mandated foreign on Indian and Pakistani behavior; to target affairs laws. the governments, rather than the people; and Specifi cally, the sanctions prohibited U.S. to minimize the damage to other U.S. inter- bilateral assistance in all areas, including mili- ests.15 tary assistance. They also banned licensing Almost immediately after the imposition of all items on the U.S. Munitions List and of sanctions, the discussion and debate began for certain dual-use items to a particular set in Washington over whether they were going of end users. It should be noted that many to achieve their goals and whether and under restrictions on military and dual-use sales had what circumstances the United States should already been in place for India since the 1970s consider lifting them on either country. Since after it became clear that it had an active nucle- the tests, the debate over the easing of sanc- ar program. Probably most telling, the law tions has refl ected the broader issue within the required the United States to oppose any loan U.S. policy community of how best to infl uence or fi nancial or technical assistance from any the behavior of India and Pakistan in the nucle- international fi nancial institution that did not ar realm and how to balance that desire with support basic human needs. This last provision the need to address global nonproliferation con- had some fl exibility in that the United States cerns. The debate has evolved considerably could choose to abstain from voting rather than since 1998, when the rhetoric of “cap, roll- vote “no” on these international support pro- back, and eliminate” in terms of India’s and grams. In fact, in December 1998, the United Pakistan’s nuclear capabilities dominated U.S. States did allow an International Monetary thinking and, particularly, what New Delhi Fund (IMF) package to go through for Pakistan and Islamabad believed they were hearing most on a one-time basis. Most of the restrictions loudly from Washington and the international and sanctions enacted in 1998 had already community. been in place for Pakistan under the provi- As noted above, the sanctions imposed sions of the Pressler Amendment, but some on both India and Pakistan after the tests new prohibitions were applied under the Glenn were required by law, and the Clinton admin- Amendment. For India, most of the sanctions istration had no choice but to impose them. were new.16 Moreover, the leveling of sanctions supported While U.S. goals vis-à-vis India’s and Pak- its general philosophical approach to the issue. istan’s nuclear programs did not offi cially Deputy Secretary Talbott’s Brookings speech change over the last three years of the Clinton noted that the sanctions were designed to be a administration, the view on the utility of sanc- disincentive to other states who might be con- tions did evolve for a number of reasons. First,

The Tests and Their Aftermath 9 it became clear relatively quickly that the sanc- tant secretary of state for South Asia, Christina tions were not going to be suffi cient to coerce B. Rocca, reportedly stated in her nomination either India or Pakistan into giving up their hearing that she believed that sanctions against newly acquired nuclear capabilities. In fact, the both India and Pakistan had outlived their util- Clinton administration never believed that this ity and were an obstacle to engagement with would be the case and never used that argu- the subcontinent.20 In a visit to India in July ment in its rationale for applying sanctions. 2001, the chairman of the U.S. Joint Chiefs of Second, other countries, even those with very Staff, General Henry H. Shelton, hinted that strong nonproliferation credentials, began eas- sanctions on India might be removed soon.21 ing their own unilateral sanctions after a short In August, Deputy Secretary of State Richard period.17 Third, in internal debates within Armitage stated that the administration would the administration, it was argued that the begin working with Congress in fall 2001 to lift sanctions were counter-productive and limited sanctions on India “…in order to clear the way India’s and Pakistan’s willingness to discuss for more military planning, joint operations the whole issue until the sanctions were eased and eventual sharing of weapons technology or removed entirely. In support of this specifi c with New Delhi.”22 Finally, in the aftermath concern over the impact of sanctions on dia- of the terrorist attacks on the United States logue in South Asia, there was a more general on September 11, 2001, and the U.S. need for belief that the executive branch should have greater cooperation from Pakistan and India in fl exibility to apply, waive, and if necessary lift combating that threat, Washington lifted the sanctions as a tool of foreign policy. sanctions it had imposed on both countries in The result of this changing view during the May 1998.23 Clinton administration was a gradual easing of the sanctions against both states. On Octo- Assessing International and ber 25, 1999, an amendment to the defense U.S. Policy Since the Tests authorization bill provided the president with Not surprisingly, after the long buildup in their the authority to waive sanctions under the nuclear programs and the tremendously posi- Glenn Amendment. The Clinton administra- tive public responses in each country to the tion promptly exercised the new authority and tests, neither India nor Pakistan was in any waived a portion of the sanctions. On the next- hurry to comply with the demands of the UN to-last day Clinton’s term, the administration Security Council. Security Council resolution waived sanctions again to allow a one-time sale 1172, so far, has been largely ignored. Neither of spare parts (with U.S. content) by the United country has halted its nuclear weapons devel- Kingdom to India for British-manufactured Sea opment program, and neither has expressed King helicopters.18 any interest in adhering to the NPT as a non- The Bush administration came into offi ce nuclear weapons state. The latter move would promising to review sanctions policy globally have not only meant renouncing the capabili- and stated that it would seriously consider ties proven by the May 1998 tests but also a further reductions on sanctions, at least on reversal of a long-standing philosophical oppo- India.19 Signaling a further shift, in May 2001 sition to the structure of the treaty, which the Bush administration’s nominee for assis- allows the fi ve nuclear power states to retain

10 Nuclear Stability in South Asia nuclear weapons while denying that possibility efforts to achieve them provides a useful pic- to all others. The two did reinvigorate dia- ture of where policy thinking for the future logue with one another, including the meeting must start as well as highlighting the diffi cul- between heads of state in Lahore in February ty that the Bush administration faces when 1999. Of course, following closely on the heels discussing nuclear and stability issues on the of that meeting, they clashed in Kargil, which subcontinent. demonstrably proved a lack of restraint on The Clinton administration’s fi rst goal, get- the part of Pakistan. The July 2001 summit ting both India and Pakistan to sign and ratify between Indian prime minister Vajpayee and the CTBT, was a tall order at best. Slight prog- Pakistani president and chief executive Mush- ress was made as both India and Pakistan, arraf was another attempt at restoring a shortly after the May 1998 tests, declared uni- dialogue, but the failure to issue a joint com- lateral moratoriums on further testing. Each muniqué showed the depth of the problems has extended, and observed, those moratori- between the two states and the limited maneu- ums while they continue to consider whether ver room of their leaders. to sign on to the CTBT. Both Pakistan and The United States was not very successful India have expressed interest in eventually in achieving either its long-term or short-term joining the CTBT, but India continues to have goals regarding the Indian and Pakistani nucle- reservations about the treaty that are not nec- ar programs. U.S. exhortations had no evident essarily related to whether Pakistan chooses effect on Indian or Pakistani views on giving to adhere to it.24 The continuing moratori- up nuclear capabilities and signing on to the ums on testing, and expressions of interest, NPT as non-nuclear weapons states. On the could be characterized as limited progress since shorter-term goals outlined by Deputy Secre- India would not have even considered the trea- tary Talbott, Washington made little progress ty before its tests because it believed that the as well. While it is impossible to prove that CTBT was part of the discriminatory practices India or Pakistan might have taken further of the world’s nuclear- weapons states. steps to weaponize and deploy their nuclear However, moving India and Pakistan from capabilities had Washington and others not unilateral moratoriums to a legally binding halt pushed and prodded, it is possible to show that is highly unlikely for a number of reasons. very little concrete has happened on the fi ve On the Indian side, some Indian scientists and short-term objectives. In fact, it is possible to analysts are not completely convinced that the show that some have been ignored and might series of tests conducted in May 1998 produced have to be taken off the table in future talks suffi cient data to allow India to cease testing because they are no longer relevant. To be fair, altogether, particularly if they believe that they the reasons that Washington’s policies have will have more robust deterrence requirements not succeeded are a mix of how they were car- in the future.25 Specifi cally, some Indian ana- ried out, the diffi culty of the objectives sought lysts have argued that further testing may be given the domestic politics and foreign policies needed on hydrogen bombs to address the Chi- of both India and Pakistan, and intervening nese threat.26 The potential need for future events that undermined Washington’s goals. A testing has also been raised by analysts out- brief review of the fi ve short-term goals and side of India for the same reasons.27 On the

The Tests and Their Aftermath 11 Pakistani side, Islamabad benefi ts from having such a step. Basically, neither side believes it tested a highly enriched uranium weapon with has enough for its credible minimal deterrent, a more proven design, which some analysts and Pakistan wants to produce some plutoni- argue was provided directly by China. Regard- um for more advanced weapons designs before less of the origins of either the weapon itself agreeing to any cutoff of production. India has or its design, if Pakistan wishes to move on to rhetorically supported a cutoff knowing that more sophisticated plutonium-fi red weapons, it the negotiations will not get anywhere, and may also have a requirement for further test- Pakistan’s position is one that would take cur- ing.28 rent stockpiles into account, a position that The Clinton administration’s second goal, India opposes. Regardless, the negotiations for to get both states to enter into a moratorium a global ban at the Conference on Disarmament on the production of fi ssile material pending have only just moved out of the “pre-negotia- the conclusion of a global treaty that would cut tions” or mandate phase. It is unclear whether off such production on a worldwide basis, went actual negotiations will begin any time soon or nowhere. The problem with this goal was that whether they will continue to be stymied by Indian and Pakistani adherence would consti- Russian and Chinese insistence on linkage to tute a direct and complete cap on their nuclear outer-space arms race talks – essentially part forces and capabilities, an outcome that neither of their campaigns against U.S. missile defense country is prepared to accept. Whereas test- efforts.29 ing moratoriums and even signing the CTBT The third practical step urged by the Unit- would represent limitations on future nucle- ed States during the Clinton administration ar capabilities, they would not necessarily be was a limitation on the development and absolute limitations. Enough data may have deployment of missiles and aircraft capable of been gathered in the May tests to suffi ce for a carrying weapons of mass destruction. Essen- wide range of nuclear force structures and doc- tially, the United States was looking to prevent trines. Moreover, computer simulations and an arms race from developing on the subcon- a range of other experiments and activities tinent. Washington was also seeking to keep involving nuclear materials not prohibited by the degree of weaponization and deployment to the treaty may also provide information that a minimum to ease concerns about command would enable India and Pakistan to move and control and safety of dispersed weapons. forward with their respective nuclear pro- Exactly what types of limitations Washing- grams. The same cannot be said for a cutoff of ton was looking for has not been made public, fi ssile material production. Such a commitment but details were developed in the bilateral dia- would have capped the numerical size, and pos- logues and included requests for moratoriums sibly also to some degree the structure, of each on missile tests and the non-mating of war- side’s nuclear forces at a level corresponding heads with either missiles or aircraft along to how much (of each type of) fi ssile material with a geographic separation of warheads from they currently had stockpiled. Given the still- delivery systems.30 Success in this step is easy evolving nature of each side’s force structure to measure in some areas, less so in others. and doctrines, neither New Delhi nor Islam- In terms of missile development, Washington’s abad has shown any serious interest in taking pleas appear to have had little to no impact.

12 Nuclear Stability in South Asia Shortly after the tests, Indian offi cials to see restraint, but little evidence exists to sup- began to make statements about efforts to port such a claim. In fact, it appears that the only weaponize their nuclear capability, including constraints on missile and other delivery vehicle miniaturizing warheads to fi t on the short- developments since 1998 were either technical range Prithvi missile.31 Such statements may or fi nancial. not necessarily refl ect actual steps taken but The fourth step that the Clinton adminis- rather lobbying on the part of advocates with- tration requested of India and Pakistan, the in India. However, other steps that India tightening of export controls on materials and took clearly refl ected an unwillingness to com- technologies that could be used to manufac- ply with U.S. requests for limiting weapons ture nuclear weapons or other weapons of mass development. The ruling BJP revived the nucle- destruction or missile delivery systems, has ar-capable Agni-II intermediate range missile been the area where the most progress has program in 1998, and the Agni-II was tested apparently occurred. This is not surprising. It for the fi rst time on April 11, 1999, and again is one of two requests that can be undertaken on January 17, 2001.32 Since the second test, without actually accepting any limitations on Indian offi cials have said that the Agni-II will their own nuclear programs. While, as Neil be inducted into the armed forces.33 Offi cials Joeck points out, both New Delhi and Islam- also announced an intention to build and test abad, having been targets of export control a third and longer-range version of the mis- regimes, may be reluctant to set up such con- sile, the Agni-III, as well as an intercontinental trols themselves, they also realize that it is a ballistic missile (ICBM) in early 2002.34 In way to demonstrate a degree of responsibility at addition, India has continued purchases of fairly little cost.38 Both countries already had advanced fi ghter aircraft from Russia and the some export control mechanisms and guide- development of its own fi ghter aircraft. lines in place, and new legislation is being Pakistan has also failed to show any measur- considered to tighten controls further. Both able restraint in the area of testing ballistic mis- New Delhi and Islamabad were receptive to siles. Three days after India tested its Agni-II discussions on these issues and have agreed to in 1999, Pakistan responded with a test launch consider at least some of the suggestions and of its Ghauri-II (also known as the Hatf-VI) recommendations from Washington. In June and followed one day later with a test of the 2001, Islamabad announced still more reforms Shaheen-I.35 In late 2000, Pakistan began serial and new laws designed to bring its export con- production of the Shaheen-I.36 Following the trols into alignment with Washington’s desires. January 2001 Agni-II test fl ight conducted by Again, this type of request is easier because India, Pakistan responded by saying that it was it focuses on the issue of further proliferation ready to fl ight test its Shaheen-II and that both (something that neither India and Pakistan this missile and its earlier version were not only can say they support) rather than what George in “regular production” but had already been Perkovich terms “unproliferation” (giving up inducted into the army.37 It is possible to argue nuclear weapons and nuclear weapons produc- that the pace of missile development would have tion capability) or crisis stability. been even faster had not the United States and The last near-term step requested by the world community made it clear that they wanted Clinton administration, direct, high-level and

The Tests and Their Aftermath 13 frequent dialogue between India and Pakistan, contending issues that exist between the two began promisingly with the meeting of the two states remains to be seen, but the results of the heads of government in Lahore, Pakistan, in fi rst meeting were not overly promising. On February 1999. The historic bus trip across the the one hand, the leaders met and apparently border by Indian Prime Minister Vajpayee and got along well. In addition, General Musharraf the statement issued by the two sides at Lahore has issued an invitation for Vajpayee to make a seemed to have set the tone for serious discus- reciprocal visit to Pakistan.42 On the other, the sions about a range of grievances.39 This was two sides failed to issue a joint communiqué, important as it meant that tensions could pos- and it was clear that little progress was made on sibly be reduced and chances of inadvertent central issues such as Kashmir. The dynamics escalation might be lowered as regular dialogue of the post-September 11 relationship have yet was initiated. This early optimism, however, to really develop and doubtless will take some was dashed by the Kargil confl ict. Even after months, or more, to shake out. India succeeded in bringing the forces in the Dialogue between India and Pakistan region to heel, New Delhi was not ready to about their relationship has been taking place begin a dialogue again swiftly because it sus- in a sporadic fashion for years on both the pected that Islamabad had been readying the offi cial and the non-offi cial “track 2” levels. operation at the same time it engaged in the Its frequency, intensity, and level of success Lahore meeting, which India saw as particular- are driven primarily by the individual leaders ly duplicitous.40 Domestic politics complicated involved and domestic politics in both states. renewed dialogue as India underwent yet The fact that both sides met in Lahore in 1999 another general election and Pakistan suffered and again in July 2001 is not in response sole- through yet another military coup. ly to U.S. or international pressure. Even if In 2001, some positive signs emerged. India international pressure played a part in push- agreed to engage in dialogue with the politi- ing the two sides to meet, it cannot make the cal groups in Kashmir (the All Parties Hurriat dialogues fruitful. Sanctions, promises of their Conference) although it initially refused either being lifted, and external exhortations alone to include or to talk simultaneously with the are not going to produce a stable and sustained Pakistani government. New Delhi also engaged dialogue that will lower the probability of con- in several unilateral cease-fi res in the fi rst half fl ict. Domestic politics and the realities on the of 2001 in Kashmir although, again, Pakistan ground in Kashmir, particularly the level of was not directly involved. Finally, in mid-2001, violence, will be the primary factors in deter- New Delhi issued an invitation to Pakistan’s mining whether the current dialogue is any chief executive, General Musharraf, for talks, more successful than the previous ones. and Pakistan accepted the invitation in late In sum, not many of either Washington’s May.41 Having been eager to restart dialogue or the international community’s goals have and having offered to meet the Indian govern- been achieved in South Asia since the May ment any time, anywhere, Musharraf could 1998 tests. This does not mean necessarily that hardly turn down the invitation. Whether the the goals were wrong or that they should be July 2001 Agra summit will have marked the discarded. It simply means that they should be start of sustained and serious dialogue on the re-evaluated and possibly re-ordered in accor-

14 Nuclear Stability in South Asia dance with current realities in the region and goal is that other states will follow India and recent experience along with the approaches Pakistan and openly develop nuclear weapons, that the international community, and Wash- negating decades of diplomatic work that has ington in particular, has been using to attempt helped keep the number of declared nuclear to achieve them. The experience of the past weapons states to fi ve. For this reason, non- three-plus years has shown that policies that proliferation advocates argue that India and are too tightly bound (or are perceived to be Pakistan should not be recognized as nuclear so in New Delhi and Islamabad) to a goal of weapons states in legal terms (by inclusion “unproliferation” are doomed to failure. The as such in the NPT). In addition, they argue fact that a nuclear-armed China continues to that policies by the international community assist one side in the confrontation and belit- should punish the two governments for their tle the other at regular intervals of course actions or at the very least should not reward also weighs heavily against any unprolifer- their actions. ation goals. At the same time, events over Again this focus is based on the concern the past three-plus years have shown that gen- by nonproliferation advocates that other states eral regional stability problems remain (i.e., with the potential to develop nuclear weapons Afghanistan, Kargil) and the possibility of the may see the Indian and Pakistani example and use of nuclear weapons in a crisis has not been adjust their policies on weapons development. abated by any of Washington’s or the interna- Exactly how those policies would be adjusted tional community’s policies. In fact, some of depends on the state in question. The states the policies designed to foster nuclear stabil- fall into two categories, and the concerns vary ity – exhortations to dialogue and promoting a accordingly. The fi rst category includes those freeze on nuclear weapons and missile develop- states currently pursuing a nuclear weapons ment – have clearly failed. The bottom line is capability, including Iran, Iraq, and the Demo- that unproliferation is a chimera and needs to cratic Peoples Republic of Korea (DPRK). No be downgraded in priority relative to the goal policy makers or analysts believe that they of stability, particularly crisis stability, on the will be dissuaded from their pursuit of nucle- subcontinent. ar weapons by international responses to India and Pakistan or by sanctions. Nor will Nonproliferation Concerns the demonstration effect of India and Paki- While it is fairly easy to proclaim that unpro- stan becoming overt nuclear weapons states liferation should be downgraded as an objective necessarily accelerate their pursuit of these in South Asia because it is not achievable in capabilities. Nonproliferation advocates, how- the short to medium term, there are those who ever, do worry about how closely these states argue that such an approach is irresponsible.43 are watching international reaction to adjust It is important to understand the concerns of their own political and diplomatic tactics as nonproliferation advocates and make sure that they approach an overt nuclear capability. The any shift in priorities and subsequently pro- concern is that countries such as these three, posed measures do not damage U.S. interests in watching international reactions to India and the area of nonproliferation. The primary con- Pakistan, would adjust their programs to mini- cern of those who focus on the nonproliferation mize punishment and maximize rewards. The

The Tests and Their Aftermath 15 other set of countries, those with a technologi- could have addressed issues of stability on the cal capability to become nuclear weapons states subcontinent. Not all are compelling, and some but that have either chosen not to do so or who have lost some of their force three years after have given up this capability (Japan falls into the tests. It is also becoming clear that the Bush the former category and Ukraine into the lat- administration does not hold these views and ter) could feel betrayed if India and Pakistan may view nuclear deterrence as stabilizing in were somehow rewarded for their actions. In certain circumstances. fact, Clinton administration offi cials were very Taking the concerns about proliferation clear that they did not want to break faith with in order, the fi rst point to note is that the countries such as these who they felt were at global nuclear nonproliferation regime is fairly the core of the global nonproliferation effort.44 robust. Given a prediction made by President They felt that if the United States were seen as John F. Kennedy in the 1963 that there would rewarding India and Pakistan after the tests, be twenty-fi ve nuclear-weapons states within then these countries might either reconsider a decade, the current reality of fi ve recognized, their decisions to forego nuclear weapons or, two overt, and one or two clandestine nucle- less dramatically, lessen their support for global ar weapons states is a fairly good result.45 nonproliferation regimes. The reasons for the success of the regime, and A related argument for holding the its likely continued success in the future, are nonproliferation/unproliferation line and post- twofold. First, a consensus exists among most poning any discussion of specifi c stability states on restricting access to critical technolo- measures that could imply acceptance or for- gies and components, particularly fi ssionable mal recognition of nuclear weapons status was material, necessary to construct nuclear weap- that it simply was not yet necessary. This “now ons, and regimes have been established to is not the time” argument is a familiar one in formalize controls on transfers. Second, states diplomatic circles on a variety of issues, but it is choose to develop nuclear weapons for a variety important to consider it carefully in this case. of reasons. Scott D. Sagan offers three mod- Immediately following the tests, it was not clear els: security based, domestic-politics based, and that either India or Pakistan was actually going norms based.46 For each of these models as to to weaponize its nuclear capabilities, mate war- why states choose to “go nuclear,” a number of heads with delivery systems, or deploy nuclear policy options exist to change the decision-mak- forces with military units. Offering to dis- ing calculus. If the state has security concerns, cuss or provide assistance for permissive action security guarantees in the form of alliances links (PALs), command and control, or other with a nuclear-weapons state, negative security issues related to making fi elded nuclear forces assurances, and confi dence-building measures safer in both peacetime and crisis could have are possible alternative solutions. If a decision been seen as not only acquiescing in but also to go nuclear has domestic roots, changing the actually encouraging the deployment of sophis- incentives of the various domestic coalitions ticated nuclear weapons arsenals. for and against developing nuclear weapons These arguments were the basis for deci- is diffi cult but not impossible. Finally, if the sions by the Clinton administration not to norms-based model is the reason, then the consider or pursue a variety of measures that continued existence and enhancement of non-

16 Nuclear Stability in South Asia proliferation treaties and regimes as well as the bilities in South Asia are troubling, they do not international community’s overall response to have the immediate global implications for the the South Asian tests should suffi ce to hold the spread of such capabilities to a region such as line. the Persian Gulf. On the issue of the time not being right The question about whether the interna- to address stability issues, it is true that more tional community should continue to punish than three years after the tests it is still not India and Pakistan through sanctions for the clear exactly what the nuclear force struc- tests has undergone a substantial change since tures of the two states will be. However, they September 11, 2001. No consensus ever existed have taken steps to continue nuclear weapons for formal multilateral sanctions through the development and are either considering or are United Nations or even regional organizations making decisions about command and control, such as the European Union. The United States deployment, mating of weapons, and delegation was beginning to loosen its own sanctions of release authority. If the United States were to on both countries before September 11, and wait still longer before engaging in discussions other leading nonproliferation advocates such about these issues, all of the relevant decisions as Japan have been considering doing so as will have been made, and the United States well.47 Once the need for substantial Paki- will have exerted no infl uence on those deci- stani cooperation in the war in Afghanistan sions. As will be shown in this study, not every became apparent, the Bush administration aspect of stability may have to be discussed or removed or waived almost all nuclear and be the subject of assistance, whether technical democracy-related sanctions on Islamabad, and or fi nancial. Choosing carefully which areas the corresponding nuclear-related sanctions of stability are the most useful on which to on New Delhi.48 Some restrictions, related to focus and how they are approached may help Pakistani importation of M-11 missiles from ameliorate concerns that such an approach China, remain in place, but they could be could undermine global nonproliferation goals waived in the future and in any case expire in or regimes. November 2002.49 The question therefore, at Finally, concerns are overblown that coun- least for U.S. policy regarding nuclear stability, tries such as Iraq, Iran, and North Korea might has changed from the question of punishment glean tactics from India and Pakistan’s dealings to one of inducement. The U.S. is already pro- with the international community after the posing an aid package for Pakistan as part of tests. The international community’s reaction its program of stabilizing the regime there and to a nuclear test by any of these other states will rewarding it for cooperation in the war on ter- be very different from its reaction to India’s rorism.50 Whether this greatly enhanced level or Pakistan’s actions. This will be particularly of support can be leveraged to persuade Islam- true of the United States. For Washington, the abad to adopt more stable and secure nuclear simple reason is that neither New Delhi nor policies is unclear. More specifi c assistance, in Islamabad threatens vital U.S. national secu- the form of technical or fi nancial measures to rity interests or those of its allies. Clearly, address specifi cally nuclear safety and stabil- that is not the case with Baghdad, Tehran, or ity, has yet to be formally proposed although Pyongyang. While overt nuclear weapons capa- some initial discussions on these issues took

The Tests and Their Aftermath 17 place in October 2001, apparently without any has also served to highlight both the general results.51 Whether specifi c assistance can or instability in the region and the potential should be provided depends on two sets of domestic instability in Pakistan – factors that analysis. The fi rst is whether any specifi c assis- do not necessarily affect the Indo-Pakistani tance proposed is legal or acceptable under dynamic but have implications for concerns domestic law, the provisions of the NPT, or about the safety and security of nuclear weap- other legally or politically binding nuclear tech- ons and technology. nology nonproliferation regimes such as the However, some scholars of the region as Zangger Committee. This depends on the exact well as numerous policy makers and analysts nature of the assistance offered (the transfer in the region have argued that Indo-Pakistani of certain nuclear-related technologies is pro- relations are actually quite stable and that hibited under the provisions of the NPT and the chance of nuclear weapons use or large- Zangger Committee rules) and the policy com- scale war between them is quite low. In a munity’s interpretation of existing laws and detailed study in 1997, before the tests but well commitments. On a broader political level, after both sides were assumed to have nuclear it depends upon whether the proposed assis- capabilities, noted India scholar Ashley Tellis tance crosses the line to legal acceptance or argued that a type of “ugly stability” exists encouragement of India’s and Pakistan’s nucle- between India and Pakistan because neither ar weapons status. state could achieve its desired political objec- tives through conventional war. This stability Is Crisis Stability a Legitimate Worry? would be “ugly” because unconventional In addition to arguments made by those pri- confl icts in the form of state-supported ter- marily concerned with nonproliferation, there rorism or state-supported insurgency would is the question of whether a focus on safety be the principal basis for security competition and stability, particularly crisis stability, is between Pakistan and India. This prediction warranted. At fi rst blush, it would appear to seems to have been proven by the Kargil con- be obvious. Most media coverage of the nucle- fl ict, although it could be argued that this ar tests notes the long history of antagonism confl ict began to push the upper limits of what between India and Pakistan, the four major could be called a low-intensity confl ict because wars fought since their creation, and the fact of the involvement of regular forces and the that ballistic missile fl ight times between the Indian Air Force. In fact, Kargil shows that capitals are under ten minutes.52 The media low-intensity confl ict cannot be deterred even are not the only ones worried about stability in if both sides have nuclear weapons although, the region and the potential for use of nuclear as will be shown in chapter 3, nuclear weapons weapons by India and/or Pakistan. The U.S. capability may have deterred certain types of director of central intelligence, George Tenet, conventional military responses to insurgency testifi ed to the U.S. Congress in early 2001 and terrorist operations. that: “…relations between India and Pakistan Policy makers and analysts – primarily in remain volatile, making the risk of war India – argue that the United States is overly between the two nuclear-armed adversaries worried about crisis stability and the use of unacceptably high.”53 The war in Afghanistan nuclear weapons.54 Their arguments take a

18 Nuclear Stability in South Asia number of forms. One is that the wars that Moreover, analysis of this sort implies a need have occurred between India and Pakistan for outside involvement, which India has resist- are not serious – mere “communal riots with ed particularly regarding the Kashmir issue tanks.”55 Presumably because the wars were and relations with Pakistan. Pakistan, on the not large and signifi cant numbers of soldiers other hand, appears eager to discuss these and civilians did not die, there is very little issues and involve the international commu- chance of escalation to the nuclear level. A nity. Underlying this willingness to engage are second argument is that the disputes between the following motives. First, Pakistan has tradi- India and Pakistan are between people, and tionally attempted to involve the international particularly militaries, who come from a com- community in its disputes with India, seeking mon background and tradition. Each side to balance the economic, geopolitical, and mili- therefore understands the other very well tary might of its neighbor. Second, after the and would not allow something as terrible coup, the military government has continued as nuclear escalation to occur regardless of to seek to engage with the outside world in their disputes. A third argument that has been almost any manner to shore up its legitimacy. presented, at least in India, is that nuclear While these are not arguments for or against deterrence is very robust on the subcontinent. the existence of a stability problem, they are Laying out a sweeping Indian nuclear doctrine views that must be taken into account as one of countervalue, devastating second strikes crafts policies to address nuclear safety and sta- against any country that uses nuclear weap- bility on the subcontinent. ons fi rst, analysts in New Delhi argue that While reassurances about nuclear stability Pakistan would never escalate to the nuclear are widespread, they are not at all certain and level.56 Essentially, nuclear use by Pakistan they are not immutable. In fact, Tellis’s analy- would equal its total destruction, so no Paki- sis (discussed above) was written before the stani leader would risk it. This explanation nuclear tests, and he notes that his conclusions then usually, and paradoxically, ends with the about ugly stability only hold if three under- assertion that Pakistan has an irresponsible lying conditions do not change: “Indian and fi rst-use nuclear policy and therefore should be Pakistani military capabilities remain rough- the focus of any efforts to improve stability. It ly similar to what they are now (in relative is unclear, therefore, whether the Indians truly terms), the domestic political environment in believe that Pakistan will be deterred by New South Asia remains reasonably stable, and Delhi’s nuclear capabilities or not. there is no dramatic shift in the regional bal- Finally, from a policy-making perspective, ance of power.”57 It could be argued that two it is important to understand how both India of these conditions have indeed changed in the and Pakistan view this foreign worry about the intervening four years and that the other could region being a fl ashpoint and there being a dan- change substantially over the next fi fteen years ger of unintentional nuclear use or escalation – approximately the period of time for this to a nuclear exchange. To New Delhi, such dis- analysis. cussion and analysis have a condescending and First, India and Pakistan’s relative conven- colonial tone, implying that India and Paki- tional military capabilities continue to diverge. stan cannot manage nuclear weapons or crises. The increases in India’s defense budget, its

The Tests and Their Aftermath 19 increasing access to Russian high technology The line that the outside world should not weapons, and progress in its own research and worry, put forward by some Indian analysts, development are all helping it slowly to pull does not ring true for a number of reasons. away from Pakistan in terms of relative mili- First, the common military background and tary strength. Whether the gap will continue history are not at all comforting given the to grow or whether it will grow to the degree history of wars and crises in the region. Look- that it undermines Tellis’s conclusions is open to ing at other examples, common heritages and debate, but the situation is dynamic. The ongo- military backgrounds and training have done ing military supply relationship between China little to prevent wars in other regions of the and Pakistan, especially in nuclear and missile world, the former Yugoslavia being only the technology, also adds to the uncertainty of the most recent example. As will be detailed later situation. Second, the domestic political envi- in chapters 3 and 4, the presumption of com- ronment in South Asia is not stable. The mil- mon backgrounds and thinking has led to some itary coup in Pakistan and the uncertainties dangerous mirror-imaging in crises, leading about the restoration of civilian rule (and the not to their resolution but to their escalation. instability that may come with it) are one source Second, the robustness of nuclear deterrence of instability. The ongoing tensions between the is arguable at best and highly suspect at worst. current leadership and many of the Islamic fun- As Keith Payne notes, deterrence policies damentalist groups, exacerbated when Mush- can and have failed for numerous reasons, arraf sided with the United States in the war ranging from misapplication to their being on the Taliban, could still explode at any point. applied against actors who operate outside the While the BJP strengthened its control in the boundaries assumed in deterrence theory.58 most recent elections, Indian politics remain Theoretically, it has been argued that nuclear unpredictable, and Prime Minister Vajpayee’s deterrence is quite robust generally and that health continues to be a matter of concern. more nuclear weapons will provide stability in When he eventually passes from the scene, it is relationships that had been unstable previous- diffi cult to predict the course of the BJP, which ly.59 However, this theory has been challenged contains much more nationalistic and hard-line with theories of organizations, bureaucratic elements. Conversely, the parties in the coalition politics, and the role of individuals.60 Either with the BJP have aspirations to the prime of these theoretical perspectives could provide minister post, and many of them are not as the underpinning for policy choices. However, nationalistic, so changes of all types are possible. enough historical examples exist of the types Third, the interplay of potential alignments and of factors that undermine deterrence policies antagonisms among China, India, Pakistan, and that it appears prudent to err on the side of the United States could signifi cantly shift the caution and assume that deterrence may not regional balance of power in the coming years be perfect.61 This prudent approach is strongly in ways that could either be stabilizing or desta- supported by the more general literature on cri- bilizing. The role of the United States in the sis management and the concerns about the region, particularly in a post-Taliban Afghani- types of misperceptions that are common dur- stan, is still unclear and could have a destabiliz- ing crises.62 ing effect as easily as a stabilizing one.

20 Nuclear Stability in South Asia Finally, it should be noted that both Wash- ington and the international community are concerned about the safety and security of nuclear weapons in South Asia as well as about the likelihood of escalation in crises. During the Clinton administration, Washing- ton engaged in discussions about these issues with both New Delhi and Islamabad in bilat- eral dialogues. After September 11, the Bush administration stepped up the dialogue on safe- ty and security with Pakistan.63 However, this is only part of the picture, and it does not appear that the quickly set up discussions made any progress on these sensitive issues. To devel- op and implement successful policies to address safety, security, and particularly stability, the United States needs to get more specifi c and develop an integrated long-term plan. First, the emphasis on the long-term nonproliferation or unproliferation goal versus the goal of improv- ing safety and stability should be adjusted in favor of the latter. Second, the United States can and should become more specifi c in its dis- cussion of ways in which India and Pakistan could improve safety, security, and stability, particularly in times of crisis. Even the discus- sions with Pakistan in the fall of 2001 were apparently limited by old interpretations of U.S. and international laws.64 Such limitations need to be reviewed. Since Washington and the international community are already engaged with India and Pakistan on these issues, it would serve them well to get specifi c and to do so before decisions in New Delhi and Islamabad are made about doctrines and force structures or before the next crisis arises. (See chapter 5 for specifi c policy recommendations in this regard). To get specifi c, Washington needs to have a better sense of the possible futures of Indian and Pakistani doctrines and postures. This is the focus of chapter 2.

The Tests and Their Aftermath 21 Endnotes for Chapter One

1 John F. Burns, “Pakistan Atom Experts Held Amid Fear of Leaked Secrets,” New York Times, November 1, 2001, p. A1. 2 Yvonne Abraham, “Pakistani Questioned in Possible Nuclear Plot,” Boston Globe, November 18, 2001. 3 James Risen and Judith Miller, “Pakistani Intelligence Had Ties to al-Qaeda, U.S. Offi cials Say,” New York Times, October 29, 2001, p. A1. 4 Celia W. Dugger, “India Shells Kashmir Area, Imperiling Delicate Balance,” New York Times, October 16, 2001, p. B3. 5 Indian Embassy to the United States, press statement, May 11, 1998, http://www.indianembassy.org/pic/ PR_1998/May98/prmay1198.htm. 6 John F. Burns, “Nuclear Anxiety: The Overview; Pakistan, Answering India, Carries Out Nuclear Tests, Clinton Appeal Rejected,” New York Times, May 29, 1998, p 1. 7 George Perkovich, India’s Nuclear Bomb: The Impact on Global Proliferation (Berkeley: University of California Press, 1999), 4. 8 Dennis Kux, The United States and Pakistan, 1947-2000: Disenchanted Allies (Washington, D.C.: Woodrow Wilson Center Press), 308. 9 U.S. Department of State, “The Pakistani Nuclear Program,” secret memorandum dated June 23, 1983, declassi- fi ed. http://www.gwu.edu/~nsarchiv/NSAEBB/NSAEBB6/ipn22_1.htm 10 United Nations Security Council, Resolution 1172 (1998), S/RES/1172, June 6, 1998. 11 Hilary Synnott, The Causes and Consequences of South Asia’s Nuclear Tests, Adelphi Paper 332, International Institute for Strategic Studies (London: Oxford University Press, 1999), 27. 12 Deputy Secretary Strobe Talbott, “U.S. Diplomacy in South Asia: A Progress Report,” address at the Brookings Institution, Washington D.C., November 12, 1998. 13 Ibid. 14 Bureau of South Asian Affairs, U.S. Department of State, “Glenn Amendment Fact Sheet,” Washington D.C., September 2000. 15 U.S. Senate Subcommittee on Near Eastern and South Asian Affairs, Committee on Foreign Relations, Assistant Secretary of State for South Asian Affairs Karl Inderfurth, “India and Pakistan: What Next?” July 13, 1998. 16 Bureau of South Asian Affairs, U.S. Department of State, “Glenn Amendment Fact Sheet,” Washington, D.C., September 2000. 17 “Japanese Help Sought for Lifting of Sanctions,” Dawn, the Internet Edition, July 14, 2001, http://www.dawn.com/2001/07/14/nat4.htm. 18 “It was Clinton, not Bush, Who Lifted Sanctions,” The Times of India Online, February 4, 2001, http:// www.timesofi ndia.com/040201/04indi3.htm. 19 “No Honeymoon Likely for New President’s Foreign Policy Team,” New York Times, January 20, 2001. 20 Tahir Mirza, “Lifting of Economic Sanctions Advocated: US Offi cial for Better Ties Between India and Pakistan,” Dawn, the Internet Edition, May 19, 2001, http://www.dawn.com/20001/05/19/top1.htm. 21 Agence France Presse, “Sanctions on India May Be Lifted, Joint Chiefs of Staff Chairman Suggests,” Washing- ton Times, July 20, 2001, p. 15. 22 Alan Sipress, “U.S. Seeks to Lift Sanctions on India,” Washington Post, August 12, 2001, p. A1. 23 Jane Perlez, “U.S. Plans End For Sanctions on Islamabad,” New York Times, September 22, 2001, p. 1. 24 Sharat Pradham, “Nuclear Powers Have to do Something More Before India is Ready to Sign CTBT: PM,” Rediff on the Net News September 11, 1998, http://www.rediff.com/news/1998/sep/11ctbt.htm.

22 Nuclear Stability in South Asia 25 A number of former Indian nuclear scientists have called for further testing, including An Prasad, M. R. Srinivasan, P. K. Iyengar, and R. Chidambaram. Andrew Koch, “India, Pakistan: Nuclear Arms Race Gets off to a Slow Start,” Janes Intelligence Review, January 2001, p. 40. P. K. Iyengar “Nuclear Nuances,” The Hindu, August 22, 2000. 26 R. R. Subramanian, “A Technical Note on the Tests, Appendix One” in Amitabh Mattoo, ed., India’s Nuclear Deterrent Pokhran II and Beyond, New Delhi: Har-Anand Publications Pvt. Ltd., 1999, 351. V.R. Raghavan, “Dangerous Nuclear Uncertainties,” The Hindu, March 13, 2000. 27 Gregory S. Jones, From Testing to Deploying Nuclear Forces: The Hard Choices Facing India and Pakistan, issue paper, IP-192, RAND, Project Air Force, http://www.rand.org/publications/IP/IP192/. 28 Andrew Koch, “Nuclear Friction,” Janes Intelligence Review, December 2000, p. 24. 29 Ambassador Robert T. Grey, Jr., United States Representative to the Conference on Disarmament, “U.S. Urges Disarmament Conference to Launch Statement,” Geneva, February 15, 2001, http://www.fas.org/nuke/control/fmct.news/treaty-fmct-010215.htm. 30 Authors’ interviews with State Department offi cials, January 2000 and March 2001. 31 Michale Kraig, “The Indian Drive towards Weaponization: Policy Statements,” analysis for Federation of American Scientists, http://www.fas.org/nuke/guide/india/missile/agni-improvements.htm. 32 Barry Bearak, “India Tests Missile Able to Hit Deep Into Neighbor Lands,” New York Times, April 12, 1999. Associated Press, “Report: India Tests Missile,” January 17, 2001. 33 Both India and Pakistan use the term “inducted” to connote the turning over of a system to the armed forces. Once inducted, a system may be put into the reserves (stored) or may be put into unit sets, deployed, and exercised regularly. “Agni-II now ready for induction,” The Hindu, January 26, 2001. 34 Reuters, “India to Build New Version of Agni Missile,” February 8, 2001, http://www.insidechina.com/ news.php3?id=285432. “India denies plans to launch ICBM,” The Times of India Online, May 6, 2001 http:// www.timesofi ndia.com/060501/06indi12.htm. 35 Pravin K Sawhney, “Pakistan Scores over India in Ballistic Missile Race,” Jane’s Intelligence Review, November 2000, p. 35. 36 Shakil Shaikh, “Pakistan Begins Serial Production of Shaheen-I Missiles,” The News, September 21, 2000. 37 B. Muralidhar Reddy, “Pak. Claims Shaheen Inducted into Army,” The Hindu, January 26, 2001. 38 Neil Joeck, “Nuclear Developments in India and Pakistan,” Access Asia Review, no. 2 (July 1999): 38. 39 See: http://www.stimson.org/cbm/sa/drs/lahore.htm, for documents from the Lahore meeting. 40 The Kargil Review Committee, From Surprise to Reckoning: The Kargil Review Committee Report (New Delhi: Sage Publications, 1999), 69. 41 Barry Bearak, “Leader of Pakistan Agrees to Summit Talks With Rival India,” New York Times, May 29, 2001. 42 DH News Service, “Musharraf’s Letter of Invitation to PM Handed Over to Government,” New Delhi, July 27, 2001. http://www.deccanherald.com/deccanherald/july28/iletter.htm. 43 Samina Ahmed and David Cortright, South Asia at the Nuclear Crossroads, policy report and joint publication of the Managing the Atom Project at Harvard University, the Fourth Freedom Forum, and the Joan B. Kroc Institute for International Peace Studies at the University of Notre Dame, March 2001. 44 Strobe Talbott, “Dealing with the Bomb in South Asia,” Foreign Affairs 78, no. 2 (March/April 1999): 119. 45 Ibid., 112. 46 Scott D. Sagan, “Why Do States Build Nuclear Weapons? Three Models in Search of a Bomb,” International Security 21 no. 3 (winter 1996/97). 47 “Japanese Help Sought for Lifting of Sanctions,” Dawn, The Internet Edition, July 14, 2001, http://www.dawn.com/2001/07/14/nat4.htm. 48 Miriam Rajkumar, “Pakistan’s Sanctions Waivers: A Summary,” Carnegie Analysis, October 29, 2001, http://www.ceip.org/fi les/nonprolif/templates/article.asp?NewsID=1672.

The Tests and Their Aftermath 23 49 Eric Schmitt, “U.S. Says Aid To Pakistan Won’t Include F-16 Fighters,” New York Times, November 13, 2001. 50 Joseph Kahn, “U.S. Is Planning an Aid Package for Pakistan That Is Worth Billions,” New York Times, October 27, 2001, p. B4. 51 Douglas Frantz, “U.S. and Pakistan Discuss Nuclear Security,” New York Times, October 1, 2001, p. A3. 52 R. Jeffrey Smith, “A Feared Scenario Around the Corner: Preventing Nuclear Strike Capability Has Been a U.S. Goal for Subcontinent,” Washington Post, May 14, 1998, p. A29. 53 George J. Tenet, director of Central Intelligence, statement before the Senate Select Committee on Intelligence on the “Worldwide Threat 2001: National Security in a Changing World,” February 7, 2001, http://www.odci.gov/public_affairs/speeches/UNCLASSWWT_02012001.htm. 54 Authors’ interviews in New Delhi and Islamabad with a range of senior government offi cials and defense and military analysts, September 2000. 55 Authors’ interview with Bharat Karnad in which he cited this description by an Indian major general, September 19. 2000. 56 Author’s interviews with Bharat Karnad, September 19, 2000, and with K. Subrahmanyamu (chairman of NSAB), September 21, 2000. 57 Ashley J. Tellis “Stability in South Asia,” Documented Briefi ng, RAND Arroyo Center, prepared for the United States Army, 1997, 31. 58 Keith B. Payne, Deterrence in the Second Nuclear Age, (Lexington: University of Kentucky Press), 1996, 100. 59 Kenneth N. Waltz, The Spread of Nuclear Weapons: More May be Better, Adelphi Paper no. 171, International Institute for Strategic Studies (London: Oxford University Press, 1981). 60 Scott D. Sagan, “The Perils of Proliferation Organization Theory, Deterrence Theory, and the Spread of Nuclear Weapons,” International Security 18, no. 4 (spring 1994) and Scott D. Sagan, “The Perils of Prolifera- tion in South Asia,” Asian Survey, forthcoming. 61 Payne, 100-119. 62 See Robert Jervis, Perception and Misperception in International Politics, (Princeton: Princeton University Press, 1976) and Gordon Craig and Alexander George, Force and Statecraft: Diplomatic Problems of Our Time, 3rd ed. (New York: Oxford University Press, 1995). 63 Steven Mufson, “U.S. Worries About Pakistan Nuclear Arms; Offi cials Try to Guard Against Arsenal, Radioac- tive Material Going to Terrorists,” Washington Post, November 4, 2001, p. A27. 64 Douglas Frantz, “U.S. and Pakistan Discuss Nuclear Security,” New York Times, October 1, 2001, p. A3.

24 Nuclear Stability in South Asia CHAPTER TWO

Nuclear Force Structures & Doctrines

THE NUCLEAR WEAPONRY that the two rival states will acquire in the future and the evolving principles that will underwrite their capabilities will shape the contours of strategic and crisis stability. This chapter examines the vari- ous alternative nuclear postures and doctrines that each side could pursue over the course of the next fi fteen years. Each possible force structure and doctrine of India and Pakistan will exert different pressures on the calculus of the two states in times of crisis. Indeed, certain mixes of nuclear capabili- ties may prove far more stabilizing or destabilizing than others. Therefore, setting forth a series of alternative nuclear force postures and doctrines can contribute to an understanding of the implications of such nuclear postures for crisis management.

Nuclear Stability in South Asia 25 As a fi rst set of analytical inputs, this chap- the political elite and the military leadership ter examines 1) external and internal forces becomes crucial in times of crisis and confl ict. that could drive the nuclear force postures in It is still unclear whether the civilians have South Asia; 2) emerging Indian and Pakistani clearly mapped out a transfer of authority to force structures and doctrines; 3) several plau- their military counterparts that is responsive sible nuclear futures; and 4) implications of the to the uncertainties of crisis. Pakistan, in con- various force structures for strategic stability. trast, is in the fi rm grip of a military regime. The timeline for analysis is the period through Moreover, the armed forces have traditionally 2015. While crises and confl icts could cer- retained almost absolute control over the nucle- tainly erupt well before 2015, this timeframe ar weapons program with some minor role for provides a useful planning parameter against the civilian bureaucracy, a phenomenon deeply which to measure the evolution of Indian and rooted in the nation’s history.1 Even if civilian Pakistani nuclear programs. Moreover, the authority is restored, the military’s infl uence nuclear futures of India and Pakistan will like- over national security affairs will continue ly look far different from today. Therefore, and the armed forces will retain the option of projecting alternative nuclear futures for both reemerging from the barracks. Beyond these states will help clarify the level and type of sweeping observations, narrow sets of internal instability that might emerge in the South infl uences will shape each country’s nuclear Asian nuclear balance. policies in unique ways. In India, political upheaval and major Accelerators and Constraints on realignments in the Lok Sabha, the Indian Nuclear Developments parliament, could directly infl uence govern- Several different factors will help to determine ment policy on nuclear capabilities. It is widely the direction and pace of nuclear developments accepted that the rise of the BJP, a Hindu-based in India and Pakistan: domestic politics, strate- nationalist party, had a decisive infl uence on gic competition, technological considerations, India’s decision to break out of its self-imposed economic considerations, and the policies of nuclear constraints.2 Fulfi lling India’s quest the international community. for nuclear power status was a central compo- nent of the BJP’s ideology, which was driven in Domestic Politics large part by a virulent brand of nationalism. First, domestic politics will have a critical Indeed, the BJP’s election manifesto included a infl uence on how India and Pakistan pursue promise to induct nuclear weapons. Some spec- their nuclear postures and on the future course ulated that the party sought to use nuclear tests of the nuclear balance. In India, absolute civil- to increase its popularity and cement its politi- ian supremacy over the military is a deeply cal position. It is clear, then, that the rise and embedded political tradition that is virtually fall of political parties in national elections, ironclad. This has led to tight civilian control an inherently unpredictable affair, could bring over decision-making processes and weapons about shifts in nuclear policy. development that are inherently military in Beyond party politics, other powerful ele- nature. While less problematic in peacetime, ments in India will continue to lend momentum the issue of command and control between to expanding the nuclear program. Outspoken

26 Nuclear Stability in South Asia and infl uential political fi gures and individuals In Pakistan, the relative longevity of the have argued persistently in favor of new stra- military regime suggests that a quick return tegic thinking for India that would presumably to “normalcy” remains a questionable prop- justify a robust nuclear posture. India’s foreign osition. While the military government has and defense minister, Jaswant Singh, has been indicated that it intends to step down in among the most vocal. He has argued that the 2002, its infl uence will almost certainly be felt lack of a strategic culture and tradition in India beyond the general elections. General Pervez has proved devastating to the nation’s security Musharraf’s self-appointment as president in and interests, lamenting: June 2001 has further entrenched the power of In the last half century the Indian polit- the military in the political system. This very ical-military leadership has not displayed uncertainty increases the unpredictability of the required strategic sense. Indian national Islamabad’s evolving nuclear posture. Should interests were not adequately served…What, Musharraf give up his authority at some future therefore, is the legacy of the past fi fty years? date, the transition to democratic governance An absence of certainties in security-related could inaugurate a period of political confu- issues; no established land boundaries; an sion or even instability. This is particularly absence of a secure geopolitical environ- true given the crisis in South Asia that began ment; a devaluation of India’s voice in global in fall 2001. Despite this crisis, Musharraf has affairs and worrisomely, not even a begin- maintained that he will stick to his promised ning of any institutional framework for con- deadline for a return to civilian rule. Converse- ceptualizing and managing the country’s ly, a failure to make the shift back to civilian defence.3 rule could spark further instability. The res- toration of civilian authority, if and when He concludes that India would have to if occurs, will not necessarily guarantee clar- break out of its self-imposed strategic trap. In ity or predictability for Pakistan’s nuclear the nuclear realm, he asserts, “An examination program. Given the nation’s nascent state of of the fi rst fi fty years of Indian independence democracy, which has been punctuated by mil- reveals that the country’s moralistic nuclear itary coups and political upheaval, a civilian policy and restraint did not really pay any regime may be subject to even greater vola- measurable dividends.”4 Given this, he argues, tility. Indeed, Prime Minister Nawaz Sharif’s India had to conduct the tests in order to decision to engage in tit-for-tat nuclear tests ensure that its nuclear option did not erode in was largely a result of political maneuvers by the future. Moreover, the scientifi c communi- opposition parties and the military to force his ty in India, or what Itty Abraham has termed hand.6 Sharif’s efforts to strengthen the prime the “scientifi c enclave,” created a sprawling minister’s offi ce through extra-constitutional nuclear infrastructure that set the stage for measures before his ouster in 1999 also sug- conducting the tests. India’s Defense Research gests that ambitious personalities could have a and Development Organization (DRDO) has direct impact on Pakistan’s nuclear decisions. been a powerful force behind the movement Among all domestic political infl uences, toward testing.5 These vested interests will virulent nationalism and sectional politics certainly drive India’s future nuclear develop- (both are at times intertwined and thus indis- ments in weaponry and doctrine.

Nuclear Force Structures and Doctrines 27 tinguishable) are perhaps the most volatile in tries (including landmass, population, military India and Pakistan. National pride and notions power, and economic health) between the two of prestige could drive the two states to pursue nations. Pakistan’s geography poses a major riskier courses of action. The jubilation in the defense planning challenge. In contrast to streets after the Indian tests demonstrated the India’s vast strategic depth, Pakistan suffers widely shared perception that nuclear weapons from limited defensive depth and is vulnerable equated to great-power status. In the after- to being cut in half at its mid point. Its con- math of the Indian tests, 70 percent of urban ventional military capabilities are far inferior Pakistanis supported an immediate response to India’s. In addition to these tangible mea- in kind.7 Moreover, incendiary events in one sures of power, both sides continue to suffer country could directly shape threat perceptions from severe misperceptions about the inten- and even prompt a corresponding response in tions and capabilities of the other. Posturing by the other. Whipped up xenophobia and reli- both sides over the years has further clouded gious or nationalistic fervor among the public a true picture of each. So long as this fog per- could infl uence government decisions and hold sists, both sides will likely base their nuclear each nation’s policy hostage to public demands planning calculations on worst-case scenarios. for reprisals. In India, clashes between Hindu Should either side unexpectedly stumble across nationalists and the Muslim community could information that fundamentally challenges its spark tensions between Islamabad and New conventional wisdom of the other, a sudden Delhi. The destruction of the Ayodhya mosque acceleration of capabilities is a likely out- by Hindu radicals underscores the dangers of come. Moreover, “the fear and uncertainty such religious confrontation. In Pakistan, the induced by such actions could precipitate coun- rise of increasingly powerful fundamentalist tervailing responses that set off a destabilizing Islamic elements within the Pakistani military, action-reaction spiral between both states.”8 state, and society could color decision-making While popular perceptions have focused processes. President Musharraf’s crackdown primarily on the bilateral Indian-Pakistani on the fundamentalist elements in the wake of dimension of the nuclear standoff, other, broad- September 11 has set these movements back, er strategic factors could also drive nuclear but it is possible that it might eventually result developments in South Asia. For instance, in a backlash. India confronts a dual threat from Pakistan and China. Former Indian defense minister Strategic Competition George Fernandes proclaimed that China was Second, historical strategic competition, includ- India’s enemy number one. This perception ing the intractable dispute over Kashmir, is is derived both from history and contempo- a defi ning element in their respective nuclear rary developments. China’s punitive invasion programs. Divergent strategic considerations of India in 1962 and its 1964 nuclear test left could drive India and Pakistan to pursue rad- indelible imprints on the Indian psyche. More ically different nuclear force structures. The recently, China’s economic growth, military most prominent strategic realities that will modernization, potential geostrategic interests determine the future paths of their nuclear in the Indian Ocean (a vital passageway for postures are the fundamental power asymme- energy supplies for both China and India), and

28 Nuclear Stability in South Asia efforts to shore up ties with India’s neighbors India, and Pakistan and the potential strategic are all worrying trends for New Delhi. Most competition between Beijing and New Delhi vexing, China and Pakistan have long shared could set off a chain reaction that severely com- a broad strategic alliance directed at coun- plicates Islamabad’s nuclear planning. Indeed, terbalancing India. Beijing’s proliferation of India’s perception of a Chinese threat could nuclear weapons-related and missile technol- directly infl uence the future path of Pakistan’s ogies to Pakistan has been at the center of nuclear arsenal. However, Islamabad’s inter- India’s immediate security concerns. China’s nal strategic calculations qualify this grim decades-old proliferation to Pakistan has been assessment. One senior Pakistani nuclear pol- far more intense and extensive than that of icy maker argued that the future of India’s North Korea. Moreover, Islamabad has played nuclear forces would not affect Islamabad’s and will continue to play the “China card” nuclear posture. He noted that the number of at every opportunity to keep India strategical- targets (presumably civilian centers) in India ly off balance. This card, however, does not would not change, implying a countervalue always have the expected effect, and at times strategy.9 In other words, internal assessments Islamabad has overestimated its weight. of what constitutes suffi ciency could offset The perceived China threat will fuel fears externally driven factors. in New Delhi that Pakistan might be embold- ened to take risky actions to destabilize the Technological Considerations subcontinent. In the longer term, Chinese and Third, several technological considerations Indian military modernization programs will will infl uence the scale and sophistication of enable the two states to project power hitherto nuclear weapons and associated delivery sys- beyond their reach that could pit them against tems. New Delhi’s dogma of self-reliance has each other over strategic regions such as the long been a double-edged sword. After early eastern Indian Ocean or the South China Sea. reliance on foreign technology transfers for its Whether China poses a genuine conventional nuclear and missile programs, India’s emphasis military or nuclear threat is debatable. Bei- on indigenous production has often hindered jing has demonstrated no visible territorial rapid development. Despite an impressive tech- ambitions since 1962 and harsh geographic nological base, the nation’s industrial capacity realities have long separated the two powers remains backward compared to world-class from direct strategic competition. Neverthe- standards. Many of India’s domestically pro- less, the Chinese threat provides a compelling duced weapons, including those under foreign and convenient political rationale for New license, suffer from quality control problems Delhi to continue its nuclear program. and fail to meet the military’s requirements.10 Pakistan faces no such two-front dilemma, In the nuclear weapons realm, technologies and its strategic focus is locked exclusively on to design and construct warheads to sit atop India. Islamabad’s nuclear force structure will ballistic missiles are very diffi cult to master. be designed to deter New Delhi’s nuclear use Without signifi cant foreign assistance, India against Pakistan and to counter India’s con- could face an uphill battle to fi eld an advanced, ventional military superiority. However, the reliable nuclear force. Pakistan has tradition- uneasy triangular relationship between China, ally relied on foreign sources for its nuclear

Nuclear Force Structures and Doctrines 29 program, particularly from China and, to ways. Advances in space launch vehicles could a lesser extent, North Korea. As a result, improve India’s ballistic missiles. The success- Islamabad has been able to leapfrog various ful launches of communications satellites in technologies and even acquire fully assembled recent years also create opportunities for devel- missiles from China. Despite Beijing’s repeat- oping military reconnaissance, command and ed pledges to curtail its proliferation behavior, control, communications, and battle damage China has often reneged on its promises. assessment, all of which are critical compo- As noted in chapter 1, future nuclear tests nents of a robust nuclear posture. to improve weapons design and reliability are also a major technological issue for both coun- Economic Considerations tries. Given the strong political symbolism Fourth, while the level of economic growth attached to testing, nuclear detonations for will frame the resources devoted to devel- whatever purpose are an inherently interac- oping nuclear weapons, the historical record tive process. Neither side will be able to ignore indicates that economic under-performance them and refuse to respond in kind. The tit- has not deterred past advances in both coun- for-tat tests in 1998 demonstrate this powerful tries. India’s slow growth before the economic domino effect. India may also perceive a great- reforms of the 1990s did not prevent expendi- er need to test in order to validate certain tures on a sprawling nuclear weapons program. technologies that cannot be tested through sim- Pakistan has long suffered from serious struc- ulations. Pakistan, with the help of China, has tural and systemic economic problems leading had the opportunity to test its capabilities and to persistently sluggish growth. Yet Islamabad’s may therefore feel less compelled to test. nuclear weapon and ballistic missile programs The growth and level of technological have enjoyed fi rst call on fi nancial resources sophistication in the private commercial sec- and have been largely insulated from economic tor could have spin-on effects for the nuclear pressures. As a result, Pakistan has been able forces in India and, to a lesser extent, Pakistan. to amass an impressive nuclear arsenal. Since China’s space program demonstrates the close New Delhi abandoned economic planning in linkages between commercial and military the early 1990s, healthy growth has replaced space: Beijing’s Dong Feng ICBM launchers decades of under-performance. India will con- became the basis for its commercial boosters. tinue to enjoy a slightly lower rate of 5 percent These benefi ts arising from civil-military syn- to 6 percent growth in the coming years assum- ergies will hinge on continued economic ing that reforms continue. India’s economic size growth, which puts Pakistan at a relative in absolute terms also means that it can spend a disadvantage. Cutting edge research and devel- relatively smaller amount of its GDP to compete opment from the private sector could help the with Pakistan (but probably not with China). military to leapfrog generations of technolo- In contrast, Pakistan struggles with immense gies. The telecommunications industries in external debt problems estimated at $34.8 India and Pakistan could provide advanced billion for 2000 – a staggering sum given communications equipment for secure nuclear its small and enfeebled economy – which command and control. India’s space program pose a daunting economic challenge for Islam- could directly benefi t the military in several abad.11 Pakistan has relied heavily on foreign

30 Nuclear Stability in South Asia assistance and the infusion of funds from drivers enumerated above. As Stephen Cohen international fi nancial institutions, particular- forcefully states, “Indian and Pakistani nuclear ly the IMF, in order to keep the economy policies are driven by calculations of national running. Economic forecasts predict that Pak- security and pride, and it will take more than istan’s economy will slow in the coming years. lectures, the blunt application of sanctions, or Consequently, India will be less constrained to the example of other states which have chosen pursue independently more ambitious nuclear not to build nuclear weapons, such as South modernization programs than Pakistan. This Africa or Argentina, to signifi cantly infl uence suggests that Islamabad will continue to look them.”12 to outside powers – such as China and North The multifaceted nature of the accelerators Korea – for access to technologies and weap- and constraints on nuclear weapons develop- onry. ment in South Asia highlights the wide range of alternative futures for each state. This very External Considerations unpredictability makes policy formulation for Finally, the policies of the United States and the United States an inherently complex but other major powers toward the region as nevertheless necessary task. In order to miti- well as the common position of the inter- gate the uncertainties for policy makers, we national community more broadly, can have must not only set forth the current state of an infl uence (albeit less dramatic) on nuclear nuclear weapons programs and doctrines for developments in South Asia. Before September both countries but also develop alternative 11, the Bush administration’s regional policies futures that they might pursue. Such an analyt- were evolving slowly. However, the trend was ical exercise can highlight both the stabilizing toward lifting sanctions on India and Paki- and destabilizing aspects of the current and stan and engaging more comprehensively on possible future nuclear balance. the issues of nuclear stability. After Septem- ber 11, these policies evolved rapidly, sanctions Current Nuclear Doctrines and were lifted, and U.S. engagement – particularly Capabilities in South Asia with Pakistan – skyrocketed. Despite this dra- The secrecy surrounding the nuclear programs matic change in U.S. involvement in the region, in India and Pakistan poses signifi cant method- it should be noted that in the past Washington ological and intelligence problems in analyzing and other capitals have had very little direct doctrines and capabilities. Outside observers political infl uence over Indian and Pakistani simply do not have adequate and reliable infor- policies on nuclear issues. The tests themselves mation to account for the number and types of attest to the collective impotence of the inter- weapons they possess, to categorize their over- national community to change the behaviors all capabilities, or to measure their levels of of both countries on matters that they consider readiness. Moreover, there are important diver- vital to their national security interests. Inter- gences over defi nitions and interpretations of national backlash after certain nuclear moves terms related to doctrine and capabilities. India may at most give pause to both governments, and Pakistan have already developed a com- but external diplomatic and economic consid- mon vocabulary to describe their respective erations will always be subordinate to the other nuclear strategies, postures, and doctrines that

Nuclear Force Structures and Doctrines 31 often differs from Western defi nitions and con- twenty nuclear-armed intercontinental ballistic ceptions. Given this difference, it is important missiles. to clarify and delineate the terminologies that The term “deployment” in the South Asian appear in this analysis. The following discus- context poses another analytical dilemma that sion analyzes “credible minimum deterrence” requires further clarifi cation. First, it is dif- and “deployment” both to highlight the nuanc- fi cult to verify deployment in either country es of these terms and to avoid defi nitional based on open sources. Second, some analysts pitfalls as they appear in this study. believe that neither country has the actual Both Indian and Pakistani offi cials have capacity or the will to deploy its arsenal. They proclaimed that their nuclear strategies are pre- contend that India and Pakistan would remain mised on “credible minimum deterrence.” What content with their respective declared nucle- does this mean? Unfortunately, this phrase has ar capabilities while continuing research and a variety of connotations and analysts still differ development without fi elding the weaponry. on its exact meaning. Both India and Pakistan Third, full deployment requires the integration have failed to provide a clear defi nition them- of several distinct components and stages. The selves. For the purposes of this chapter, it is warheads must be prepared for mating with the defi ned as the minimum nuclear force structure delivery systems and transported to the stag- necessary to deter a nuclear or overwhelming ing site. The warheads must then be placed on conventional military attack. Under the cred- the delivery systems. Command and control ible minimum deterrence strategy, the country arrangements, which could be either highly possesses a basic number of warheads and deliv- decentralized or tied closely to the national ery systems that are or can be targeted at the authority, then must be handed over to the adversary. The target set would consist largely military commanders operating the weapons. of population centers, but may also include some In the fi nal stage, the nuclear force would be infrastructure and military forces. The key dispersed to mobile or fi xed sites for launch. to credible minimum deterrence is determin- Depending on its strategy or readiness require- ing the essential numbers and types of nucle- ments, it is conceivable for either side to ar weapons, together with a set of targets that consider any of these phases as the “fully would deter the other side from resorting to deployed stage.” Without an extremely sophisti- force using nuclear and/or conventional weap- cated intelligence capability or highly intrusive ons. In order for the deterrent to remain cred- inspections, determining whether India or ible, a suffi cient number of nuclear weapons Pakistan has reached complete deployment must be able to survive a fi rst strike and then is an impossible task. To complicate matters have the ability to retaliate. Clearly, the required further, Indian and Pakistani offi cials have numbers and sophistication of the nuclear force avoided using the term “deployment.” Rather, will vary depending on the intended adversary. as mentioned in chapter 1, the authorities have As an example, China’s minimum deterrence preferred to describe their weapons as having against American nuclear weapons use has been been “inducted.” Both sides have remained premised on the ability to hold an urban center ambiguous about the exact defi nition of induc- such as Los Angeles hostage with its estimated tion. It is clear that concrete defi nitions of certain terms will remain elusive and observ-

32 Nuclear Stability in South Asia ers must be careful when interpreting Indian been virtually nonexistent in India’s statecraft, and Pakistani understanding of these terms. which is often said to be ad hoc in nature. Oth- ers point to the deeply embedded normative India considerations in India’s political culture.14 India’s nuclear doctrine remains in a state of According to Raja Menon, “Independent India transition and a comprehensive nuclear strate- started life with the belief that international gy has yet to emerge. Indeed, confl icting offi cial relations, like society, could be governed by statements on how the country intends to ethics and morality rather than practical real- harness its nuclear weapons program have ism.”15 Whatever the explanations proffered, emerged since the tests. Several explanations the secrecy surrounding India’s nuclear doc- can be attributed to this seemingly confused trine makes projections of its future force state of affairs and frustratingly slow pace of structure diffi cult. doctrinal development. First, New Delhi may At present, only a semi-offi cial draft doc- be continuing the policy of opaqueness that it ument has approached a full articulation of adopted after the fi rst “peaceful nuclear test” India’s future direction in nuclear doctrine. in 1974. This low-profi le approach sustains However, the subsequent controversy sur- a high level of uncertainty to keep its rivals rounding the report led the government to off balance. Second, nuclear doctrine must disavow its authoritativeness. In August 1999, constantly evolve to refl ect strategic realities. the National Security Advisory Board (NSAB), Finding a suitable nuclear doctrine that fi ts an organ of the newly reorganized National India’s unique strategic requirements will Security Council, issued a well-publicized and remain a constant challenge. Third, the cacoph- highly controversial draft report on India’s ony of (often confl icting) opinion among policy nuclear doctrine. Members of the NSAB makers, retired offi cials, and academics further included retired offi cials, military offi cers, clouds any clarity in India’s nuclear doctrine. and prominent academics. The content of the India’s democracy has enabled a variety of report initially caused a major stir in Paki- interests to voice their positions, which has in stan, the international community, and India turn generated a tremendous amount of impen- itself. The draft document outlined the main etrable intellectual “noise.” In many ways this contours of a proposed nuclear doctrine for noise provides a convenient cover and vehicle India:16 for denial and deception by overwhelming • A doctrine of credible minimum deter- observers with an overabundance of informa- rence, a phrase left undefi ned in the tion. Finally, some contend that India still document, that would be based on retal- lacks a coherent strategic tradition to guide its iation only; in other words, a no-fi rst thinking on nuclear strategy. Jaswant Singh use policy argues that India’s lack of a sense of history, • A credible, survivable, effective, secure, the absence of a “territorial consciousness,” the and safe nuclear posture persistence of foreign domination, and the pre- • Nuclear forces based on a triad of land-, occupation with internal affairs have diverted air-, and sea-based platforms intellectual energies from fostering a strategic • An extensive command and control culture.13 As a result, strategic thinking has system and an early warning system,

Nuclear Force Structures and Doctrines 33 including space-based assets, to support posture enables India to retaliate at a time these forces (days, weeks, or even months) and on a scale According to the document, the elements of its own choosing.17 India’s relative strate- together would enable India “to shift from gic invulnerability in terms of geographic size, peacetime deployment to fully employable forc- vast population, and large numbers of targets es in the shortest possible time…for rapid affords New Delhi such an unhurried option. punitive response.” Contrary to offi cial reas- These analysts further assert that New Delhi’s surances after the tests, the NSAB document force structure can be limited in size for both seemed to endorse an open-ended long-term Pakistan and China. They claim that India development path for India’s nuclear forces. It only needs to target an opponent’s largest essentially presented a wish list of capabili- population centers. In Pakistan’s case, the tar- ties that the advisory board hopes to acquire. geting of three major urban areas (Islamabad, Moreover, the report did not resolve or recon- Lahore, and Karachi) would probably suffi ce to cile many crucial issues confronting India. The infl ict unacceptable and devastating damage. document failed to specify details on command In China’s case, India cannot hope to com- and control, survivability, civil-military chain pete numerically or technologically in the short of command over nuclear use, and costs. The to medium term. Moreover, China poses no quest for “rapid punitive response” also sug- cross-border conventional military threat to gested a far more aggressive, sophisticated, the Indian heartland that would warrant a and decentralized capability, which provoked more robust posture. Until longer-range deliv- great global consternation. In response to the ery systems, such as the Agni-III, required to overwhelmingly negative domestic and inter- hit far-fl ung cities like Beijing become available national responses to the report, the Indian in India’s inventory, holding hostage large cit- government was careful to distance itself from ies in central China (Chengdu and Chongqing) the document. Indeed, to date, the draft has would be suffi cient. The leisurely pace with received no offi cial sanction. The Indian gov- which New Delhi has sought to develop its ernment may have concluded that in the future doctrine and force structure appears to con- it should avoid discussions regarding its nucle- fi rm a less ambitious national policy than the ar doctrine that could generate controversy. NSAB’s recommendations. This suggests that The views of other authoritative offi cials India will continue to engage in “creeping and scholars in and outside of India have both weaponization” until a force suffi cient to con- conformed to and differed with the NSAB’s stitute a nuclear posture emerges. report. Most individuals agree that India’s The country’s evolving command and con- nuclear posture centers on a no-fi rst use doc- trol will be a major determining factor in the trine and an assured second-strike capability sophistication and scale of India’s force struc- against soft targets. However, in contrast ture. At fi rst glance, this assessment appears to the NSAB’s proposals, a signifi cant num- counterintuitive. The evolving force struc- ber of Indian analysts discount the need to ture should logically dictate the command and build a large and sophisticated force for a control architecture. However, the inherent rapid, overwhelming response against a nucle- asymmetries in India’s civil-military relations ar attack. They argue that a second-strike will likely empower the command and control

34 Nuclear Stability in South Asia desiderata to drive force structure require- ture. Consequently, this element of India’s ments. At present, the democratically elected current doctrine will be generally conservative civilian government has deliberately chosen to in nature. Civilian dominance further rein- place a high premium on civilian dominance forces the widely shared belief that India’s over the nation’s nuclear command structure. forces should not be on a hair-trigger, launch- The bureaucracies governing defense and mili- on-warning footing. This suggests that in the tary policies are designed to ensure an absolute scenarios where nuclear weapons could be civilian control over the decision-making pro- used, there will be a measure of warning time cess and complete subordination of the military during which nuclear warheads could be mated services. This institutional framework is deep- with delivery systems and the launch author- ly rooted in the history and tradition of India’s ity passed on to the military commanders. This peculiar civil-military relationship. As a result, also implies that the Indian leadership has a any decision to use nuclear weapons rests high level of confi dence in the ability of the entirely with the prime minister and the cabi- evolving command and control system to sur- net. Both the civilians and the military have vive a nuclear attack and then to initiate the joint custody of the nuclear capabilities. Each authorization process to retaliate. oversees different components of the assets Data on India’s nuclear program has varied that constitute the nuclear force. The military wildly. The nuclear tests in 1998 provided a controls the delivery systems while the civil- glimpse of the types of warheads that India was ians safeguard the nuclear warheads. Should able to assemble. According to India’s Bhabha the need for nuclear use arise, the civilians Atomic Research Centre, the fi rst round of and the military would convene to integrate tests on May 11 purportedly included a 43-kilo- these separated elements of the force into an ton thermonuclear device, a fi ssion device with employable capability. The civilians would a 12-kiloton yield, and a sub-kiloton device. then incrementally decentralize the authority Two days later India detonated two devices for the military commanders to carry out the within the 0.2- to 0.6-kiloton range. If these missions. claims are to be believed, India is capable of So far, the civilian elites have deliberately producing a relatively wide range of nuclear obfuscated the exact nature of this rigid, highly warheads. However, independent seismic anal- centralized arrangement. Whether processes yses suggest that the actual yields were lower exist by which a transfer of authority to the than the offi cial pronouncements. A British military can be made in an orderly manner, report published in July 2001 claims that the particularly in times of crisis, remains an open combined yield of all the devices totaled closer question. This reluctance to clarify the chain to 20 kilotons, implying that the thermonuclear of command is partly driven by the self-rein- weapons failed to detonate fully.18 These coun- forcing impulse to maintain exclusive civilian ter estimates have in turn cast doubt on New control. The political elites worry that any Delhi’s ability to develop a hydrogen bomb in debate could weaken the rationale for this the near term.19 awkward system and, in turn, undermine civil- The state of India’s nuclear infrastructure ian dominance. They are therefore unwilling and stockpiles of fi ssile materials available for and afraid to decentralize the command struc- weaponization is equally unclear. With ori-

Nuclear Force Structures and Doctrines 35 gins dating back to Prime Minister Nehru’s program, it remains unclear how much of the fi rst comprehensive four-year plan on develop- fi ssile material has been converted into usable ing India’s nuclear power capabilities in 1952, warheads. In March 2000, an MSNBC report, India’s quest for a weapons program began well based on leaked U.S. intelligence, hinted at before 1970. 20 Analysts generally agree that India’s progress. It indicated that India might India already enjoys a self-suffi cient nuclear have completed only fi ve bombs. The assess- fuel cycle, with the full complement of natu- ment also claimed that Pakistan’s nuclear ral resources, nuclear research facilities, heavy arsenal could be far larger than previously water nuclear power plants, and fuel fabrica- reported, suggesting that Islamabad may have tion, enrichment, and reprocessing facilities.21 weaponized most of its fi ssile material at a India’s nuclear infrastructure has centered much faster rate than expected. How much of on plutonium reactors, although the country the information is true or relevant is debatable. has also made forays into uranium-enrich- As mentioned above, unlike Pakistan with its ment programs. It is believed that most of relationships with China and North Korea, India’s high-quality weapons-grade plutonium India does not currently have a major external is derived from unsafeguarded research reac- supplier of nuclear or missile technology. While tors.22 For example, the one hundred-megawatt it is widely assumed that India has the technol- Dhruva research reactor was the major source ogy to design and manufacture gravity bombs, of the fi ssile materials for the 1998 tests the lack of foreign assistance suggests that New (although other reactors may have contributed Delhi may be behind Islamabad in its ability as well). The ability to siphon off materials to miniaturize warheads for ballistic missiles from ostensibly civilian reactors for the weap- with a proven design. ons program in a short period of time is not India’s strategic delivery systems are at in doubt (it took only two months after the different stages of development and their opera- new BJP government came to power to con- tional status remains unclear. At present, New duct the tests). The ability to exploit its civilian Delhi’s frontline capabilities probably still cen- infrastructure means that India has a larger ter on its air leg. The British-made Jaguar potential stockpile of fi ssile materials than Pak- and the Russian-supplied MiG-27 Flogger are istan. There is still debate on whether India ground-attack fi ghters that can carry out nucle- used its unsafeguarded civilian power reac- ar missions with few modifi cations. Given that tors to extract weapons-grade plutonium for India may have had a nuclear weapon well its 1998 test. While estimates of accumulated before it developed ballistic missiles, it is like- stockpiles of fi ssile material available for wea- ly that its earlier devices were designed to be ponization have differed (at times drastically) air dropped. In contrast to the unreliable MiG, among various sources, reliable numbers have the Jaguar boasts an advanced suite of elec- generally converged on upwards of three tronics, sophisticated navigation and attack hundred kilograms of plutonium, enough to systems, and a larger combat radius. Among produce between sixty and seventy-fi ve early these assets, the Jaguar is perhaps best suited generation devices.23 for nuclear strikes although it is an aging Given that India has maintained a stran- platform. While India’s Su-30s, MiG-29s, and glehold on data regarding its nuclear weapons Mirage 2000s are designed primarily for air

36 Nuclear Stability in South Asia superiority missions, they could also be con- sion for the army, 250-kilometer version for verted to serve effectively as nuclear delivery the air force, and 350-kilometer version for platforms. In a journalistic account of India’s the navy).27 In August 2000, the Indian nuclear program, Raj Chengappa claimed that government apparently authorized the produc- India might have used a Mirage 2000 to drop a tion of three hundred nuclear-capable Prithvi dummy gravity nuclear bomb in a 1994 fl ight missiles.28 Given the limited range and ques- test.24 The uncertainties over the penetration tionable operational reliability of all three capabilities and survivability of fi ghter aircraft types, the Prithvis are widely believed to be after a preemptive strike have already forced stop-gap measures until the more robust capa- India to shift its force structure toward its land- bilities become fully deployable.29 Indeed, based ballistic missiles. for genuine operational usefulness, the Prith- The principal delivery systems for India’s vis would have to be deployed close to the strategic ballistic missile force will be the Pakistani border, a move that could provoke medium-to intermediate-range Agni series. The conventional or nuclear fi rst strikes. India does Agni-I was successfully tested in 1989 and not yet possess a sea-launched leg of the nucle- reportedly has a range of 1,500 kilometers with ar triad. In June 2001, a joint Indian-Russian a 1,000-kilogram payload. The Agni-II is cur- team of scientists successfully tested a verti- rently the most potent delivery system and has cally launched supersonic cruise missile, which apparently superseded the Agni-I. According to can be fi tted on board a ship or submarine. the DRDO, Agni-II has a 2,500-kilometer range However, given the persistent low levels of with a 1,000-kilogram warhead. India success- funding for the navy, substantive progress fully tested the missile in April 1999.25 With remains a long-term proposition. this missile, India can easily put at risk all major targets in Pakistan. India announced in May Pakistan 2001 that the missile is fully operational and Similar to its Indian counterpart, Pakistan’s would be deployed in the 2001-02 timeframe. nuclear doctrine is equally vague and not well India also formally approved the development understood. Three major forces have shaped and eventual deployment of the more capable the nation’s strategic doctrine. First, the birth Agni-III, which will have an estimated range of Pakistan in 1947 is an obvious contextual of 3,500 kilometers. This missile would enable starting point. War with India and the incom- India to hold at risk targets throughout China. plete partition are at the core of Pakistan’s However, some analysts believe that it may national identity, raison d’etre, and strategic take up to ten years before the missile enters culture. In short, India and its potential threat large-scale production.26 The media in late 1999 (both imagined and real) to Pakistan’s survival reported that New Delhi is also considering the lie at the heart of Islamabad’s strategic thinking development of ICBMs, the Surya, which would in both the conventional and nuclear realms. be designed to carry a nuclear warhead over Pakistan has consistently maintained that it 10,000 kilometers. developed a nuclear weapons program to coun- India currently operates three variants of ter India. Islamabad’s response in kind to the nuclear-capable, liquid-fueled, short-range the 1998 Indian tests and subsequent rhetoric Prithvi ballistic missiles (150-kilometer ver- on the need to restore regional military bal-

Nuclear Force Structures and Doctrines 37 ance conform with Pakistan’s obsession with weapons capability and thus to prepare for a India’s existential threat to the country. Sec- conventional strike. Second, Pakistan wants ond, unlike India’s defense community, which to have a limited capacity to deter or blunt is an amalgam of personalities and interests an overwhelming Indian conventional attack. (predominantly civilian) in and out of govern- Third, it wishes to have the capacity to use ment, Pakistan’s armed forces have dominated nuclear weapons as a demonstration instru- every aspect of strategic thinking. The pre- ment to internationalize the crisis and invite eminence of the military has in turn resulted outside intervention if the circumstances prove in a Pakistani nuclear doctrine and posture unfavorable to Pakistan. that are quite distinct from India’s (elaborated Subsequent offi cial statements also suggest- below). Third, Pakistan suffers from perenni- ed that Pakistan had adopted a doctrine of al military and economic weaknesses. These minimum credible deterrence but – in line vulnerabilities have forced the country to rely with its long-standing opaque posture – refused almost exclusively on international support for to specify or quantify what constituted this fi nances, technology transfers, and, in times of form of deterrence. The release of India’s crisis, intervention. Hence, Islamabad is more NSAB draft report deeply alarmed Islamabad. sensitive to currying favorable external per- The open-ended conclusions of the document ceptions of its behavior. Ironically, this has prompted Islamabad to issue several pointed led Pakistan to demonstrate relatively greater responses. Throughout 2000, Pakistani offi - responsibility and transparency in its nuclear cials declared that the country would need to posture and doctrine as a way of garnering develop a more proactive nuclear posture in favorable foreign perceptions. order to counter India’s conventional military Immediately after the 1998 tests, Prime superiority. Islamabad consistently rejected Minister Sharif announced that Pakistan was a policy of no-fi rst use. Some strategists in forced to conduct tests in order to reduce the Pakistan urged the government to deploy and military imbalance in South Asia. At the time disperse nuclear-tipped ballistic missiles at of his announcement, Pakistan already pos- high-levels of readiness. sessed warheads that were products of proven There is strong evidence that the govern- designs and tests. Islamabad conducted the ment is still grappling with its options for tests for political rather than strictly military doctrine and force structure and because of reasons. He also stated that the weapons could this it has taken a minimalist approach to be used to deter aggression during both conven- its doctrine for the time being. While reports tional and nuclear wars, suggesting a fi rst-use claim that the government formally established policy. The exact nature of the fi rst-use policy a nuclear doctrine in 2000 (withheld from remains unclear. But three distinct policy public scrutiny), many Pakistani offi cials have objectives can be inferred. First, Islamabad admitted that the principles for nuclear use hopes to deter a fi rst nuclear use by India. It are still evolving. In the most authoritative fi rmly believes that New Delhi, despite India’s statement yet, General Musharraf publicly repeated pronouncements of a second-strike noted that the warheads and delivery systems policy, would contemplate attacking fi rst in a remain decoupled. This announcement dove- crisis, perhaps to destroy Pakistan’s nuclear tailed with Pakistan’s policy of not deploying

38 Nuclear Stability in South Asia its nuclear forces in order to demonstrate and control system that would prevent acciden- restraint to the international community. As tal or unauthorized launches. Islamabad also a part of its public relations campaign, Islam- took steps to institutionalize its command and abad has also declared that it has achieved a control mechanism. In February 2000, the gov- suffi cient level of confi dence in the survivabil- ernment published a document that described ity of the nuclear forces against a fi rst strike. in very general terms the country’s nuclear Whether such confi dence is merely rhetorical command and control structure. In a brief state- or real, nuclear survivability from a fi rst strike ment, Pakistan’s National Security Council will be an ongoing challenge for Pakistani declared that it had approved the establishment defense planners. of the National Command Authority (NCA).30 Pakistan’s fi rst-strike policy stems from The NCA comprises the Employment Control three major interrelated vulnerabilities. First, Committee, Development Control Committee, the country lacks strategic depth geographical- and the Strategic Plans Division, which func- ly. India can easily threaten the few major cities tions as the secretariat and coordinator of without which Pakistan would cease to exist as command and control networks. The main a political entity. This stands in sharp contrast purpose of the organization is to formulate pol- to India’s landmass and far-fl ung and dispersed icy on the development and use of all strategic population centers. Second, Pakistan is highly nuclear forces and associated organizations. exposed to a deep conventional military thrust The members of the NCA include the head of that could geographically cut the nation in two. state or prime minister, ministers of foreign Such a blow would spell national disintegra- affairs and defense, chiefs of all the armed ser- tion. Third, there is a substantial asymmetry vices, and directors of strategic organizations. in conventional military power between the In early 2001, Pakistan established the Nuclear two rivals. As India’s economy promises to Regulatory Authority to supervise the secu- grow consistently while Pakistan’s stagnates, rity, safety, reliability, and survivability of this imbalance of power is likely to tilt further the country’s nuclear-related installations. On in favor of India. Pakistan therefore must dem- October 7, 2001, immediately prior to the com- onstrate the capability not only for deterrence mencement of the U.S. bombing campaign but also for at least a limited use in warf- against Afghanistan, the Pakistani govern- ighting. Pakistan could use nuclear weapons ment changed the key fi gures in the chain against the massed Indian ground forces prior of command for nuclear weapons and also to confl ict in order to degrade the threat to its altered their storage procedures for nuclear national survival. It could also signal the poten- devices.31 tial for nuclear use in order to internationalize Despite offi cial Pakistani reassurances that the crisis and draw sympathetic external play- its arsenal is both safe from theft and secure ers into the fray. to use or not use in times of crisis, it will take In contrast to New Delhi, Islamabad has years if not decades to put in place effective been relatively open about its command struc- mechanisms to manage the nuclear arsenal. ture. Government offi cials and military leaders The assurances originally were designed to have repeatedly claimed that Islamabad has win sanctions relief; now they are also designed devised an elaborate and effective command to demonstrate effective and strong leadership

Nuclear Force Structures and Doctrines 39 in a country with potentially severe internal tems, the country probably enjoys a relatively problems as the war in Afghanistan winds robust and proven nuclear infrastructure for down. Islamabad faces several fundamental the short to medium term. China is suspected challenges stemming from its nuclear doctrine of providing warhead designs, highly enriched that must be addressed as it considers how best uranium (HEU), and uranium enrichment to secure its arsenal. The presumed fi rst-strike technologies. Beijing may even have permit- posture that Pakistan has adopted could have ted the Pakistanis to carry out detonations of signifi cant implications for command and con- weapons designs without the physics package trol over its nuclear forces, especially in the at China’s Lop Nor testing ground as long ago absence of a sophisticated C2 architecture. The as 1983. After the 1998 tests, the Pakistani conviction that India would consider using its government declared that it had detonated weapons fi rst implies that the Pakistani lead- six enriched-uranium based boosted-fi ssion ership may need to disperse its forces early in devices. These have a more effi cient trigger any crisis, if not in peacetime, to avoid losing mechanism that produces a higher explosive its retaliatory option should India conduct a yield relative to the mass of the fi ssionable disarming fi rst strike. The short fl ight times material. Islamabad claimed the successful needed for both Indian fi ghter aircraft and tests of one twenty-fi ve kiloton, two twelve- ballistic missiles to reach their targets add to kiloton, and three sub-kiloton devices. As with the pressure to disperse. Moreover, Islamabad the Indian tests, independent seismic evidence faces a nuclear survivability problem as a result indicated that the detonation yields were much of its conventional military weakness vis-à- lower than Islamabad claimed. vis India. A massive Indian invasion could Pakistan’s nuclear infrastructure centers overrun facilities and launch sites containing on uranium enrichment technologies that Pakistan’s nuclear arsenal – essentially per- Islamabad acquired primarily from China. forming a disarming conventional fi rst strike. Pakistan also operates an unsafeguarded plu- Achieving the necessary connectivity between tonium separation facility. Islamabad may be the leadership and military commanders is working on other enrichment and reprocess- therefore a critical feature of Pakistan’s com- ing plants that could increase the capacity for mand and control, particularly in times of producing weapons-grade uranium and pluto- crisis when response times are often very nium.32 Estimates have varied on the quantity short. Additionally, there are many conceiv- of fi ssile material and warheads that Pakistan able circumstances (ranging from uncertainty now possesses or has the potential to produce. during crisis to miscommunication) that could Most calculations have converged on approxi- tempt a commander to launch a preemptive or mately six hundred-plus kilograms of HEU and unauthorized strike. over fi ve kilograms of plutonium for thirty-fi ve Pakistan initiated its nuclear program in to forty warheads.33 Substantial Chinese assis- the early 1970s in response to its defeat in the tance, as noted above, may have enabled 1971 war against India. The program accel- the Pakistanis to develop warheads small erated after India’s 1974 nuclear test. Given enough for arming ballistic missiles. More- Pakistan’s external sources of support for its over, the March 2000 MSNBC report suggests nuclear program and associated delivery sys- that Islamabad may have weaponized far

40 Nuclear Stability in South Asia more fi ssile material than Western intelligence short-range missiles are stored clandestinely expected. farther inland to avoid preemptive fi rst strikes. With the help of China and North Korea, Given their limited range, they would then be Pakistan’s nuclear ballistic missiles are based readied and deployed along the Indian border on both liquid- and solid-fueled technologies during crisis or war. (although substantial confusion persists over Pakistan likely had assembled nuclear the designations of the missiles). Pakistan devices before ballistic missiles became avail- apparently acquired from North Korea tech- able, and some of its aircraft may therefore nologies for the liquid-fueled Ghauri-I and -II have been modifi ed for nuclear missions. The with 1,500- and 2,000-kilometer ranges respec- A-5 Fantans, the French Mirage-III/5s, and tively.34 Pyongyang’s designations for the two F-16s in Pakistan’s current fl eet are all possible missiles are the No Dong-I and -II. Pakistan candidates for dropping nuclear weapons, but successfully tested the Ghauri-II in April their limited combat ranges, penetrability con- 1999. Islamabad may also be developing the cerns, and vulnerability to fi rst strikes have Ghauri-III based on the 3,000-kilometer range likely relegated these fi ghters to last resort mis- Taepodong-I ballistic missile. This missile sions. The availability of ballistic missiles has would enable Pakistan to hit any target in also lessened the relevance of aircraft deliv- India. The Chinese have also supplied Islam- ery. At present, sea-based nuclear capabilities abad with technologies and fully assembled belong in the theoretical realm for Pakistan. missiles for the solid-fueled Shaheen series. The The dominance of the army and air force will Shaheen-I, -II, and -III – with ranges of 300 ensure that developments in the naval leg of the kilometers, 800 kilometers, and 2,500 kilome- nuclear force structure will be slow over the ters, respectively – are based on the Chinese next ten to fi fteen years. M-11, M-9, and M-18. In April 1999, Islamabad Several preliminary conclusions can be tested the Shaheen-I. Pakistan indicated that drawn from this assessment of Pakistan’s cur- the Shaheen-II would satisfy the nation’s long- rent nuclear doctrine and force structure. First, range strike capabilities, presumably against because of Pakistan’s geographic, convention- India. Indeed, the missile would put most of al military, and economic vulnerabilities, the India’s major population centers (even in the doctrine of fi rst use will dominate Pakistani south) at risk. The solid-fuel technologies fea- nuclear thinking. Second, Islamabad will assid- tured on the Shaheen will enable the Pakistanis uously build up a comprehensive arsenal of to respond more rapidly and fl exibly. Both the ballistic missiles with ranges allowing it to Ghauri and the Shaheen are mobile, suggesting strike deep into the Indian heartland. Third, that concealment and dispersement are central Pakistan’s military and industrial disadvan- to Pakistan’s nuclear strategy. The indigenous- tages will constrain its ability to expand the ly produced Hatf-I and -II have been deployed arsenal independently, and it will look to China only on a limited basis because of their modest and North Korea for ongoing technological ranges (100 and 280 kilometers, respectively) assistance. and lack of sophisticated guidance. The total inventory and the operational status of the Hatf force are unknown.35 It is likely that the

Nuclear Force Structures and Doctrines 41 Future Nuclear Doctrines and In ten to fi fteen years the interaction Force Structures in South Asia between doctrine and force structure within Some analysts have speculated that “recessed each country could create an irresistible inter- deterrence” – an opaque nuclear posture nal logic of its own. Technological advances that both sides assumed before the tests – could change the force structure in ways would continue to suffi ce for India and Paki- that would require doctrinal adjustments to stan. These observers contend that the tests manage and harness new capabilities. Techno- were primarily politically motivated and had logical constraints could also shape the future few strategic implications. Jasjit Singh defi nes command and control architectures on both recessed deterrence as a strategy that “would sides. Safety mechanisms to prevent acciden- require a non-weaponised status, but where tal or unauthorized launches are still in early all necessary steps for ‘weaponisation’ and its stages of development in India and Pakistan, usability have been taken.”36 A recessed deter- largely because arsenals are only now begin- rent would permit either side to assemble a ning to be weaponized and possibly deployed. nuclear arsenal within a few months. The abil- More centralized command structures and ity to deploy overtly would provide each with restrictive procedures on nuclear use are there- suffi cient leverage to safeguard its national fore likely to dominate for some time to come. security interests.37 Whether India and Pak- Alternatively, perceived satisfaction with doc- istan have abandoned or are in the process trine and strategic requirements could freeze of abandoning their recessed deterrence pos- further developments in the nuclear arsenals. tures remains unclear. But recent evidence For example, the Indian government appears indicates that India and Pakistan have weap- content with an assured second-strike capabil- onized (by some estimates in larger quantities ity, which suggests a relatively minimalist level than expected) and have deployed or are in the of suffi ciency to deter Pakistan and China. process of deploying associated delivery sys- Beyond the interactive process within each tems. As mentioned above, “deployment” or country, there is also an action-reaction rela- “induction” – as Indian authorities prefer – can tionship between the two states. Both sides manifest itself in various confi gurations that are fundamentally at odds with each other do not fi t neatly with familiar Western con- over clashing national interests. Since each ceptions. For example, nuclear warheads and side views the strategic competition as a zero- delivery systems that are fi elded but de-mated sum game, any perceived gain on one side and geographically separated by signifi cant would likely spur the other to respond in distances could constitute “full deployment” kind or devise a counter. As mentioned above, in the minds of Indian and Pakistani strate- each side suffers from grievous mispercep- gists. How far New Delhi and Islamabad are tions (both in strengths and weaknesses) of prepared to carry on with these nuclear devel- the other, which could exacerbate acute anxi- opments remains unanswerable. But it is clear eties on each side. Lessons learned from past that the evolving force structures and doctrines and future crises could further feed such mis- outlined above suggest that the nuclear pos- understanding, thereby forcing the leaderships tures of both states will continue to grow in to reassess and formulate radically different size and sophistication over the next decade. nuclear strategies. Nuclear stability or volatil-

42 Nuclear Stability in South Asia ity will depend on the potent and potentially established stance on no fi rst use. Second, the combustible mix of these complex internal and doctrine would be focused on guaranteed rath- external factors. er than rapid reaction to a nuclear fi rst strike. This section outlines three general nucle- Third, New Delhi’s targeting would be inher- ar doctrines and force structures that each ently countervalue in nature. In peacetime, side could assume in the fi rst ten to fi fteen the components of the nuclear weapons would years of the twenty-fi rst century. Again, the be dispersed geographically and placed under fi fteen-year timeline provides a useful plan- separate authorities. The civilian leadership ning parameter and allows suffi cient time for would maintain control over the unassembled the nuclear futures of each side to mature in nuclear warheads. The military would over- meaningfully measurable ways. For purposes see the control of the delivery systems, which of simplifi cation (and to avoid the common pit- would be stored away from the launch sites and falls of defi nitional confusion), the analysis the warheads. The delivery systems would be will focus on three alternative postures (low, maintained at varying levels of readiness. In medium, and high ranges) that India and Paki- times of “supreme crisis,” the civilian and mili- stan could pursue. The three alternatives are tary leaders would set in motion the process not defi nitive. Indeed, it is entirely conceivable that would integrate the various parts of the that certain components of all three will coex- nuclear weapon for deployment and employ- ist. Moreover, incremental weaponization over ment. For this leisurely posture, an inventory the next decade would allow certain elements of 150 weapons with ten- to twenty- kiloton of the force posture to develop ahead of others. yields by 2010 would suffi ce for deterring Paki- Each of these force structures will also be sub- stan and China. The delivery systems would ject to the strategic drivers (such as deep-seated center primarily on tactical strike aircraft and, threat perceptions) mentioned at the outset of assuming that India has successfully designed this chapter. a miniaturized warhead, a handful of Agni-II missiles. India’s Low Nuclear Posture38 A linear projection of trends since the 1998 India’s Medium Nuclear Posture tests suggests that India could pursue a rela- This posture resembles the fi rst model with tively conservative nuclear strategy, essentially a few modifi cations. Deteriorating strategic maintaining its current doctrine and force trends and domestic considerations could structure. Open source materials indicate that undermine the constraints imposed by nor- New Delhi has not radically expanded the mative considerations and rigid civil-military number of warheads or overtly deployed bal- relations, thereby pushing India to pursue a listic missiles. India’s inherent advantages over numerically larger and possibly technologically Pakistan and its satisfaction with the current more sophisticated posture. New Delhi’s per- level of suffi ciency for deterring China could ceptions of the Chinese threat could worsen lead New Delhi to adopt a restrained posture. over the next decade and a half. Recent evi- The nuclear doctrine that underwrites this dence already indicates that Chinese nuclear low projection is based on a credible threat of posture will not remain static. Solid-fuel, road- punishment. First, India would abide by its mobile strategic missiles, which will replace

Nuclear Force Structures and Doctrines 43 obsolescent liquid-fuel, fi xed ICBMs, will be which would form the backbone of its nucle- slowly incorporated into the Chinese inven- ar deterrent. The induction or deployment of tory. Some analysts have even argued that these missiles could take on several confi gu- Beijing’s nuclear doctrine has shifted grad- rations. The warheads and delivery systems ually (a process underway for more than a could be de-mated but placed in closer proximi- decade) from minimum deterrence to limited ty or located in the same facility. The warheads deterrence, which connotes the capacity to use and delivery systems could be assembled but nuclear weapons for calibrated escalation and stored in a secured location away from the warfi ghting. launch site. Both of these options enhance How would Beijing’s ongoing nuclear mod- responsiveness while retaining a high level ernization affect India’s security? What kind of procedural control over the employment of of Indian response would such developments nuclear weapons. trigger? First, since nuclear weapons are close- However, several additional considerations ly linked to prestige and national pride, India would have to be taken into account for this might feel compelled to keep pace with the posture. First, in order to heighten its deter- Chinese. Second, a larger and more capable rence credibility against China, India would nuclear arsenal based on limited deterrence have to expand its target set to include far-fl ung could enable Beijing to impose a limited form of strategically important cities such as Beijing escalatory dominance on India. The prospect and Shanghai. Whether the longer-range Agni- of such Chinese pressure or intervention on III will be ready for deployment in ten years behalf of Pakistan in a future crisis could prove remains an open question. Second, India may politically unpalatable. Third, China may have to increase the destructive power of its build a missile defense system that could warheads, such as boosted fi ssion and possibly fundamentally erode the deterrent value of thermonuclear devices, to bolster its deter- India’s small nuclear force. Fears about China’s rent. Given the questionable track record of nuclear superiority would undoubtedly create India’s warhead development, it is unclear tremendous domestic pressures and political whether New Delhi would be able to construct justifi cations for a more powerful arsenal in high-yield warheads and, more importantly, India. miniaturize them to sit atop ballistic missiles. Similar to the fi rst model, this model proj- On a related point, India may have to conduct ects that India would still build a nuclear more nuclear tests for quality control and to force structure designed for an assured second- validate miniaturized designs. Third, a larger strike capability. The civilian leadership would and dispersed force would require the civilians continue to maintain strict control over the to improve command and control to tighten decision-making process and the employment their control over the arsenal. Within the 2015 of the nuclear assets. In contrast to the fi rst timeframe, it will be challenging for India to model, however, India would shift away from acquire the more technologically sophisticated its tactical strike aircraft for the delivery of C2 system that would have to accompany such nuclear weapons. New Delhi would induct a process. New Delhi would have to address larger numbers of survivable mobile Agni-II a range of issues, including security (theft intermediate range ballistic missiles (IRBMs), prevention), safety (accident prevention), reli-

44 Nuclear Stability in South Asia ability, and survivability at a higher and more has been long overdue. They believe that India sophisticated level with such a force posture. needs to energize the process further to catch up with the rest of the world.39 In other words, India’s High Nuclear Posture whether the security environment worsens or India may take enough steps in ten to fi fteen not, India should nonetheless pursue a nuclear years to ensure that it could acquire a rad- posture commensurate with its international ically larger and more sophisticated nuclear status and its security interests. They are con- force. Such a ready arsenal would most resem- vinced that a credible nuclear capability at ble Western conceptions of full deployment. more sophisticated levels would endow India The warheads and delivery systems would be with the necessary prestige and power on the mated for employment. These assembled weap- world stage. ons could be stored near or at the launch site. Externally, increased hostilities (actual or The missiles could be at a lower readiness level rhetorical) with China and Pakistan or Chi- whereby targeting, fueling, and other prepara- na’s continued proliferation to Pakistan could tions could only commence with authorization all exacerbate New Delhi’s threat percep- from the civilian leadership. Assuming that tions. Indeed, it could strengthen the hands India could design and/or acquire a robust of the hawks noted above. China’s nuclear command and control and early warning sys- modernization program could accelerate faster tem it might even place the nuclear force on than anticipated. Should Pakistan accumulate higher levels of readiness including launch enough weaponry and associated intelligence on order and, most destabilizing, launch on and targeting assets that it could threaten a warning. The military would enjoy a greater disarming fi rst strike, India would have to autonomy or pre-delegated authority to launch. develop an even more robust second strike that While this model is the least likely politically could survive preemption. These events togeth- and technologically, it is still worth examining er could also force a gradual transformation in the drivers that would compel India to move India’s strategic culture, international outlook, toward a far more ambitious nuclear posture in and behavior. India may feel compelled to shed the next ten to fi fteen years. its normative constraints and move toward a Exogenous shocks or internal assessments realpolitik orientation. This may unleash its of the nation’s defense requirements could desire to play a greater role on the global stage. infl uence Indian thinking on nuclear issues Provided that sustainable economic growth independently of each other. As the NSAB continues, the political will to carry the fi nan- report clearly demonstrates, there are many cial burden would enable India to pursue a individuals and more hawkish schools of larger and more diverse nuclear force. thought in and out of government that believe These domestic or external forces would India should pursue a robust nuclear posture. drive India to develop and begin to deploy Infl uential fi gures, such as the current foreign a land-air-sea triad. While estimates of the and defense minister, Jaswant Singh, assert required numbers of warheads vary widely, that India needs to abandon its normative con- one study calls for an end state of an arsenal straints and pursue a realist orientation. Some with three hundred-plus warheads. Jaswant analysts go further and argue that this shift Singh asserts that a credible deterrent for India

Nuclear Force Structures and Doctrines 45 would require the ability to hit as many as management, and command and control to sixty primary and secondary targets.40 India help India assess the growing threats from would have to devote substantial resources to two fronts. India already enjoys a relatively developing the sea leg of the triad to ensure advanced space-launch industry from which maximum survivability and to safeguard its to exploit commercial space technologies. As assured second-strike capability. This sea- a larger nuclear state in this projection, India based element could include both ballistic and might seek a missile defense capability to cruise missiles launched from a surface ship or counter accidental or unauthorized launches a submarine. Again, while analysts differ on from either Pakistan or China or to render how many submarines would be needed, most Islamabad’s deterrent obsolete. Escalatory dom- concur that at least one to two submarines, inance could then become an even more each carrying a minimum of twelve ballistic attractive option for New Delhi. Many of these missiles, would have to be operational at any new elements would be in early stages of devel- given time. Taking into account maintenance opment. The technological constraints noted in and refi tting cycles, a fl eet of three to four sub- the second model would persist, and it seems marines could fulfi ll this requirement. Given likely that India would be struggling to master the technical diffi culties associated with a sea- these sophisticated technologies and the doc- launched capability, particularly in the SLBM trines and management systems necessary to confi guration, India would likely still be strug- use them optimally. gling to achieve this capacity by 2010 but may reach it by 2015. Pakistan’s Low Nuclear Posture India would seek to deploy ICBMs capable Pakistan is probably edging toward such a pos- of reaching as far as Beijing and all other major ture at present. Several drivers could motivate cities in eastern China. Bharat Karnad esti- Islamabad to pursue this restrained approach. mates that India would have at least fi ve ICBMs First, the external environment and relations by 2010 and as many as fi fteen or more by with India could stabilize suffi ciently for 2020. Given the advanced stage of its IRBMs, Pakistan to feel secure enough to freeze fur- their inventory could grow to ten by 2010 and ther nuclear developments. Second, external up to twenty-fi ve by 2020.41 With such growth, resources from China and North Korea could India would surely have to conduct missile dry up unexpectedly. For example, the North tests to validate accuracy, range, and reliabil- Korean regime could conceivably collapse in ity in the intervening years. As mentioned the next decade. Alternatively, nonprolifer- for the medium-level model, India may fi nd it ation pressures from the United States or necessary to increase the yield of its nuclear improved relations with Washington could warheads in order to strengthen its deterrent decrease the willingness among external sup- force. There would then be more political pres- pliers to provide further support to Pakistan sure to develop thermonuclear weapons, which and infl uence decisions in Islamabad itself would heighten the likelihood of additional about how far to go in developing its arsenals. nuclear tests. Some additional components This minimum posture would entail the to the posture would be required, including deployment of a limited number of weapons. space-based assets for early warning, battle There would be a mix of aircraft and ballistic

46 Nuclear Stability in South Asia missiles as delivery vehicles. The latter, a com- ture. As noted earlier, intensifi cation of the bination of liquid-fuel and solid-fuel rockets, Sino-Indian rivalry may automatically raise would over time begin to dominate the force Pakistan’s perceived level of suffi ciency for structure. Most of the weapons would be deterrence. So long as Chinese and North Kore- targeted primarily against India’s major cit- an technology transfers continue, Pakistan ies. The rest would be kept in reserve for would be able to sustain a larger force and keep demonstration purposes or actual use on a con- up with new technologies. ventional battlefi eld. Some weapons may be This medium posture would consist of a fully deployed (mated and otherwise prepared larger and more sophisticated nuclear force. for launch) while others remain at low levels of The solid-fueled, road-mobile ballistic missiles readiness. The nuclear weapons would likely would begin to eclipse the liquid-fuel rockets. be mounted on road-mobile launchers. Other Many would be based on the Shaheen-II and weaponized warheads would be de-mated from perhaps the Ghauri-III. This would enable Pak- their delivery systems. Both the warheads and istan to deploy its forces farther from India, delivery systems might be in different storage making them less vulnerable to preemption. facilities or be located at the same facilities. Aircraft delivery would be largely relegated Both the ready weapons and the separated to last-resort missions. Some of the weapons components of the arsenal would be widely dis- would be targeted at a larger set of popula- persed in hidden storage sites. The command tion centers, including cities deep in southern and control structure governing this force India. The remaining force would be reserved would still be relatively rudimentary. Com- for demonstrations and limited warfi ghting. manders and operators on the ground would Warfi ghting weapons would be focused pri- have to await orders from the highest levels marily against an invading conventional force of authority to launch. During crisis, the sep- within Pakistani territory or close to the Indo- arated weapons would be reconstituted and Pakistan border. Most of the missiles would deployed relatively quickly while the fully have warheads and be fully deployed at rela- deployed missiles would be able to engage tively low levels of readiness. A small number in evasive maneuvers to avoid preemptive of weapons might still be de-mated and hidden strikes. in separate storage facilities. The command and control system would be more sophisti- Pakistan’s Medium Nuclear Posture cated, with some ground-based radar for early This more robust posture does not deviate warning and a more sophisticated secure com- far from the minimum stance outlined above. munications network linking the leadership to In this case, the Pakistani leadership would the fi eld commanders. believe that a larger arsenal is necessary to maintain a credible fi rst-strike capability. The Pakistan’s High Nuclear Posture security environment would still be tense and The security environment would have to dete- punctuated by regular standoffs with India riorate drastically to drive Islamabad to pursue over Kashmir and other issues. India would such a course, resulting in a full-blown arms also be in the process of building up a nuclear race on the subcontinent. China’s continued capability that is moving beyond the low pos- nuclear modernization could enable Beijing to

Nuclear Force Structures and Doctrines 47 fi eld a force based on credible limited deter- grams would still be in the early stages by rence, forcing India to accelerate its nuclear the 2010-15 timeframe and would likely take program in response and setting off a chain another decade to become fully operational. reaction in South Asia. Major reconfi gurations in regional relations could also radicalize Pak- Implications for Strategic Stability istan’s threat perceptions. The United States A key condition that would ensure strategic and India could form a new strategic part- stability in ten to fi fteen years is mutual vul- nership. This could force Pakistan to foster nerability. As long as both sides can credibly deeper political and military cooperation with infl ict unacceptable damage on each other, China and possibly Russia as a strategic mutual deterrence should work if neither counterweight. India could succeed in fi eld- side falls prey to mistaken perceptions or ing a sea-launched nuclear deterrent. Unlike faulty decision making (as noted in chapter 1). India’s ground-based and air-launched nuclear For India, such a capability must include an force, a submarine fl eet would be able to oper- assured countervalue second-strike capability ate virtually undetected, thereby creating an to deter and, failing that, to respond to a Paki- impregnable deterrent. Given Pakistan’s deep stani fi rst strike. Over the next decade, New distrust of Indian intentions, a submarine lurk- Delhi’s level of suffi ciency to retaliate against ing in the open seas could fuel fears (however a fi rst strike would still center on the ability to unwarranted) of a bolt-from-the-blue attack. target three or four major population centers in In response to these trends, Pakistan could Pakistan. As mentioned above, Pakistan would be compelled to develop a much more com- cease to function as a state with the destruction prehensive and highly decentralized nuclear of those cities. As long as India can ensure that arsenal for a wider range of targets and its forces targeted at Pakistan are survivable, missions. A larger number of solid-fuel, road- New Delhi may not feel compelled to build a mobile, long-range ballistic missiles would force beyond the low to medium postures or to dominate the force structure. Tactical fi ghter move toward a fi rst-strike policy. Even if India delivery would play a minor role if any at all. pursues one of the more advanced postures A larger counterforce arsenal would accom- outlined above, strategic stability would not pany the capacity to decimate major cities necessarily suffer. In fact, a more sophisticated throughout India. An inventory of shorter- force, particularly with an SLBM component, range missiles would be built up for use against would bolster India’s confi dence in its ability India’s conventional forces, while the lon- to survive a fi rst strike and punish Pakistan, ger-range weapons would be targeted against thereby reinforcing its second-strike policy. India’s nuclear command and control centers While an increase in the size of the nuclear and launch facilities. Islamabad would also force could help India expand the list of Paki- begin to engage in the research and develop- stani targets, it is more likely that India would ment of a naval nuclear option. Pakistan would divert those weapons for use against the more harness a more sophisticated command and formidable threat from China. control system that integrates its conventional Unlike India, Islamabad cannot strategical- forces with the nuclear posture for a limited ly afford a posture based entirely on retaliation. fl exible-response capability. Many of the pro- Pakistan’s fi rst-strike policy encompasses a

48 Nuclear Stability in South Asia broad range of objectives: 1) to deter Indian invasion while Islamabad would be too weak fi rst use (a widely accepted assumption among (and even weaker in the future) to challenge Pakistani strategists); 2) to defend the home- New Delhi on the battlefi eld. However, it is land against an Indian conventional invasion; unclear whether nuclear weapons would deter 3) to demonstrate its resolve; and 4) to inter- confl icts below the conventional threshold. nationalize any situation that could prove Indeed, as will be seen in the next chapter, unfavorable. Given these more demanding both the 1990 war and the 1999 Kargil con- requirements, Pakistan’s level of suffi ciency fl ict suggest that Pakistan might be emboldened has always been higher than that of India. It to engage in low-intensity confl ict under the would in fact increase as India’s nuclear and protective cover of its nuclear deterrent. If conventional capabilities improve in the ten- to Islamabad believes that it can resort to simi- fi fteen-year timeframe. In terms of deterrence, lar strategies in the future and if New Delhi Islamabad would have to hold hostage larger is deterred from escalating a low-intensity numbers of Indian cities to ensure deterrence confl ict, then nuclear weapons could in fact credibility. In terms of warfi ghting, the num- prolong the ongoing struggle over Kashmir. ber of weapons required would vary depending Such an outcome, as will be shown in chapter on the intensity and severity of the crisis (such 4, also presents the possibility for escalation as how many conventional thrusts it must con- through a combination of misperceptions and tend with). This suggests that Pakistan will domestic political pressures. pursue postures with higher numbers of weap- The gravest danger to strategic stability in ons than India over the next ten to fi fteen years ten to fi fteen years would be the emergence to ensure mutual vulnerability. Given these of an asymmetry (real or perceived) between factors, a more robust Pakistani nuclear pos- Indian and Pakistani nuclear postures. Dras- ture may be the most conducive to strategic tic shifts in the force posture on either side stability. For example, a larger, fl exible, and could heighten threat perceptions and trigger responsive nuclear force would alleviate Islam- a vicious action-reaction cycle. A qualitatively abad’s persistent fear of preemption by an or quantitatively higher posture in the nucle- Indian nuclear or conventional fi rst strike, ar arsenal could spark counter moves that thereby minimizing or even avoiding the “use could prove highly destabilizing. In the case it or lose it” concern. of Pakistan, it could receive large numbers Assuming that both sides maintain sat- of fully assembled warheads or delivery sys- isfactory levels of suffi ciency within the tems from China or North Korea during the fi fteen-year timeframe, nuclear weapons in 2015 timeframe that signifi cantly increase its South Asia would likely deter nuclear and fi rst-strike force. A more robust Pakistani fi rst- conventional wars. Pakistan would not con- strike force that New Delhi perceives to be a template a nuclear fi rst strike because this precursor to a disarming capability might lead would invite national annihilation. India’s sec- India to reconsider its second-strike policy or to ond-strike policy would ensure that it would acquire additional capabilities in order to pre- never attack Pakistan unless it was struck clude a successful Pakistani fi rst strike. India fi rst. Pakistan’s fi rst-strike policy would pre- could introduce indigenously produced or for- vent India from contemplating a conventional eign built sea-based platforms into its nuclear

Nuclear Force Structures and Doctrines 49 arsenal, which would virtually guarantee its assured retaliatory posture. The Pakistanis might then contemplate preemptive attacks before India fi elds such capabilities. This interactive dynamic suggests that sudden or unexpected transitions to higher postures on either side would likely be strategically desta- bilizing. Even if such posture changes do not occur suddenly, the perception that they have or are about to could be just as destabilizing. Such a situation could not be resolved with changes in hardware and would have to be addressed at a more fundamental level, as dis- cussed in chapter 5. Whether the future force structures actu- ally turn out to be stabilizing or destabilizing depends largely on the future political and mili- tary context in which they exist and that, in turn, depends in part on the history of the region and the lessons that each side has taken from that history and will apply in future cri- ses and wars.

50 Nuclear Stability in South Asia Endnotes for Chapter Two

1 See Samina Ahmed, “Pakistan’s Nuclear Weapons Program: Turning Points and Choices,” International Security 23, no. 4 (spring 1999): 178-204. 2 Neil Joeck, Nuclear Developments in India and Pakistan (Seattle: National Bureau of Asian Research, 1999), 12-13. 3 Jaswant Singh, Defending India (New Delhi: Macmillan Press, 1999), 268. 4 Ibid., 326. 5 See George Perkovich, India’s Nuclear Bomb: The Impact on Global Proliferation (Berkeley: University of California Press, 1999). 6 Joeck, 22-23. 7 Ibid., 23. 8 Ashley Tellis, “The Changing Political-Military Environment: South Asia,” in Zalmay Khalilzad, David Orletsky, Jonathan Pollack, Kevin Pollpeter, Angel Rabasa, David Shlapak, Abram Shulsky, and Ashley Tellis. Eds., The United States and Asia, Toward a New U.S. Strategy and Force Posture (Washington, D.C.: RAND, 2001), 234. 9 Author’s interview, September 29, 2000. 10 Scott D. Sagan, “The Perils of Proliferation in South Asia,” Asian Survey, forthcoming. 11 See the Economist Intelligence Unit, Country Report: Pakistan (London: EIU, May 2001). 12 Stephen Cohen, “Policy Brief: Moving Forward in South Asia” (Washington, D.C.: Brookings Institution, May 2001), 3. 13 Jaswant Singh, Defending India, 1-60. 14 Deepa M. Ollapally, “India’s Strategic Doctrine and Practice: The Impact of Nuclear Testing,” in Raju G.C. Thomas and Amita Gupta, eds., India’s Nuclear Security (London: Lynne Rienner Publishers, 2000), 69-71. 15 Raja Menon, A Nuclear Stratety for India (New Delhi: Sage Publications, 2000), 25. 16 For details, see “Draft Repot of the National Security Advisory Board on Indian Nuclear Doctrine,” at http://www.meadev.nic.in/govt/indnucld.htm. 17 Authors’ interview with India’s national security advisor, Brajesh Mishra, September 26, 2000. 18 “British Report Disputes Indian Claim on Yield of Nuclear Test,” Indian Express, July 23, 2001, www.nci.org/01/07/22-ind-indian_yield_test.htm. 19 Hilary Synnot, Causes and Consequences of South Asia’s Nuclear Tests (London: International Institute for Strategic Studies, December 1999), 55. 20 Perkovich, 22. 21 Raju G.C. Thomas, “India’s Nuclear and Missile Programs: Strategy, Intentions, Capabilities,” in Thomas and Gupta, 109. 22 Rodney W. Jones and Mark G. McDonough with Toby Dalton and Gregory Koblentz, Tracking Nuclear Proliferation (Washington, D.C.: Carnegie Endowment for International Peace, 1998), 112. 23 See Gurmeet Kanwal, India’s Nuclear Force Structure (New Delhi: Institute for Defence Studies and Analysis, 2000). David Albright, India and Pakistan’s Fissile Material and Nuclear Weapons Inventory, End of 1999, (Washington, D.C.: Institute for Science and International Security, 2000). Federation of American Scientists, Nuclear Weapons: India Nuclear Forces, http://www.fas.org/nuke/guide/india/nuke/index.html. 24 Raj Chengappa, Weapons of Peace: The Secret Story of India’s Quest to be a Nuclear Power (New Delhi: Harper Collins, 2000), 382-384. 25 Duncan Lennox, ed., Jane’s Strategic Weapon Systems (Alexandria, Virginia: Jane’s Information Group, 2001), 80-81. 26 Andrew Koch, “Nuclear Friction,” Jane’s Defense Weekly, December 13, 2001, 22.

Nuclear Force Structures and Doctrines 51 27 Lennox, 83-84. 28 Carnegie Endowment for International Peace, “India’s Slow-Motion Nuclear Deployment,” Nonproliferation Project 3, no. 26 (September 7, 2000). 29 Ibid. 30 Pakistan Ministry of Defense, press release: “NSC Approves Establishment of National Command Authority,” February 2, 2000. 31 Molly Moore and Kamran Khan, “Pakistan Moves Nuclear Weapons; Musharraf Says Arsenal Is Now Secure,” Washington Post, November 11, 2001, p. A1. 32 Jones and McDonough, 134-136. 33 Albright, 3. 34 Lennox, 126-127. 35 Ibid., 122-123. 36 Jasjit Singh, “A Nuclear Strategy for India,” in Jasjit Singh, ed., Nuclear India (New Delhi: Institute for Defense Studies and Analyses, 1998), 320. 37 Ashley Tellis, India’s Emerging Nuclear Doctrine: Exemplifying the Lessons of the Nuclear Revolution (Seattle: NBR Analysis, May 2001), 8. 38 For a comprehensive assessment of this posture, see Ashley J. Tellis, India’s Emerging Nuclear Posture: Between Recessed Deterrent and Ready Arsenal (Santa Monica, Calif: RAND, 2001). 39 See Menon, 23-38; Jaswant Singh, Defending India, 1-60; and Amitbah Mattoo, “India’s Nuclear Policy in an Anarchic World,” in Amitbah Mattoo, ed., India’s Nuclear Deterrent; Pokhran II and Beyond (New Delhi: Har Anand Publications, 1999), 9-33. 40 Bharat Karnad, “A Thermonuclear Deterrent,” in Thomas and Gupta, 143. 41 Ibid., 144.

52 Nuclear Stability in South Asia CHAPTER THREE

Stability Issues from Past Wars and Crises

In thinking about future crisis stability between India and Pakistan, it is necessary to move beyond the deterrence concepts that evolved over two generations of standoff between the United States and the Soviet Union. Most South Asian analysts suggest that such concepts have limited utility on the subcontinent and that Western observers simply ignore the history and complexities of the region. In a study of the role of deterrence in the 1965 war Sumit Ganguly declares “…that notions of deterrence have to be grounded in specifi c historical circumstances because of the crucial role that domestic politics play in both shaping and unraveling deterrence strategies.”1 In chapter 2, the discussion began to move beyond the U.S.-Soviet Cold War experience by sketching out, as clearly as possible, the evolutionary paths that India’s and Pakistan’s nuclear strategies, doctrines, and force sizes and structures may take. Several important factors were presented, including domestic politics and history that help shape these possible futures.

Nuclear Stability in South Asia 53 Both New Delhi and Islamabad have obvi- of them anew. Rather it draws on key studies ously learned from observing the U.S.-Soviet and histories of each, supplementing them with experience, even though they are starting from insights from interviews with U.S. policy mak- different technological levels for both weap- ers and analysts as well as with current and ons and delivery systems. The accompanying past Indian and Pakistani policy makers and technologies available for command, control, analysts. None of the accounts are perfect, communications, computers, intelligence, sur- even those that have been subjected to lengthy veillance, and reconnaissance (C4ISR) are also and rigorous historical analysis. Furthermore, at signifi cantly different levels than at the India and Pakistan each viewed the crises and beginning of the U.S.-Soviet nuclear competi- their outcomes very differently and for the tion. In South Asia the technical level of the most part continue to do so. Moreover, no two weapons and delivery systems far outstrips crises are the same because of differing issues those of the available intelligence, reconnais- as well as personalities involved and lessons sance, and C4 capabilities, prompting the learned – correctly or not – from every pre- potential for different types of instability. vious experience. However, the point of this In this chapter, we seek to build yet more study is not to predict perfectly the next cri- region-specifi c information that will help in sis and how it will unfold. Rather it is to analyzing future crisis stability. We will do provide some parameters for speculating about this by examining the wars, crises, and near possible future crises so as to guide analysis crises that have characterized Indo-Pakistani regarding potential stabilizing measures that relations since 1947. The goal of this examina- would reduce the possibility of nuclear escala- tion is to identify a number of critical variables tion and enhance the prospect that such crises that could infl uence future crises, how they might even be prevented. might unfold, how they may escalate, and how Before discussing the various wars and cri- they can be resolved. In looking at previous ses that have occurred between Pakistan and wars, crises, and near crises, we will seek to India, several points should be noted. First, it is identify their cause or causes; the patterns of important to remember that despite the heat- escalation in each; the critical turning points; ed rhetoric and the breathless news coverage the methods of de-escalation attempted; the les- of Kashmir, the disputed territory is not the sons learned by each side; and fi nally, the role only reason that India and Pakistan have gone of nuclear weapons, if any. This chapter iden- to war or have had disputes. The causes of tifi es key trends and patterns in these crises wars and crises have varied over time although and wars. These trends and patterns will then Kashmir has been a recurring theme in the be used in chapter 4 as a basis for developing antagonisms between the two states. There future crisis scenarios. Combined with the have also been long periods of relative peace analysis set forth in chapter 2, chapter 4 will between New Delhi and Islamabad. Second, then provide a sense of future crisis dynamics, not all crises involving these two states have escalation potential, and instabilities. involved antagonisms between them. Indians Given the large body of work that exists are quick to point out, quite rightly, that they on the various India-Pakistan wars and crises, had a war with China, and indeed that war was this study does not attempt to reassess each important in pushing decisions by India and

54 Nuclear Stability in South Asia Pakistan with regard both to their conven- crises, and near crises that had implications for tional military capabilities and, at least in nuclear weapons development or had nuclear- India’s case, the public debate about the pace related threats associated with them. and purpose of their nuclear program.2 Future crises involving China could include nuclear Kashmir Divided: 1947-1948 escalation as well, but for this study, we are The antagonism between India and Pakistan concentrating on the India-Pakistani relation- is as old as their independence.3 Before the ship, as the probabilities for future confl ict British government granted independence in along that axis at the moment seem signifi cant- 1947, there was major strife among Hindus, ly greater. In addition, the crisis in South Asia Muslims, and Sikhs, which resulted in vast precipitated by the September 11, 2001, terror- population shifts in anticipation of the end ist attacks on the United States has obvious of British rule. Only two months after inde- nuclear stability implications, as noted at the pendence, India and Pakistan were at war outset of chapter 1. While the full implications over Jammu and Kashmir, a disputed territo- of this crisis will take months if not years to ry that, before independence, was a princely unfold and assess, it shows clearly that crises state. Instead of following the partition formula in the region not necessarily directly involving whereby Muslim majority areas “located in ter- confl ict between India and Pakistan can affect ritories contiguous to Pakistan would accede to nuclear stability in numerous ways. Pakistan and the rest would go to India,” indi- Nuclear weapons have already been vidual rulers of princely states were allowed to involved in some way or another with almost accede to either India or Pakistan, with inde- all of the major wars or crises between Paki- pendence not being considered a real option.4 stan and India. The involvement has varied Communal violence escalated in Kashmir, and, as will be shown below, has included deci- the ruler vacillated, and rebels from Paki- sions to build nuclear weapons, threats to build stan’s Northwest Frontier Province crossed them, to use them, and to attack nuclear pro- into Kashmir rallying behind Kashmiris from duction facilities. The fi rst three wars between what Pakistan now calls Azad Kashmir. The India and Pakistan – 1947-48, 1965, and 1971 Pakistani government supported the rebels, – occurred before either state had developed who advanced on Kashmir’s capital of Srina- nuclear weapons capabilities, but the latter two gar, causing the monarch to fl ee. The next each had nuclear-related consequences, partic- day, Kashmir’s ruler signed an accession agree- ularly the 1971 confl ict. Additionally, all of the ment with India. India airlifted troops into the disputes between the two states have their ori- region and pushed the rebels back. gins in the confl ict that accompanied the two The UN brokered a cease-fi re that went states’ independence. The crises and wars after into effect on January 1, 1949, leaving the fi nal 1971 took place when at least one side was able political status of the region unresolved and to demonstrate nuclear weapons capabilities. the situation on the ground divided along a Given the focus of this monograph on ques- cease-fi re line. India was left in control of the tions of nuclear stability, the fi rst major war majority of what had been Jammu and Kash- between Pakistan and India will be described mir. Kashmir’s political limbo has given rise only briefl y. The focus will be on those wars, to numerous disputes between the two states,

Stability Issues from Past Wars and Crises 55 internal political crises, and attempts by each as a future threat and saw a window closing side unilaterally to determine defi nitively the for challenging India militarily over Kashmir. area’s status. Neither side has been willing to Nehru’s death in 1964 and the subsequent compromise on the territory for two reasons. struggles for leadership between Shastri and First, the land is strategically important for Indira Gandhi also led Islamabad to believe both, providing strategic depth for Pakistan that India was entering a period of political and constituting critical military high ground weakness and division.7 Pakistan worried for India, somewhat like the Golan Heights for that India’s legal maneuverings to remove Israel. More importantly, the territory is sym- Kashmir’s special legal status and the joining bolic for each country in terms of its legitimacy. together of the state’s National Congress party Pakistan was founded on the theory that Mus- with the Indian Congress party constituted lims on the subcontinent could only have their an attempt by New Delhi to settle Kashmir’s rights ensured if they had their own country. If status unilaterally. Finally, and according to a Muslim-majority area (Kashmir) could thrive one prominent scholar, more speculatively, the in Hindu-dominated India, then Pakistan’s Pakistani leadership was worried about India’s raison d’être could be called into question.5 progress in building nuclear weapons. Based Conversely, India prides itself on being a sec- on statements by Indian weapons scientists, ular democracy that protects the rights of all Pakistani president Ayub Khan and Foreign minorities. If it cannot successfully incorpo- Minister Bhutto were worried that India would rate Kashmir, then its legitimacy could also be have the bomb soon. In fact, Bhutto’s famous challenged. statement that Pakistan would eat grass if nec- essary to build a nuclear weapon was made to a Kashmir Again: British journalist in 1965 in response to a ques- Indo-Pakistani War of 1965 tion of what Pakistan would do if India got the After the early battles over Kashmir, India and atomic bomb.8 Pakistan managed not to fi ght a major war for As part of President Ayub Khan’s policy another sixteen years. The next one, howev- of “leaning on India” to test Prime Minister er, was again largely about Kashmir, although Shastri’s mettle, in early 1965 Pakistan sent conventional forces also clashed farther south. patrols into the Rann of Kuch, an area only The 1965 war demonstrates one way in which above water about one-half of the year, near conventional (and possibly nuclear as well) Karachi and the Arabian Sea.9 India responded deterrence can fail on the subcontinent.6 with its own military patrols, but the Paki- In late 1963, unrest broke out in Kashmir stanis outmaneuvered them, and the Indians because of the disappearance of an Islamic reli- withdrew as the rainy season began. Shastri gious relic. Although the relic was returned a agreed to submit the Rann dispute to legal year later, the unrest was one of several vari- arbitration, drawing howls of protest from the ables that caused the Pakistani leadership to Indian press and Parliament, and feeding Pak- believe the time was ripe to test India’s con- istani beliefs that India was politically and trol of Kashmir. In addition, Pakistani leaders militarily weak. saw India’s military buildup after New Del- The summer of 1965 began with escalated hi’s defeat by Beijing in a 1962 border war rhetoric and a probing attack by Pakistan in

56 Nuclear Stability in South Asia the Kargil sector of Kashmir. India successfully joined in, calling for a halt to the fi ghting. UN counterattacked and took some territory held Secretary General U Thant’s early September by Pakistan since 1948. Under pressure from visit to the region was unsuccessful in stop- the United Kingdom, the two sides agreed to a ping the fi ghting, but after ten days of intense cease-fi re on 27 June. In early August, Pakistan fi ghting, forces on both sides stalled, neither launched Operation Gibraltar, and thousands being able to advance upon key objectives – for of irregulars and special units infi ltrated across Pakistan the road to Srinagar and for India, the cease-fi re line into Kashmir from Azad Lahore. Kashmir and possibly Pakistan proper in hopes In an interesting twist to the crisis, on Sep- that the locals would rise up to help them over- tember 17 China threatened to broaden the throw Indian rule.10 However, local Kashmiris confl ict. It gave India an ultimatum to remove did not rise up and in fact called the police to certain construction works in Tibet or face inform on the raiders.11 In late August, India what it termed “grave consequences.”13 China successfully counterattacked, pushing many of had provided rhetorical support for Pakistan the infi ltrators back, capturing some key pass- at the outset of the crisis, but this escalation es, and seizing more Pakistani territory. In brought both the United States and Moscow in response, on September 1, Pakistan escalated, on the other side. Washington and Moscow’s striking with armor into southern Jammu near warning led Beijing to extend its deadline, the narrow area known as the “chicken’s neck” diluting the ultimatum. Finally, faced with and threatening a key road approaching the a third UN Security Council resolution, this city of Srinagar, which if captured would have one demanding a cease-fi re, and recognizing cut off India’s forces in Kashmir. Both sides’ both a stalemate on the ground and ammuni- air forces engaged at this point, and India tion shortages, India and Pakistan accepted a responded by crossing the international border cease-fi re on September 20 and 22.14 Both sides on the ground farther south, attacking with claimed victory, but the result was a stalemate, its armored forces toward Lahore and Sailkot. forces on both sides returned to their original This move raised the stakes of the confl ict positions, and control of territory reverted to considerably, as Indian forces were striking at the status quo ante. The Tashkent Declaration the heart of Pakistan and threatening, if they of January 10, 1966, brokered by the Soviet advanced, to cut the country in two. Pakistan’s Union, included return of prisoners of war and armored counterattack against this thrust was a withdraw to original lines of control, but ambushed at Khem Karan in Punjab, and the it did not attempt to settle the larger issue of overall outcome of this large tank battle was a Kashmir.15 stalemate.12 Lessons Attempts to De-escalate the Crisis New Delhi was pleased with the military out- By this point, the international community come of the 1965 war; India had successfully was concerned and directly involved in try- held Kashmir, countered Pakistan’s escalation, ing to stop the fi ghting. The United States, the and fended off China’s ultimatum. Some ele- United Kingdom, and the United Nations all ments in India were not pleased with the pressed both sides for restraint. Even Moscow Tashkent Declaration because India gave up

Stability Issues from Past Wars and Crises 57 some of the territory it had captured, par- Pakistan would move to developing an active ticularly some key passes in Kashmir, and program, starting in 1971. because the accord failed to name Pakistan as One analysis postulates that the 1965 war the aggressor.16 Oddly enough, the victory pro- provides a good study as to why conventional duced renewed calls in India for the accelerated deterrence can fail, particularly in the South development of nuclear weapons. The drivers Asia context. Sumit Ganguly posits that the for these calls were China’s ultimatum (Beijing military balance at the time would have argued had tested its fi rst nuclear weapon less than against Pakistan attacking and that the clas- a year before, in October 1964) and Washing- sic elements of deterrence were present: India’s ton’s cutting off of aid during the war. Both capability and credibility, recognized by Paki- incidents led many politicians and analysts to stan. His conclusion is that decision makers believe that India had to have nuclear weap- were not entirely rational and dispassionate ons. While Prime Minister Shastri publicly in their decision making, largely because of rebuffed these calls, India’s scientists apparent- domestic political factors. Specifi cally, India ly accelerated their research and development did not realize the impact that its rearmament efforts, moving India even closer to a weapons (directed at China after 1962) would have on capability.17 Pakistan’s calculations about the future. On Pakistan’s morale suffered a deep blow by the other side, Pakistan had fundamentally the loss. It had believed that it was martially fl awed estimates of military success, which in superior to India, if only in spirit. Moreover, turn were based on skewed estimates of the the fact that Pakistan had been unable to pro- military balance and an underestimation of voke a popular uprising in Kashmir was keenly Indian resolve. Not only was Pakistan’s anal- disappointing to the Pakistani leadership, who ysis about the static nature of the balance had been sure it would occur. Islamabad also fl awed, but it had the incorrect, but powerful felt for the fi rst time a threat by India outside of belief that time was not on the side of Islam- Kashmir as New Delhi showed both a willing- abad.18 The conclusion to be carried forward ness and a capability to threaten the heartland from this war is that rational deterrence theo- of West Pakistan with armored maneuver forc- ry, at least that based on conventional weapons, es. This lesson, and worry, were to be repeated needs to be adjusted for cognitive and affective in 1971, 1987, and 1990. Finally, Pakistan too limitations brought on by such variables as per- learned that self-reliance was imperative for sonalities and domestic politics. future confrontations. While China made the right noises at the outset of the crisis and The 1971 War – Bangladesh tried an ultimatum, it was trumped by the two The 1971 war had its origins in a domestic superpowers. As noted above, Pakistan might political crisis in Pakistan. This in turn was have been worried about India’s progress in the the result of the fact that, at the time, Paki- nuclear fi eld, and this concern could have been stan was a geographically divided country, each part of the reason for the 1965 confl ict. If so, half with a different ethnic makeup, but it was it would then make sense that Pakistan began ruled by a central government, and military, thinking seriously at this point about develop- dominated by Punjabis from the west. The ing its own atomic bomb. As described below, political issues this situation raised came to a

58 Nuclear Stability in South Asia head in December 1970 when the ethnically Attempts to End the War Bengali, Awami League from more populous This war occurred amid a swirl of interna- East Pakistan won a majority of seats in Parlia- tional activity involving the United States, the ment. West Pakistan, home to the political and Soviet Union, and China that colored the inter- military elite of Pakistan, did not want to sur- national reaction to it and skewed attempts to render its control of the state, and negotiations bring it to an early conclusion. Throughout ensued between Zulfi kar Ali Bhutto and Yahya the spring of 1971, the United States was Khan for the west and Bengali leader Muji- negotiating its opening to China, which was bur Rahman for the east, seeking to fi nd formally announced on July 15, 1971. Pakistan some modus vivendi. The negotiations broke played a facilitating role in this entente, and down without agreement in March 1971, and Washington therefore was inclined to support the West Pakistan-controlled military cracked Islamabad’s position in confl icts with India. down on a burgeoning independence move- Nixon and Kissinger decided that the United ment in East Pakistan. East Pakistan’s political States must tilt toward Pakistan, the weaker of rulers fl ed to India and formed a government- the two sides, as part of its grand strategy. This in-exile, and refugees from East Pakistan fl ed tilt, and its consequences in terms of China, the spiraling violence into India, setting off an took tangible form when Kissinger appar- international crisis that would lead to India ently told the Indians in August of 1971 and Pakistan’s third war. that the United States would be “unable to Bengali insurgents from East Pakistan help you against China in the event of Chi- began operating out of India, and in July, na’s involvement in an Indo-Pakistan war.”21 India decided that it would support a war to Given Washington’s involvement in convinc- “liberate” East Pakistan (eventually renamed ing Beijing to back off its ultimatum in 1965, Bangladesh).19 In the fall of 1971, India’s assis- this confi rmed for New Delhi its need for a tance to the rebels was ramped up to include different alignment. At the same time, India artillery support for their raids, logistical sup- unveiled its new treaty of friendship with the port, and bases in India. India also apparently Soviet Union, further alienating Washington had planned for an all-out invasion in early and coloring the latter’s view of the struggle on December to liberate the East Pakistani capital the subcontinent. of Dhaka, but the Pakistanis acted preemptive- As a result of this set of events, the ly with air strikes on India on December 3 United States sided against India. Specifi cally, and formally declaring war on December 4. Washington cut off all assistance to India on India responded with its planned invasion on December 2 and on December 10 it sent the December 6, and it captured Dhaka ten days aircraft carrier USS Enterprise and its battle later. On the same day, Pakistani forces in East group into the Bay of Bengal. Simultaneously, Pakistan surrendered, and a day later, Pakistan Washington advised Pakistan not to accept agreed to an Indian-proposed unconditional a cease-fi re limited only to the east. It also cease-fi re in the west as well, where clashes gave Moscow an ultimatum, calling on it to had been occurring along the cease-fi re line in reign in its new “friend,” India. The Soviet Kashmir and south in Punjab.20 Union responded that India had no aggressive intentions in the west, and Indira Gandhi

Stability Issues from Past Wars and Crises 59 echoed this position publicly on December relationship with Moscow. In addition, China 12.22 Finally, on the same day China told was shown not to be willing to challenge the United States that it had no interest in India’s security directly. For all of these rea- intervening, pulling out the last prop from sons, George Perkovich argues that the 1971 under Pakistan. The international intervention war “…did not create a security imperative to was crucial because each side refused to develop nuclear weapons.”26 Some argue that negotiate or communicate in any way with the the Enterprise (and the implicit nuclear threat other during this crisis.23 it carried) created a rationale for India to take The 1971 war was closed out with the the next step in its nuclear development, but negotiation of the “Text of India-Pakistan Perkovich argues persuasively that this was Agreement on Promotion of a Friendly Rela- more indirect and merely built on India’s desire tionship,” more popularly known as the Simla for prestige and an improved self-image that in Pact, in June and July 1972. Pakistan argued turn fed its nuclear program. In fact, the chro- that it negotiated the agreement under duress nology of the decisions on the subsequent 1974 as India still held large numbers of Pakistanis “peaceful nuclear explosion” is very diffi cult as prisoners of war.24 The pact, which com- to piece together, but much evidence points to mits both countries to solving their differences decisions on moving ahead having been taken peacefully, is to this day interpreted differently before both the 1971 war and the entry of the by the two sides. Indians emphasize the provi- Enterprise into the Bay of Bengal.27 sion that all issues, including Kashmir, should The lessons Pakistan took away from the be settled bilaterally while Pakistan points out 1971 war, however, were quite different. First, the mention of the United Nations in the pact the war confi rmed the worst fears of Pakistanis and argues therefore that the UN and its earlier – that India had both the desire and the abil- resolutions on the Kashmir issue should still ity to dismember Pakistan using conventional be applied.25 Finally, the pact called for both forces. In fact, Pakistanis believe that India was sides to respect and newly demarcate the cease- behind the Awami League and Bengali region- fi re line in Jammu and Kashmir, turning it into al separatism in the east.28 Second, Pakistan’s what was termed a line of control – essentially potential ally, China, was reducing its inter- a de facto border. est in the subcontinent and would not be much help in the future conventionally if Pak- Lessons istan were threatened by India. Unlike in The lessons of 1971 for India were of two types 1965, China this time did not even attempt to – one related to Pakistan and another related pressure India and even Pakistan’s role in facil- to its place in the international order. In terms itating U.S.-Chinese ties was insuffi cient to of Pakistan, the 1971 war was a clear victory bring about helpful actions from Beijing. for India militarily, bolstering its confi dence U.S. intervention against India, while wel- vis-à-vis its rival and eliminating its diffi cult comed by Islamabad, came too late to “save” two-front strategic situation with the inde- East Pakistan. The result was that Pakistan, pendence of Bangladesh. On the international under Bhutto’s leadership, took formal steps to front, Soviet diplomatic support coupled with launch a nuclear weapons program in January the formal agreement solidifi ed New Delhi’s 1972.29

60 Nuclear Stability in South Asia Issues Asian media about possible attacks on Paki- One issue that became clear in the 1971 war stan’s nuclear facilities. The sources for the and the international community’s reaction to articles were most often intelligence from the it, is that crises in the region do not occur in United States or statements by Pakistani lead- a vacuum. The level of activity and engage- ers in response to questions about these press ment by the United States, the Soviet Union, articles. The scenarios variously had Israel, and China were all dictated by larger consid- India, the two in some combination, or the erations and their maneuverings vis-à-vis one Soviet Union preparing to attack Pakistan’s another. A second issue is that again domes- nuclear facilities, including reprocessing and tic politics and political changes were catalysts enrichment facilities in Kahuta.31 The prece- to war. Neither side anticipated that the situa- dent set by Israel’s 1981 preemptive attack on tion in East Pakistan would result in a war, the Osirak nuclear reactor in Iraq apparently at least not at the outset of the crisis. Third, fi red the imagination of the media, Pakistanis, each viewed the intentions of the other as intelligence analysts in the United States, and implacably hostile. This colored all planning possibly military planners in India. and decision making, and decisions that result- In September 1984, leaks of a U.S. intelli- ed were based on narrowed expectations and gence briefi ng to the Senate caused the United fi xed perceptions based on history and fi xed States to make public statements that it would patterns of thinking.30 Finally, the relation- support Pakistan if it were attacked in such ship between security and decisions about the a way. India responded by seeking a corre- need for nuclear weapons remains complex. sponding assurance from the Soviet Union India detonated a nuclear weapon three years that it would support its client if the United after the war despite the fact that its victory States took any action.32 The United States and the attendant solidifi cation of its ties with also apparently told President Zia of Pakistan the Soviet Union should have reduced a secu- that it saw no immediate threat and that it rity motive for moving ahead with its nuclear would warn Islamabad if it saw signs of an program. While as Perkovich argues, prestige imminent attack.33 Other reports allege that and other domestic infl uences were signifi cant Indira Gandhi considered such an attack but in Indian decisions on moving ahead with its dismissed it because of a concern that Pak- program, the recent war, regardless of outcome, istan would retaliate against Indian nuclear likely helped to support the advocates of test- facilities.34 A more complex scare arose in the ing India’s weapons capability, if only with middle of 1986, when a Soviet demarche to a “peaceful” demonstration. Pakistan, on the Pakistan was interpreted as a threat to attack other hand, provides clearer evidence of a more if Pakistan acquired nuclear weapons. Again, direct relationship between perceived security Islamabad engaged Washington, which sent a requirements and a desire to obtain nuclear demarche to Moscow and was told that it was weapons. all a misinterpretation.35 The early bilateral scares probably led to the agreement between The Preemption Scares of the 1980s President Zia and Prime Minister Rajiv Gan- Throughout the early and mid 1980s, various dhi, announced on December 17, 1985, not to reports surfaced in the U.S., British, and South attack each other’s nuclear installations. This

Stability Issues from Past Wars and Crises 61 agreement was formally signed on December eering expeditions mounted from Pakistani 31, 1988 and ratifi ed on January 27, 1991.36 territory to Siachen and the publication of These scares were not only prompted by several atlases showing the line of control the Israeli example. They were also based on extending to the Chinese border in a way that changes and perceived changes in the strate- substantiated the Pakistani claim to control the gic stability equation between the two sides. glacier.37 In one case, Pakistan was concerned when In response to what New Delhi perceived India acquired modern, long-range strike air- as a weakening of its claim to this territory, craft that were capable of low-level penetration the Indian army launched Operation Meghdoot of Pakistani airspace – the U.K.-manufactured on April 13, 1984. The Indian military occu- Jaguars. On the other side, Pakistan’s nuclear pied key passes in the Saltoro range dominating weapons infrastructure was still perceived to the approaches to the glacier, and subsequent be limited in size and scope – an inviting target Pakistani attempts to dislodge them failed.38 for preemption. Once Pakistan crossed the line Fearing that ceding any ground or rights in this to weaponization and had produced enough disputed area would cascade to further losses fi ssile material, it could easily be dispersed to in Kashmir, both India and Pakistan commit- various locations, making it more diffi cult to ted substantial forces to the battle. A little over target. If India was actually considering such a year-and-a-half after the battle began, the two a strike, it would have been doing so because sides agreed to negotiate bilaterally in hopes of it believed that the broad strategic balance solving the dispute and easing the fi ghting. Six between the two was changing and that its rounds of talks were held through 1992 with window for any sort of preemptive action was no resolution. Immediately before the 1990 closing. crisis, the fi fth round appeared to produce some progress, but both sides quickly retreated Siachen Glacier from any apparent concessions, and the rising The Siachen glacier, an inhospitable piece of tensions in the rest of Kashmir scuttled any territory in the north of Kashmir, became the optimism.39 Meanwhile, the fi ghting continues site of military confrontations between India to this day, at altitudes ranging from twelve and Pakistan in 1984. Largely because of its thousand to twenty-four thousand feet with altitude, inaccessibility, lack of population, and Indian forces still holding the high ground and lack of strategic importance, the UN did not Pakistani forces still seeking to dislodge them. demarcate the glacier during the cease-fi re line Since 1984, over four thousand soldiers have exercise in 1949. In the aftermath of the Simla been killed and over twice that number have Pact, India and Pakistan chose not to remedy been wounded, providing an ongoing source of this anomaly when they demarcated the line tension between the two sides.40 of control. As a result, it was unclear exactly While not a crisis in which nuclear weap- who had either control of, or more important- ons have played any part, the ongoing fi ghting ly, rights to this piece of territory. In the late in Siachen continues to poison the atmosphere 1970s and early 1980s a series of events greatly between the two parties. Moreover, it is a crisis increased worry in New Delhi. These events in which there is relatively little international included increased international mountain- intervention, largely because no one has ever

62 Nuclear Stability in South Asia seen it as having a chance to escalate. Islam- Sikh uprising. Which of these three explana- abad has decried this lack of interest, claiming tions is the truth, or whether it was some mix that the international community failed to act of the three, probably will never be known.41 when India unilaterally changed the situation The result of Sundarji and Gandhi’s early on the ground in disputed territory, some- 1986 discussions was a series of military exer- thing both sides pledged not to do in the Simla cises dubbed Brasstacks I-IV. Brasstacks I and Pact. More recently, Pakistan notes that the II were map, computer, and sand table exercis- international condemnation of the 1999 Kargil es. Brasstacks III, conducted in November and operation is evidence of bias against it because December of 1986, was a fi eld training exercise India was never similarly condemned for what conducted in segments by different services they claim is a similar exercise on Siachen in and parts of the army. Brasstacks IV, scheduled 1984. for February to March 1987, was supposed to pull it all together in a fi eld training exercise The Brasstacks Crisis of 1986-1987 using combined arms and multi-service coor- In 1986 Prime Minister Rajiv Gandhi and the dination. This last portion of the exercise was newly minted army chief of staff, General K. to take place in the Rajasthan desert near the Sundarji, wanted both to show off and to test Pakistani border. India’s growing conventional capabilities. In The political script for the exercise built on particular, Sundarji wanted to test the ability recent events and in many ways anticipated the of the Indian armed forces to conduct large- actual crisis of 1990. The exercise scenario had scale, combined-arms warfare with air support events in Kashmir spiraling out of control and and new command and control concepts, build- the simultaneous declaration of an independent ing on the air-land battle concepts that he state of Khalistan by Sikh militants. While the learned while attending the U.S. Command former was not even a remote possibility at that and General Staff College. Gandhi, it seems, time, the Sikh independence movement was still did not have strictly military purposes in mind very much in the minds of Indian planners. and was apparently fascinated simply by the Only three years before, in June 1984, the Indi- idea of conducting a very large exercise – larger an army had conducted Operation Bluestar, an than anything ever conducted in South Asia. assault on the Sikh Golden Temple complex, Other theories suggest that there were addi- resulting in the death of an estimated six hun- tional motivations for the exercises. Although dred to one thousand people. The Sikh rebel- key participants deny this, some circumstan- lion was claimed by the Indians to have been tial evidence suggests that at least Sundarji and supported by foreign powers, namely Pakistan, Defense Secretary Arun Singh may have been and the scenario for the exercise played out this setting up a situation that would coax Paki- supposition. In response to the events in Kash- stan into attacking India, thereby providing a mir and Khalistan, the exercise script called for pretext for striking back and crushing the Paki- Pakistan to make a military thrust into India to stani military. A third possibility is that India’s help liberate these two lands. In response, the leadership wanted to use the exercise to send a exercise had the Indian army halt the invasion political signal to Pakistan, warning it to cease and undertake a counteroffensive to carry the what New Delhi saw as its meddling in the confl ict into Pakistan.

Stability Issues from Past Wars and Crises 63 The exercise transformed into a crisis after United States. Regardless, the communications the Pakistani government became nervous between the two sides about the exercise dur- about Indian intentions (because of its own ing late 1986 appear to have been less than internal unrest problems in the Sindh) and satisfactory from the Pakistani perspective, responded by extending the stay of its own mil- despite the fact that they apparently occurred itary units in the vicinity of the border after at quite high levels including the directors gen- their winter exercises ended. Pakistan’s Army eral for military operations (DGMO) and even Reserve South, consisting of an armored divi- the prime ministers at the time. Interestingly, sion and an infantry division, remained in the however, as the crisis continued and indeed Bahawalpur area near the border after their escalated into 1987, direct communications via exercise ended. Army Reserve North, also con- the DGMO hotline were not used (December sisting of an armored division and an infantry 8, 1986 to January 23, 1987). division, moved to the Gujranwala areas in the In late January both sides were suffi cient- Ravi-Chenab corridor, west of the Shakargarh ly concerned to reopen several channels of salient. In addition, Pakistan took other steps dialogue with one another and begin making to prepare for possible confl ict, including acti- somewhat less alarming statements to the press. vating some reserves and moving ammunition The hotline between the two DGMOs was acti- forward. Finally, in early January 1987, Pak- vated on January 24 and public statements by istan moved its Army Reserve South across both sides indicated a desire to de-escalate the a key river into a position that was at least crisis. On January 25, the two prime ministers theoretically better for an offensive pincer spoke by phone, further easing tensions, and movement into India. These moves in turn trig- on January 31, the two sides met at the foreign gered Indian anxieties about an invasion by secretary level for several days of negotiations Pakistan roughly along the lines envisioned in on how specifi cally to de-escalate the crisis. An the exercise scenario. New Delhi responded by accord was reached on February 4 whereby both manning defensive positions close to the bor- sides promised not to attack each other, pledged der and making other military preparations to to avoid provocative actions along the border, blunt any potential Pakistani military thrusts and began to pull their units back from the Ravi- across the border. Chenab corridor over the next two weeks.42 The pullouts began, starting with the Paki- Attempts to Prevent War and stani Army Reserve North, and on February De-escalate the crisis 21, President Zia traveled to India for an Indian- From the fi rst detection of the exercise plan Pakistani cricket match (a trip dubbed “cricket by Pakistani intelligence in August 1986, Paki- diplomacy”), bringing the crisis to a close. India stani offi cials sought to gain clarifi cation about ended up holding the Brasstacks IV maneuvers the size and scope of the exercise. Some on the in March, but they were watched closely by the Indian side questioned the motives for Paki- press and the diplomatic corps. stani expressions of concern, arguing either that they were unwarranted or that they were Lessons merely intended to create a crisis atmosphere Interestingly, neither Pakistan nor India has and wring more military assistance out of the seen fi t to conduct any formal lessons-learned

64 Nuclear Stability in South Asia studies of this crisis. Scholars outside the coun- tion of the Indian Punjab is seen as decisive try or national scholars residing outside of in deterring New Delhi from attacking into South Asia have initiated the vast majority of the Sindh.44 For India, the Brasstacks crisis the analyses. In part this may be because nei- showed that at least in some situations India ther side ever had any intention to go to war was still militarily vulnerable to Pakistan, and therefore the crisis is not perceived as seri- which many policy makers had thought was a ous. There was also no nuclear play in this dead issue after it was “solved” in 1971. Both crisis, perhaps adding to the perception that its political and military hierarchies were genu- origins, progress, and resolution are not wor- inely disturbed by the crisis, and it is believed thy of extensive study.43 This is unfortunate that they therefore approached the next cri- because Brasstacks represents an interesting sis, in 1990, more cautiously. They also may chance to examine a crisis from at least two have begun to think that conventional war as a perspectives depending on which theory one method of competition was getting too danger- subscribes to about the motives behind India’s ous, not because of escalation to the nuclear initiating the exercises. If one believes that level but because of the combination of geo- India’s intentions were benign, then the crisis graphic vulnerabilities and domestic political becomes an interesting study of how the classic problems. Finally, for India, Brasstacks was security dilemma works with chances to study something of an embarrassment to the civilian the specifi cs of the action-reaction cycle in this leadership as it was led along by the military part of the world. If one believes that India into a crisis that got quickly out of hand. was either baiting Pakistan or sending it a mes- sage of resolve and military strength, then the study changes to one of how countries develop Issues such strategies, believe they can control the out- When the problem is considered from a non- comes, and de-escalate if the process gets out participant’s the point of view, a number of of control. issues arise that are useful to consider for Despite the absence of formal studies, par- preventing future crises or managing them if ticipants and analysts in each country have prevention fails. First, the general view of the drawn different lessons from the crisis. For other side in the crisis was highly negative, Pakistan, the entire affair reinforced its mis- which colored readings of intentions and led to trust of India and its intentions. Specifi cally, assuming the worst about any capabilities and it supported the theory that at least some key moves. As the crisis evolved, each side read the fi gures in Indian politics and the military con- other’s defensive moves as being designed to tinue to believe in a military solution to that enhance its opportunities for offensive action. country’s problems with Pakistan. In other Second, in a different version of mispercep- words, India’s strategy of 1971 – intervening tions than merely assuming the worst about and supporting separatist/independence move- the adversary, the Indian decision makers con- ments in parts of Pakistan – was successful sistently discounted the effect their moves and and should be considered for use again. The statements were having on Pakistan. Third, outcome, for Pakistan, is viewed widely as while intelligence was fairly plentiful on both a victory, and the countermove in the direc- sides, it was not timely and often it was dis-

Stability Issues from Past Wars and Crises 65 counted or not believed, owing to preconceived ing presidential rule from New Delhi and notions about intentions. When information sending a former governor with a hardline was exchanged by the two sides, it was often reputation to crack down. Pakistan took advan- inaccurate or sketchy, feeding worst-case sce- tage of the situation, with the Inter-Services narios and suspicions.45 Fourth, there was Intelligence Agency (ISI), army, and other orga- little outside intervention in the crisis, with nizations providing increased material support key actors like the United States playing only to the insurgents, opening more training camps minimal go-between roles. Exactly why this in Azad Kashmir, and whipping up the rheto- was the case remains somewhat unclear, espe- ric. Indian analysts claim that Pakistan had cially given the extensive involvement of the been providing support for, and in fact active- United States, the Soviet Union, and China ly fomented, the insurgency for years. Some in previous crises. One possibility is that the argue that the outbreak of high-level violence in superpowers and China did not realize that Kashmir in 1990 was a direct result of a Paki- there was a crisis until it was well on the way stani strategy to pursue low-intensity confl ict to being resolved. Finally, decision making in against India under the shield of recently weap- the lead-up to and during the crisis had con- onized nuclear capabilities.48 Others argue that siderable failures, particularly on the Indian it was inevitable given the growing alienation side. Analysts have identifi ed these failures of the Kashmiris from the rest of India and as being of two types: overcentralization of indeed from their own governing bodies. The decision making and lack of policy coordina- nuclear aspect of the crisis is discussed in more tion between branches of government. Another detail below. conclusion from the crisis, according to Dennis Had the issue remained a struggle between Hagerty, was that it drove both Pakistan and insurgents in Kashmir, whether indigenous India to increase the pace of their respective or infi ltrated, and Indian paramilitary and nuclear weapons development.46 army forces, it would not have qualifi ed as an international crisis. In fact, early moves by The 1990 Crisis the prime ministers on both sides indicated a The 1990 crisis between Pakistan and India desire to keep the crisis limited. Unfortunately, saw the revival of Kashmir as a central point neither prime minister was politically strong, of dispute. Most analysts date the beginning of and both were pressured into upping the ante the crisis from the kidnapping of the daughter both rhetorically and militarily. The Pakistani of the new Indian home minister, a Kashmiri, opposition began calling for a holy war, and Mufti Mohammed Sayeed. Sayeed’s appoint- one party called for Pakistan to develop nuclear ment by newly elected Prime Minister V.P. weapons to defend itself against India. Both Singh was meant to follow up on campaign to keep ahead of popular opinion and to fol- promises for national healing.47 After negotia- low the military lead, Pakistani prime minister tions for her release, the simmering insurgency Benazir Bhutto called for Kashmiri self-deter- in Kashmir exploded in January 1990. Violence mination. skyrocketed and the economy and daily life, At this early point in the crisis, nuclear particularly in the Vale of Kashmir, became weapons possibly came into play. In a Janu- almost impossible. India responded by impos- ary 21 meeting in New Delhi between the

66 Nuclear Stability in South Asia Pakistani foreign minister, Yakub Khan, and region in February 1990 and confi rmed that his Indian counterpart Inder K. Gujral, Khan neither army had signifi cant strike forces in made a statement about war clouds gathering positions necessary for offensive operations. that some in New Delhi took to be a veiled These reassurances did not work, and each side nuclear threat. When confronted later in continued both to fear the worst and to prepare the day by Indian offi cials, Khan apparently to take advantage of the situation should the claimed that he had been misinterpreted.49 other move precipitously. Some analysts point to the fact that Khan was By April, the conventional situation was as closely linked to the military and was given a follows. In Kashmir, India had deployed two mandate to deliver a harsh message. The same hundred thousand army and paramilitary forc- analysts note that Khan may, however, have es, supplemented by seventeen thousand local exceeded his mandate and delivered an actual police, to deal with the insurgency. On the Pak- threat.50 Thereafter the rhetoric continued to istan side of the border of Kashmir, in Azad escalate and both sides began making conven- Kashmir, Pakistan had deployed one hundred tional military preparations in areas outside of thousand troops. Indian and Pakistani troops Kashmir. were in very close contact across the line of In many ways, the military moves that control. Farther south in Punjab, both sides escalated the crisis were similar to those in the had infantry units deployed in their forward Brasstacks crisis of 1987. Certainly that crisis defensive positions, but major armor and artil- was on the minds of all of the key participants lery units were still in their cantonments some in both countries in 1990 and in the mind of distance from the border. Finally, in the Lahore the most critical international observer, the sector, India had deployed two infantry divi- United States. The Pakistanis conducted a sions, in penny packets, near the border to help conventional military exercise in late 1989 stop potential infi ltration. The Indian con- called Zarb-I-Momin. It was reminiscent of cern that prompted the last deployment was India’s Brasstacks exercise and was meant by that terrorists could infi ltrate Punjab, sabotage Pakistani army chief General Beg to send a key lines of communication, and thereby clear dissuasive message to India about meddling the way for a Pakistani army invasion.51 In with Pakistan. After the exercise ended the sum, both sides were actually restraining them- major units, including an armored division and selves, keeping their mobile units capable of a squadron of Apache helicopters, stayed in the offensive operations in their cantonments, but fi eld. India responded by moving some units to they were making signifi cant defensive prepa- defensive positions along the border and keep- rations against an expected move by the other ing some tank units at the Mahajan fi ring range side. in Rajasthan, the scene of Brasstacks III and IV. Fearing that the increased insurgent activ- Attempts to Prevent War and ity in Kashmir might be replicated by Pakistan De-escalate the Crisis in the Indian Punjab, New Delhi sent more bor- Attempts by Washington and Moscow to coax der security and army infantry units into that the two sides into a joint statement on restraint region. In an attempt to defuse tensions, Amer- failed, and in mid-May the now-famous trip ican military attachés traveled to the border to the region by U.S. Deputy National Secu-

Stability Issues from Past Wars and Crises 67 rity Advisor Robert Gates took place. What Gates visited key offi cials in Islamabad fi rst prompted the trip is still a matter of debate, and told them that Pakistan would lose any but most of the participants say that the United conventional war with India. He also outlined States had become increasingly worried that for Pakistan the political and diplomatic down- the posturing would spiral into a large con- sides of a war. Before Gates left for New Delhi, ventional war that could then, down the line, he was told by the Pakistani foreign minister lead to a possible nuclear exchange.52 However, that Islamabad would close training camps two journalistic accounts claim that the Gates for the insurgents located in Azad Kashmir.55 mission was prompted by intelligence reports This promise was later denied by the Paki- that painted a picture of Pakistan preparing to stanis. Gates was well received in New Delhi, assemble and use its nuclear weapons.53 and the conversation there apparently focused The complete record remains classifi ed, and on the conventional issue and the insurgency not all of the key participants have chosen to go in Kashmir with no mention of nuclear weap- on the record to clarify the issue. However, a ons.56 Gates also provided both sides with a number of elements appear to have been operat- list of possible confi dence-building measures ing that led to varying accounts and differing (CBMs) that could be useful in de-escalating interpretations. First, as always, the intelligence the crisis. Had the United States believed that was ambiguous and was interpreted different- there was a possibility of Pakistan’s using ly by various agencies in the U.S. government. nuclear weapons in the near term, presumably Some of Pakistan’s moves detected at the time Gates would have mentioned it. This is evident could have been in preparation for readying by the role that the United States took in the a nuclear weapon for use, or they could have preemption scares in the 1980s, promising to been precautions against a possible Indian con- provide information if a strike seemed immi- ventional strike on Pakistani nuclear facilities. nent. Islamabad could also have been trying to send a Gates’ trip was not viewed as a success signal to the United States that it would consider upon its completion. In fact it took several using its nuclear weapons in extremis, hoping weeks for the two sides to take visible steps that Washington would intervene and defuse to defuse the crisis. These included India’s the crisis. If that was the case, it was a risky announcing that its armored units, presumably gamble. Second, a number of activities related those at the Mahajan fi ring range, had been to Pakistan’s nuclear weapons programs appear withdrawn. India also proposed a list of CBMs, to have been going on at once. Sometime dur- based on those received from Gates that were ing the crisis, it appears that Pakistan restarted largely agreed to about a year later.57 elements of its weapons program that had been frozen at the insistence of the United States.54 Lessons This may not necessarily have meant that Paki- For India, the 1990 crisis provided further stan was assembling a nuclear weapon. It could evidence that Pakistan was fomenting and sup- have merely meant that it was restarting enrich- porting the insurgency in Kashmir. New Delhi ment facilities. This could have led to some believed that in the late 1970s, Pakistan real- degree of confusion both in intelligence circles ized that it could not take Kashmir by force and among policy makers. using conventional arms, so it began a strat-

68 Nuclear Stability in South Asia egy of terrorism and secession in both Punjab its own. This probably enhanced the belief and Kashmir, with the hope that the latter at in Islamabad that having a relatively robust least would eventually be split off from India. nuclear capability was going to be necessary. In support of this strategy, India believes, Pak- The U.S. detection of Pakistan’s resumption of istan developed nuclear weapons under the certain activities in it nuclear program, such umbrella of which it could carry out the insur- as enrichment of uranium above a prohibited gencies without fear of Indian escalation. In level, provides evidence of this belief. fact, many in New Delhi have concluded that Pakistan achieved a nuclear weapons capabil- Issues ity in 1987, leading to an increase in violence Unlike 1965, Siachen, or 1987, this crisis was (due to infi ltrations) in Punjab and Kashmir not set off by specifi c actions of one or the other in the 1988-90 timeframe.58 Because the early government. It resulted from the explosion of 1990 violence in Kashmir exploded so quickly insurgency in Kashmir, which Islamabad cer- into an international crisis, New Delhi did not tainly desired but did not cause directly or all review its own governing strategy in the region on its own. The violence took both govern- or have cause to believe that it was at all to ments by surprise.60 This upsurge in violence blame for the violence. India did not believe did not have to lead to an extended and seri- that Pakistan would actually use nuclear weap- ous crisis, but the domestic weakness of both ons but rather that it would bluster about them regimes helped to fuel the escalation. Both gov- in order to bring in the international commu- ernments became locked into positions early on nity to either rescue it or freeze in place any and were too weak to back down without out- advantages it had gained during its incursions side intervention.61 Neither side wanted war, into Indian-controlled Kashmir.59 but each found it diffi cult to take the concrete For Pakistan, 1990 was a crisis with mixed steps necessary to back down. This, coupled lessons. On the one hand, Kashmiris took up with the rhetoric that each side felt was neces- the insurgence cause with a vengeance, some- sary for domestic political reasons, escalated thing they had not done in 1965, 1971, or even the crisis. The fact that neither side necessarily throughout the 1980s when support from the wanted war also rules out using this crisis as ISI was on direct offer. This encouraged those a test of whether nuclear deterrence works on in Pakistan who thought that a strategy of the subcontinent. “bleeding India from a thousand cuts” might Unlike Brasstacks, the issue was not a have promising prospects. In addition, Gen- lack or breakdown of communication. During eral Beg succeeded in essentially pulling a this crisis, communication was fairly regular Brasstacks on India by staging his own large between the two sides, and as noted, included exercise that got New Delhi to react nervous- direct meetings early on that fueled rather than ly. On the other hand, the Gates mission, dampened the tensions. However, as in previ- with its stern message that Pakistan could ous crises and wars, the information provided not win and that the United States would not by each side to the other was completely mis- help it, showed that the strategic partnership trusted. Each side’s own intelligence was also with Washington was truly over, and Pakistan not of high quality, and analysts on each side would have to confront India in the future on assumed the worst of the other side. The Unit-

Stability Issues from Past Wars and Crises 69 ed States used fairly creative diplomacy in the Pakistani military and intelligence servic- an attempt to help solve the information and es, and the troops were a mix of Kashmiris, intelligence problem, providing reports from Pakistani regular forces, and mujihadeen fi ght- its military attachés to each side. However, ers from Afghanistan (some of these being these reports were not suffi ciently persuasive Saudis and others who originally came to the to move either India or Pakistan from its posi- region to fi ght the Soviets in Afghanistan). tion. Throughout this crisis, it is likely that the In addition to the military signifi cance of United States had both a clearer and a more the occupied positions, the incursion had politi- clear-eyed view of the actions of each side than cal signifi cance for three reasons. First, despite they did of one another. regular cross-border activity by insurgents and As noted above, this was not a nuclear cri- shelling by both sides across the line of con- sis. Pakistan did not directly threaten India trol, this represented the fi rst time since the with nuclear weapons, either during Khan’s 1965 war that portions of Indian Kashmir were initial trip or later through signaling actions. If occupied by Pakistani, or Pakistani-allied, Pakistan did signal to the United States its seri- troops. Second, the planning, infi ltration, and ousness through some movement of its nuclear occupation occurred when the prime ministers assets, it was a clumsy try, and it easily could of the two countries were meeting in Lahore in have been either missed by Washington or seen February 1999 in an attempt to improve rela- and quickly misinterpreted by New Delhi, lead- tions between the countries. Finally, Indian ing to an even more serious crisis. prime minister Vajpayee and his BJP-led gov- erning coalition had lost a vote of confi dence Kargil 1999 and were due to face elections in September. From February to May 1999, between six hun- Having been taken by surprise, the Indian dred and seven hundred fi ghters from Pakistan military reacted as swiftly as it could, consider- infi ltrated into Indian Kashmir in the Kargil ing the operational diffi culties associated with and Dras sectors. These forces occupied key the geography of the region. It moved army high points from which they could threaten forces into the region and on May 26 began Highway 1A, a major supply route for Indian using fi xed-wing aircraft and helicopters to forces on the Siachen glacier. These high points strike at the encampments. In addition, both were usually contested during the summer, but India and Pakistan used artillery to shell one the inhospitable nature of the geography meant another across the line of control and, in India’s that they traditionally remained unoccupied case, to interdict supply lines to the infi ltra- in the winter months. The infi ltrators took tors from Pakistan. India’s use of fi xed-wing advantage of this fact and occupied abandoned aircraft drew harsh verbal responses from Pak- positions. The identity of these forces remains istan, but both sides refrained from having a hotly contested point between India and their air forces cross the line of control. India Pakistan, with the former claiming they were lost two jets and one helicopter to surface-to-air Pakistani regular troops and the latter claim- weapons, but no air-to-air clashes took place. In ing that they were Kashmiri freedom fi ghters. a repeat of the confl icts of 1987 and 1990, both The truth is closer to New Delhi’s version, as sides alerted forces opposite one another far- the operation appears to have been planned by ther south, and armored strike forces on both

70 Nuclear Stability in South Asia sides were activated and elements were moved on July 3 to follow up. Sharif asked for a meet- to defensive positions near the border in the ing in Washington, and the next day a joint Rajasthan Desert region. statement was issued saying that concrete steps would be taken to reduce tensions. The United Attempts to End the Fighting and States elaborated on this by saying that Paki- De-escalate the Crisis stan would pull back.63 Discussions between the two sides, including The opposition, the military, and various telephone conversations between the prime jihadi groups roundly criticized Sharif for back- ministers and a visit by Pakistani foreign min- ing down. However, the then-chief of the army, ister Aziz in mid-June, failed to resolve the General Musharraf, stated at the time that he crisis. International efforts to lower tensions was comfortable with the deal that the prime did not meet with much success either. UN minister had made in Washington. As a prac- secretary general Kofi Annan’s offer to send tical matter, a pullback was at that point no a special envoy to both capitals was rejected longer the issue as India had succeeded in kill- by New Delhi, which argued that any delega- ing, capturing, or pushing out almost all of the tion should go only to Islamabad since Pakistan troops that had come into the Kargil sector was the source of the problem. American calls in the spring. Later, Prime Minister Sharif’s for restraint to both sides went unanswered attempt to distance himself from responsibil- although India did keep to its side of the line ity for the crisis and essentially throw the of control with its air attacks and it did not blame in the military led to his downfall. In send ground troops across the line of control. A October 1999, General Musharraf overthrew letter sent by Prime Minister Vajpayee to Presi- Sharif and took control of the Pakistani govern- dent Clinton in mid-June spurred Washington ment. to put more pressure on Pakistan to withdraw the fi ghters. In the letter, India warned that it Lessons might have to cross the line of control if Paki- Unlike some previous wars and crises, the Indi- stan did not back down.62 an government commissioned a formal review This warning, combined with the moves and lessons-learned study of the Kargil crisis. by both sides to activate larger conventional Interestingly, most of the lessons had to do with formations farther south, apparently led Wash- the internal Indian management of the crisis. ington to worry about escalation to a larger The broadest lesson was that India must mod- conventional war. As a result, the United States ernize its national security decision-making increased pressure on Pakistan. This pres- system to enable it to anticipate and deal with sure included a G-8 statement calling for an crises better, whether they stem from Paki- end to armed intrusion (although it did not stan or elsewhere. In addition to some minor name Pakistan directly) and a visit from the recommendations concerning years of service U.S. commander of Central Command, Gener- in the military and paramilitary forces, the al Anthony Zinni, to Pakistan to tell the army most sweeping organizational change was to and the government to pull back. After see- recommend that the entire national security ing few concrete results from the Zinni trip, management and decision-making structure be the U.S. president called Prime Minister Sharif studied and reorganized with an eye to creating

Stability Issues from Past Wars and Crises 71 a national defense headquarters and a position ously public debate and scholarly review are of chief of defense staff.64 Finally, the review extremely limited. However, discussions with committee recommended that a full-time posi- various offi cials and analysts in late 2000 tion of national security advisor be established revealed a basic consensus that Kargil was a along with a separate staff from the cabinet tactical success militarily, but a strategic disas- offi ce. ter politically.67 It was seen as enabling India to In the intelligence realm, the committee take the high ground on Kashmir in the eyes of recommended the following changes. First, global public opinion, and New Delhi was able India must vastly improve its intelligence col- to paint Islamabad both as the villain in this lection capabilities, particularly in the area of region and as an irresponsible party in a region satellite imagery and unmanned aerial vehicles with newly overt nuclear capabilities. Despite (UAVs). Second, organizational changes should this assessment, many Pakistan-based analysts be considered to bring signals intelligence into believe that Pakistan cannot or should not give a single organization and to improve coordina- up the insurgency strategy in the region as this tion and sharing of intelligence brought in by is seen as the only current point of leverage various intelligence branches. In August 2001, that Pakistan has regarding India. There was, the Indian government fi nally seemed to be however, vigorous debate over the implications moving toward adopting the committee’s rec- of Kargil at the time although some of this ommendations on this score, announcing the has been stifl ed by the ascent of the military creation, beginning in the next fi scal year, of a to a position of formal control of the govern- centralized defense intelligence agency. Lastly ment. Indeed, the failure of Indian aircraft to in the intelligence fi eld, the committee recom- cross the line of control and India’s restraint mended better “red teaming” and analysis of in moving forces outside of Kashmir may have intentions of possible adversaries, noting as a reinforced the view in Pakistan that a strat- rationale for the recommendation that Paki- egy of engaging in low-intensity confl ict under stan’s actions in Kargil were not rational.65 its nuclear umbrella was a successful one mili- In terms of nuclear lessons, the committee tarily. Finally, however, Kargil and the coup basically commended India’s decision, taken afterwards resulted in international isolation over years by various prime ministers, to slow- for Pakistan, particularly from Washington. ly acquire a nuclear weapons capability. In Almost all interlocutors in Pakistan noted the order to strengthen deterrence and further need to re-engage with the United States and to point the fi nger at Pakistan, the committee rec- ensure that Washington does not “tilt” too far ommended the publication of a white paper toward New Delhi. This isolation ended in the detailing India’s nuclear developments, show- fall of 2001 as a result of Washington’s need for ing its broad base of support and public Islamabad’s cooperation in the war in Afghani- rationale. In addition, the white paper would stan, but the risk of again becoming a pariah describe Pakistan’s continuing nuclear pro- remains if Pakistan attempts another Kargil. gram and the sources of its technology.66 Given the political situation in Pakistan Issues since the 1999 events in Kargil, no public The nuclear dimension of the Kargil confl ict report has been issued on the crisis and obvi- was threefold. First, the open possession of

72 Nuclear Stability in South Asia nuclear weapons by both sides did not deter ons use obviously continues to worry both confl ict, at least low-intensity confl ict. In fact, sides. many Indian analysts posit that Pakistan attempted the operation under the cover its Cross-Crisis Patterns and Issues nuclear umbrella, assuming that nuclear deter- Two major themes emerge from the review of rence would keep India from escalating to a past wars and crises in terms of causes that broader conventional war.68 Indian military could signifi cantly infl uence future stability offi cials have noted that the Indian air force in the region. The 1971 war was caused by did not cross the line of control because they a combination of domestic politics (the result deemed that to be one of Pakistan’s “red lines” of Pakistan’s geographic division into East that might cause it to escalate to the nuclear Pakistan and West Pakistan at that time), level.69 In support of this notion, the BJP often the Pakistani government’s unwillingness to called for hot pursuit of insurgents across the resolve the political issue peacefully and in line of control when it was in opposition, but accordance with democratic principles, and, a the BJP-led governing coalition resisted such mix, on India’s part, of opportunism and fear of temptations. This could be the moderating refugees from East Pakistan. For all other wars infl uence of being in power or the deterrent and crises, one of two factors contributed to effect of Pakistan’s weaponized nuclear capa- the cause of the confl ict. The fi rst, an obvious bilities. one, is the unresolved Kashmir question. It Second, the involvement of third parties has been central in 1947, 1965, 1984 (Siachen), to calm the crisis differed from previous cri- 1990, and Kargil in 1999, and it was a point ses. At the outset of the Kargil crisis, Pakistan of vulnerability that worried Indian planners made statements about the possibility for the during Brasstacks. Again, while not the sole crisis to escalate to conventional and then cause of all crises or wars, it remains a point of nuclear war in hopes of bringing about out- contention that has international and domestic side intervention to freeze the situation while political implications. Other situations like Islamabad still held territory in Indian Kash- Siachen help to bolster domestic saliency in mir. At the same time, in keeping with past both countries and keep it at the center of practice, India rebuffed outside efforts to inter- the political debate. Kashmir’s geopolitical vene. As the confl ict went on and it proved importance to both India and Pakistan will to be diffi cult to oust the insurgents, India not change even in the medium term, and brought in the United States. Seeing that it was its importance to the political legitimacy of winning the military war and the propagan- each state will remain an issue. Without at da war, India at that point was eager to have least the beginning of a process to resolve the Washington pressure Islamabad and did so by issue, involving both bilateral and unilateral hinting at escalation of its own, which may adjustments to national goals and expectations, have raised fears of a broader war. Finally, cer- long-term stability between the two is tain Indian and Pakistani reports indicate that impossible. both sides loaded warheads on missiles during The second major theme underlying the the crisis.70 This is impossible to verify with causes of wars and crises is the constantly open sources, but the specter of nuclear weap- changing military equation, on both the con-

Stability Issues from Past Wars and Crises 73 ventional and nuclear sides, that has led to worry military planners on both sides. Now perceptions of vulnerability and opportunity that the nuclear capabilities of each side are that in turn have triggered actions and reac- overt, the possibility for additional perceived tions. In 1965, Pakistan believed that India’s changes in the strategic balance will be rel- domestic weakness opened a window of oppor- atively greater as development, procurement, tunity to infl uence the Kashmir situation. It and deployment of delivery systems, warheads, also believed that that window would close and associated infrastructure will more often soon because of India’s defense buildup. Again, be the subject of public announcements and the preemption scares in the early 1980s were debate or be more easily subject to discovery by triggered by either Pakistani concern about intelligence capabilities. vulnerability at that time or Indian percep- tion that it may have a window of opportunity Patterns of Escalation to strike Pakistani facilities before Islamabad’s One prominent pattern of escalation across sev- nuclear infrastructure became too large or too eral wars and crises is that of both sides moving dispersed. In the 1987 Brasstacks crisis, Indian ground forces to areas near the border where the offi cials might have believed that they should other has geographic vulnerabilities. The moves seek to “solve” the Pakistan issue before Islam- were both defensive, because of perceptions of abad weaponized its nuclear capabilities. Also their own vulnerabilities, and offensive, because in 1987, Pakistan believed that India might each side also knew of the other’s geographic vul- actually team up with the Soviet Union for nerabilities. This was obvious in 1947, 1987, and an attack on its nuclear facilities. The 1990 1990. Even in 1965 when the war was initially war, while triggered by domestic turmoil in focused in Kashmir, Pakistan moved armored Kashmir, may have been the fi rst instance forces in from the south, striking at the nar- of Pakistan attempting to implement a low- row area through which India supplied its forces intensity confl ict strategy against India under in Kashmir. In response, India invaded farther the cover of its newly weaponized nuclear south with its armored forces, driving toward capabilities. Similarly, there has been much Lahore and threatening to cut Pakistan in two at speculation, particularly in India, that Kargil its narrowest point. These geographic realities 1999 represented yet another attempt to put are not going to change, so forward deployment this strategy into action, again after a perceived remains a likely move for each side during any change in the strategic equation due to the mov- future crisis – threatening the other’s vulnerable ing of Pakistan’s nuclear capabilities into the points and moving its own forces into place to open with the 1998 tests. defend against a possible deep ground attack. In sum, both of these issues (Kashmir Another pattern of escalation involved and the military balance) can be expected bringing in outside powers in an attempt to to continue to trigger crises and possibly change the balance of forces and prevail in the wars through 2015. Kashmir is unlikely to be confrontation. This tactic was used by Paki- resolved any time soon, and the continuing stan in 1965 with China, in 1971 with China shifting of the strategic equation will create and the United States, and possibly in 1990 perceptions of strategic opportunities or vul- through the use of nuclear signaling to bring nerabilities that will tempt politicians and in the United States. Pakistan also appealed

74 Nuclear Stability in South Asia for external intervention by the United States nent’s military moves. Throughout most of the during the preemption scares to deter attacks crises, intelligence information appears to have against its facilities. In 1999, any hopes by Pak- been plentiful but not necessarily of the highest istan that engaging the United States would quality. Because of technological limitations, allow it to prevail were crushed by the messag- intelligence was often derived from human es delivered by General Zinni and eventually sources and some signals interceptions. Nei- President Clinton. For the most part, India, ther state to date has imagery satellites, and being the dominant power, has sought external airborne reconnaissance is limited by both intervention less frequently, except in response technology and the other side’s air defenses to to Pakistan’s bringing in outside help. External shallow looks into the other’s territory. Intelli- powers will play a role and could potentially gence appears to have been most accurate when be brought in by one side or the other for the it involved identifying large ground maneuver purposes of deterrence, escalation or, as will be units, although for a time in 1987, India lost discussed below, de-escalation. track of a multi-division Pakistani force. This type of intelligence failure has not been unusu- Crisis Management, al (India again failed to detect infi ltrators into Communication, and De-escalation Kargil in 1999) and can lead to strategic sur- In almost every war and crisis, India and Paki- prise that in turn could lead to precipitous stan eventually negotiated directly to end the actions such as escalation. confl ict, at times employing the good offi ces of Regardless of the presence or quality of a superpower. In fact, during most of the crises information available, one consistent pattern and wars, the two sides communicated through throughout previous wars and crises was public statements, the media, hotlines, or meet- misinterpretation of that intelligence. Almost ings of senior government offi cials. In 1965, it without fail, analysts on each side would does not appear that extensive communication ascribe the most negative motives to the other, occurred during the fi ghting. In 1971, as noted resulting in a worst-case interpretation of each above, international intervention was critical and every piece of available intelligence. This to open communications because both sides pattern of behavior does not bode well for refused to communicate or negotiate directly. future crises when, in addition to concerns After 1971, communication does not appear to about large ground units and conventional air- have been a problem. With the exception of the craft located at known airbases, intelligence one period during Brasstacks when both sides gathering and analysis will be focused on appeared to have chosen not to communicate, small, highly mobile, and diffi cult to detect mis- it appears that channels existed and were used, sile systems. although not always with the intent of de-esca- As noted above, at times outside actors lating or solving the issue at hand. The physical were invited in for the purposes of tipping the ability to communicate does not appear to have balance of forces to one side or the other. How- ever been a problem. ever, more often, states and at times the UN Not all communication during the crises intervened for purposes of calming the situ- was verbal. Much occurred through each side’s ation. The success of outside intervention in observations and interpretations of the oppo- damping down crises or ending wars has been

Stability Issues from Past Wars and Crises 75 mixed. In 1965, the fi ghting ground to a halt, side powers can be refi ned so that it can best leading both sides to accept a cease-fi re. In serve to prevent crises or de-escalate ones that 1971, outside intervention was critical for end- already have begun is an area for exploration, ing the war since neither side was willing particularly considering the types of crises that to communicate, let alone negotiate. Both the could occur in the future involving signifi cant- United States and the Soviet Union indicated a ly more advanced nuclear arsenals. willingness to become involved in the preemp- tion scares to provide support for their clients. Role of Nuclear Weapons While Brasstacks was solved directly by India The wars and crises between Pakistan and and Pakistan without outside intervention, in India until the run-up to the 1990 confl ict 1990 the role of the United States was critical all had some infl uence on the issue of in ending the fi ghting. Washington provided nuclear weapons. Each war and crisis led information through attaché reports on the to consideration and reconsideration of the forces near the border, and more importantly nuclear option on New Delhi and Islamabad. it warned both sides, but particularly Paki- In Pakistan’s case, its loss in the 1971 war stan, against escalating or continuing to push led directly to the start of its nuclear weapons the other side. In this case, the United States program. India moved more steadily towards essentially withdrew its support from Paki- exercising the nuclear weapons option, but stan, letting Islamabad know that it was on its each war and crisis renewed the debate and own if it continued fi ghting. In the future, the saw India’s scientists accelerate their work in role for the United States could become more preparation for the political signal to cross the extensive, for example if the U.S. were to devel- test line in 1974 and again in 1998. As op a boost-phase missile interception capability noted in the summary of the 1990 crisis, this that could intercept missiles launched by India crisis may indeed have been the fi rst actual or Pakistan. nuclear crisis between India and Pakistan. For future crises, therefore, the issues All this may have meant was that both sides in this area are many. First, the quality of had weaponized their nuclear capabilities. communication between India and Pakistan However, it is possible that Pakistan engaged matters a great deal and should be signifi cant- in nuclear threats and signaling to both India ly improved. Second, intelligence gathering and the United States in that crisis, although should be enhanced on both sides, with mul- again if it did, the signals were tremendously tiple source types enhancing the picture with opaque. The Pakistani leadership may have which analysts and decision makers have to decided in the late 1980s to engage in a work. However, regardless of the quality and low-intensity confl ict strategy against India quantity of information, fl awed interpretation because it believed that its nuclear weapons will continue to be an issue, and one not easily would deter India from escalating. It is highly resolved. Finally, outside actors will play roles unlikely that the two sides were minutes in future crises, often as providers of informa- away from a nuclear war as some journalistic tion or good offi ces, but also as catalysts or as accounts would have it. Similarly, Kargil was interested parties preferring to see one side or likely not a nuclear crisis in the sense that the other come out on top. How the role of out- nuclear weapons were readied for use, but the

76 Nuclear Stability in South Asia risk of the crisis gradually escalating to that the driving force behind India’s acquisition of point was present as it was in 1990. nuclear weapons, but the ongoing clashes with However, both India and Pakistan have Pakistan have sparked debate on the issue and shown a keen awareness of the other side’s have often been used as a post-hoc justifi cation nuclear capabilities, in development or actual, for the program. Overall in terms of relations and have adopted strategies and adjusted tac- with Pakistan, the various wars and crises tics accordingly. Both however were reacting to have reinforced India’s belief that Pakistani covert nuclear capabilities and not to weapon- intentions remain malign and that Islamabad’s ized capabilities that may be activated in crises, actions in Kashmir are tied directly to its con- put on alert, or otherwise be visible to intel- cerns about Pakistan’s legitimacy as a state. ligence. For example, mobile launchers may be Moreover, Pakistan retains the capability to moved out of casernes and dispersed to launch- initiate crises in Kashmir. The crises in 1990 ing or hide locations. If either side ever develops and 1999 have also demonstrated to India that a sea-based nuclear weapons capability, sub- low-intensity confl ict is an ongoing problem marines or surface ships that serve as launch and that its conventional superiority and nucle- platforms may be sortied from naval bases. ar weapons have so far been insuffi cient to These actions may be part of plans to use deter the use of this type of strategy by Pak- the weapons, defensive moves intended to pro- istan. However, India does not believe that tect these assets, signals of resolve, signals this is an insurmountable problem and believes of escalatory intent designed to elicit outside that this vulnerability can be removed through intervention, or some combination of all of changes in military equipment and tactics and these. These considerations will be signifi cant changes in crisis management and decision- in future crises, and will pose a stern challenge making structures and processes in India. for both sides to learn how to react but not over- Finally, India has learned several lessons about react to weaponized capabilities that are now outside actors and their roles in the ongoing openly in the arsenals of each side. India-Pakistan relationship. First, its belief about the threat China represents to India’s Lessons Learned security – either directly or through Pakistan Finally, India and Pakistan have learned as a proxy – has fl uctuated over time. Second, numerous strategic and tactical lessons from India believes that Pakistan, as the weaker of each of their wars and crises. For India, the the two states, needs outside intervention more lessons related to conventional war with Paki- than India does in times of crisis. Finally, while stan are twofold. India can defeat Pakistan in a the United States has been a fi ckle partner over conventional war, but it also remains vulnera- time, India has at times found it useful to have ble to Pakistani conventional forces because of an outside patron either to provide military a combination of geographic realities (in both technology or to face down a threat from anoth- Kashmir and the Punjab) and Pakistan’s ability er great power. to exploit domestic unrest in India for its own Pakistan’s lessons, not surprisingly, are ends. In terms of developing a nuclear weap- very different from those of India. On the ons capability, it is not clear that any of the conventional military side, Pakistan has wars or crises with Pakistan necessarily were increasingly seen that India is both capable and

Stability Issues from Past Wars and Crises 77 willing to use its conventional forces to defeat them once they have begun. After all, most of Pakistan in limited wars. Pakistan also now the wars and crises between India and Paki- knows that India has the capability, and pos- stan were intentional, that is one side or the sibly even the desire, to use its conventional other was attempting to use force, or the threat superiority to defeat Pakistan in a general war of it, to achieve political aims. Only the 1987 and impose upon it broader settlement terms Brasstacks crisis (depending on one’s interpre- regarding Kashmir and other issues in dispute. tation of India’s initial motives in holding the Directly because of this concern about not just exercise) appears to have been a case where losing limited wars but of having Pakistan’s neither side had any intention of threatening national survival threatened by India, Islam- or using force for political gains. abad decided that it had to develop a nuclear Looking into the future, it is necessary to weapons capability. The development of this think through how these various patterns of capability led to a belief, reinforced by the con- action and interaction will play out in crises fl icts of 1990 and 1999, that Pakistan may be or wars and how they may produce different able to pursue a new strategy in its competition results because of the different nuclear force with India (particularly over the fate of Kash- strategies, doctrines, and force structures that mir), one in which it uses low-intensity confl ict India and Pakistan might adopt over time. This while counting on its nuclear capabilities to is the task of chapter 4, attempting to apply deter India from escalating to the conventional this region-specifi c information to future crisis or nuclear levels. Finally, Pakistan has learned models in an effort to identify the most danger- that outside intervention by superpowers or ous situations and points where escalation may great powers remains a requirement for suc- occur. cess in its competition with India. While it has counted on China for WMD-related tech- nology, Islamabad has learned that China has limited capability to threaten India directly and effectively. Pakistan continues to believe that it is essential to court the United States because of its status globally and in South Asia, and after 1999, wishes to re-engage with Wash- ington and keep it from tilting too far in India’s favor. As can be seen above, these lessons are often diametrically opposed and not necessar- ily conducive to stability, particularly in future crises. Most of the lessons, as evidenced most recently by the offi cial Kargil report commis- sioned by the Indian government, concentrated (not surprisingly) on how to do better in the next possible war or crisis rather than how to avoid crises in the fi rst place or de-escalate

78 Nuclear Stability in South Asia Endnotes for Chapter Three

1 Sumit Ganguly, “Deterrence Failure Revisited: The Indo-Pakistani War of 1965,” The Journal of Strategic Studies 13, no. 4 (December 1990): 77-93. 2 George Perkovich, India’s Nuclear Bomb: The Impact on Global Proliferation (Berkeley: University of California Press, 1999), 46. 3 For a complete treatment of the roots of confl ict between India and Pakistan, see Sumit Ganguly, The Origins of War in South Asia: Indo-Pakistan Confl icts Since 1947 (Boulder, Colo.: Westview Press, 1996). 4 Gowher Rizvi, “India, Pakistan, and the Kashmir Problem, 1947-1972,” in Raju G.C. Thomas, ed., Perspectives on Kashmir: The Roots of Confl ict in South Asia (Boulder, Colo.: Westview Press, 1992), 50. 5 Z.A. Bhutto, The Myth of Independence (London: Oxford University Press, 1969), 180. 6 Ganguly, “Deterrence Failure Revisited.” 7 P.R. Chari, Pervaiz Iqbal Cheema, and Stephen P. Cohen, “The Compound Crisis of 1990: Perceptions, Politics, and Insecurity,” pre-publication manuscript, version 2.14, July 12, 2000, 63. 8 Perkovich, 108. 9 Ibid., 106. 10 Chari, Cheema, Cohen, 64. 11 Ganguly, “Deterrence Failure Revisited,” 89. 12 Ibid., 90. 13 Perkovich, 109. 14 United Nations Security Council resolution 211, September 20, 1965, available at http://www.un.org/documents/sc/res/1965/s65r211e.pdf 15 Tashkent Declaration, January 10, 1966, available at http://www.stimson.org/cbm/sa/tashkent.htm. 16 Perkovich, 110. 17 Ibid., 112. 18 Ganguly, “Deterrence Failure Revisited,” 77, 84. 19 Devin T. Hagerty, The Consequences of Nuclear Proliferation: Lessons from South Asia (Cambridge, Mass.: MIT Press, 1998), 71. 20 Perkovich, 165. 21 Ibid., 162. 22 Ibid., 163. 23 Richard Sisson and Leo E. Rose, War and Secession: Pakistan, India, and the Creation of Bangladesh (Berkeley: University of California Press, 1990), 279. 24 Chari, Cheema, Cohen, 65. 25 The text of the pact is available at http://www.stimson.org/cbm/sa/simla.htm. 26 Perkovich, 165. 27 Ibid., 171. 28 Sisson and Rose, 221. 29 Perkovich, 178. 30 Sisson and Rose, 274-275, 279. 31 Milton R. Benjamin, “India Said to Eye Raid on Pakistani A-Plants,” Washington Post, December 20, 1982, p. A1; Reuters, “Concern Voiced on Atom Plants,” July 3, 1983; Philip Taubman, “Worsening India-Pakistan Ties Worry U.S.,” New York Times, September 15, 1984, p. 2. 32 William Claiborne, “U.S. Offi cial Holds Talks In India on Aid Row: Arms Supply to Pakistan at Issue,” Washington Post, October 22, 1984, p. A20. Kanti Bajpai, P.R. Chari, Pervaiz Iqbal Cheema, Stephen P. Cohen,

Stability Issues from Past Wars and Crises 79 and Sumit Ganguly, Brasstacks and Beyond: Preception and Management of Crisis in South Asia (New Delhi: Manohar Publishers and Distributors, 1995), 9. 33 Bajpai et al., 56. 34 Ibid., 148. 35 Ibid., 10. 36 Robert G. Wirsing, India, Pakistan, and the Kashmir Dispute (New York: St. Martin’s Press, 1994), 197, footnote 21. 37 Wirsing, 78-80. 38 Ibid., 77. 39 Ibid., 200-201. 40 Gardner W. Heaton, “Prospects for Peace in the Indo-Pakistani High-Altitude Border War,” MALD thesis at The Fletcher School of Law and Diplomacy, April 2001, 4. 41 The best and most detailed account of the entire Brasstacks crisis is found in Bajpai et al. 42 Text of the agreement was reproduced in “Indo-Pak Accord to De-Escalate Border Tension,” The Statesman, February 5, 1987. 43 The A. Q. Khan interview reported by investigative journalist Seymour Hersch has been shown by the more in-depth studies of Bajpai et al. and Hagerty not to have been related to the crisis or its resolution. 44 Chari, Cheema, and Cohen, 49. 45 Bajpai et al., 55-56. 46 Hagerty, 117. 47 Chari, Cheema, and Cohen, 106. 48 C. Uday Bashkar, “The May 1990 Nuclear ‘Crisis:’ An Indian Perspective,” Studies in Confl ict and Terrorism, vol. 20 (New Delhi: Institute for Defense and Security Analysis, 1997), 325. 49 Perkovich, 307. 50 Chari, Cheema, and Cohen, 115. 51 Ganguly, The Origins of War, 179. 52 Michael Krepon and Mishi Faruqee, “Confl ict Prevention and Confi dence-Building Measures in South Asia: The 1990 Crisis,” Henry L. Stimson Center Occasional Paper no. 17, April 1994, 114. 53 Seymour M. Hersh, “On the Nuclear Edge,” New Yorker, March 29, 1993: 56-73. William E. Burrows and Robert Windrem, Critical Mass: The Dangerous Race for Superweapons in a Fragmenting World (New York: Simon and Schuster, 1994). 54 Mitchell Reiss, Bridled Ambition: Why Countries Constrain Their Nuclear Capabilities (Washington, D.C.: Woodrow Wilson Press, 1995), 188. 55 Ibid., 190. 56 Bashkar, 320. 57 Chari, Cheema, and Cohen, 170. 58 Bashkar, 324. 59 Bashkar, 326. 60 Chari, Cheema, and Cohen, 217-218. 61 Ibid. 62 Thomas W. Lippman, “India Hinted at Attack in Pakistan; U.S. Acts to Ease Tension in Kashmir,” Washington Post June 27, 1999, p. A26. 63 Bradley Graham and Nathan Abse, “U.S. Says Pakistan Will Withdraw,” Washington Post, 5 July 1999, p. A15. 64 Kargil Review Committee, From Surprise to Reckoning. The Kargil Review Committee Report (New Delhi: Sage Publications, 1999), 259.

80 Nuclear Stability in South Asia 65 Kargil Review Committee, 255. 66 Ibid., 259-260. 67 Author’s interviews, Islamabad, Pakistan, October 2000. 68 The Kargil Review Committee, 183. 69 Author’s interview with senior Indian air force offi cial, September 22, 2000. 70 P.R. Chari, “Nuclear Restraint and Risk Reduction in South Asia,” presentation at the Carnegie Endowment for International Peace, February 16, 2001. In this presentation, Chari cites Indian journalist Raj Chengappa on this issue.

Stability Issues from Past Wars and Crises 81 CHAPTER FOUR

Future Crises and Escalation Potential

This chapter builds on the analysis contained in chapter 2 about potential future doctrines and force structures as well as the analyses and compari- sons set forth in chapter 3 of past wars and crises. We now posit a spectrum of future crises and project how escalation to the nuclear level could occur in those situations. This is done to gain a better understanding of how specifi c force structure and doctrine choices on the part of India and Pakistan could interact with likely crisis situations to produce destabilizing results, namely escalation to the nuclear level. This then provides a starting point for devel- oping remedial measures that could decrease the probability of escalation – stabilizing measures – in chapter 5.

82 Nuclear Stability in South Asia Deterrence and other Failures in Past nuclear forces being at least theoretically Wars and Crises part of the equation since 1974), Payne’s anal- As noted in chapter 3, India and Pakistan have ysis addresses similar outcomes. In short, had numerous failures along the confl ict spec- the potential for deterrence failure remains trum that have either led to wars, caused those substantial even if both sides have nuclear wars to be longer and bloodier than other- weapons if we base our analysis of the condi- wise would have been the case, and/or led to tions for success or failure on the conditions unsatisfactory outcomes that paved the way proposed by Payne. As seen in the survey of for future clashes. All of these failures have crises and wars in chapter 3, the strategic com- implications for the future as it is not clear that petition between India and Pakistan matches either side has learned suffi ciently from them virtually perfectly with these prescriptions to avoid repeating them. As suggested in chap- for deterrence failure. Each side has persistent- ter 3, the lessons learned from past confl icts ly misinterpreted the other’s political signals. by each side focus more on how to prevail in Both have also misread each other’s level of future confl icts rather than how to avoid or resolve. India and Pakistan often have had damp down the levels of violence in future fl awed assessments of the existing situation or fl are-ups. anticipated shifts in the military balance. They Past deterrence failures are the most cru- have either exaggerated their own capabilities cial starting point in understanding how the or underestimated those of the opponent. A two sides have found themselves at war. Most pessimistic outlook for the future, particularly important, the deep-seated hostilities between on the part of Pakistan, has led it to foresee India and Pakistan have underscored the and fear the closing of a crucial window of inherent fragility of strategies based principally opportunity. In turn this has precipitated risky on deterrence. Writing about nuclear deter- behavior to halt or reverse the perceived trend. rence, Keith Payne concludes: These elements have in their turn caused the Deterrence policies cannot succeed by deterrence mechanism to fail, leading to war. design when basic ingredients as defi ned in The clearest example of conventional deter- deterrence theory are missing…. Even the rence failure was 1965, and the 1999 Kargil careful manipulation of nuclear threats… episode provides evidence that both conven- cannot “ensure” deterrence policies against tional superiority and possession of nuclear the possibility of misperception when the weapons are not suffi cient to deter low-inten- deterrer’s policy is poorly informed, when sity confl ict, providing support for Glenn the contestants are operating outside the Snyder’s “stability-instability” paradox where- deterrence theory assumption of well- by nuclear weapons eliminate the possibility informed rationality, or when the challenger for all-out wars between opponents but raise is simply so highly motivated that it will the possibility of low-intensity confl ict or con- rationally and knowingly accept tremen- ventional confl ict in non-strategic areas. dous risks in pursuit of its goal.1 On those occasions when either deterrence has failed or both sides have become engaged Although the cases examined in chapter in a crisis short of war, both have demonstrat- 3 had as their basis conventional forces (with ed a persistent inability – either deliberate or

Future Crises and Escalation Potential 83 inadvertent – to manage the resulting crisis or measures adopted by the two sides. However, confl ict. Crisis management has internal and the intentional misuse of these channels of external dimensions. First, a complex process communications has at times obviated crisis of internal deliberations must proceed to deter- management efforts and could do so in future mine options and responses. In the past, this crises. For example, the leadership may be com- has often rested on individual leaders, who pelled to send a stronger message – that could sometimes failed to heed prudent advice or signal a willingness to escalate – than warrant- simply miscalculated. The results have often ed for domestic consumption and propaganda been disastrous (such as Nehru’s “forward pol- purposes. On a related point, the leadership icy” that triggered a massive Chinese response could fail – intentionally or inadvertently – to in 1962). More recently, both sides have make use of critical channels of communica- institutionalized the process to prevent the tions such as the hotline connecting the top overpersonalization of the decision-making leaderships. As Stephen Cohen has argued on mechanism during crises. For example, India is a number of occasions, many CBMs in South moving ahead with the Kargil-related changes Asia have evolved into “confi dence-bursting”2 to establish a the position of chief of defense measures because of abuse or misuse. Past staff, a centralized defense intelligence agency, failures have then led to expectations that and a full-time national security advisor – all the other side will not use CBMs or certain steps to provide better information and advice channels of communication designed for crisis to decision makers. However, these organiza- management properly and therefore each side tional changes and processes remain untested avoids their use altogether. and underdeveloped and in fact may produce In the past when crises have threatened similar miscalculations or as poor judgment as to spiral out of control, both sides have often would a single individual. underestimated or failed to control the rate of Second, how one side interprets the signals escalation. Indeed, the historical record tends and responses of the other is integral to man- to support the case that each side has taken aging a crisis. Reading the signals correctly actions or signaled intentions that have actu- enables one party to avoid certain thresholds ally accelerated escalation. The persistent that might lead to escalation. Unfortunately, failure to control escalation stems from the in the annals of confl ict and crisis, adversar- classic security dilemma: the action of one ies have often lacked the proper understanding side provokes a corresponding or even dispro- of each other’s intentions. Even when there portionate reaction that in turn heightens the are exchanges of information or signals, parties crisis. Both sides are locked in a vicious cycle. involved have misinterpreted their content. As noted in chapter 3, defense planners in Moreover, the history of modern South Asia India and Pakistan tend to base their decisions powerfully demonstrates that the probability on worst-case and exaggerated assumptions. for breakdown in communications remains For example, military exercises or maneuvers dangerously high. In some cases, the lack of are often perceived as a disguise or a precur- means to communicate or technical problems sor to a preemptive attack. This deep distrust has worsened the crisis. This has become less often triggers unwarranted fears and anxiet- of an issue given the various communications ies that have in turn forced the other side to

84 Nuclear Stability in South Asia engage in drastic defensive reactions that are fall of 2001 as Washington needed Islamabad viewed as offensive by the opponent. These to prosecute the war in Afghanistan against fl awed assumptions distort calculations of the the al-Qaeda terrorist network and the ruling probability for controlling or prevailing in the Taliban. Such broad realignments in the future crisis or confl ict. This has led to risk-taking will continue to be critical indicators of how behaviors that have proved disastrous. one side will react to crisis. Indeed, in the wake When prolonged crises or confl icts reaches of increased U.S. involvement in South Asia, a stalemate, both sides have failed politically to the region could once again become the nexus de-escalate and resolve the situation. Domes- of geostrategic competition between China, tic political pressures have often constrained Russia, and the United States although some the deliberations and actions of each govern- might argue that the latest heavy U.S. presence ment. These pressures have been evident both will dampen that tendency and promote in democratic India and even in Pakistan at stability. Most worrisome, fl awed assumptions times when the country was ruled by the mili- about external assistance – that may or tary. New Delhi and Islamabad often fear the may not be forthcoming – could precipitate domestic backlash that could result from public risky behavior. Should the outside power perceptions of government weakness or lack of fail to respond as anticipated, the risk taker resolve. This internal consideration has made could quickly fi nd itself in trouble. Clearly, backing down, concessions, or negotiations on future intervention by a great power or the realistic terms very diffi cult. As a result, each international community at large will affect all side has either been unwilling or unable to end aspects of the confl ict spectrum. the crisis or confl ict. As a democracy, India is These same failures are likely to resur- particularly vulnerable to domestic pressures. face in the future. Most vexing, the potential There are also exogenous factors that have involvement of nuclear weapons use could fun- affected how crises have unfolded. At the damentally alter the contours of the confl ict broadest level, geopolitics have determined spectrum. The nuclear calculus could exacer- the calculations and actions of each state. bate the failures to prevent or dampen a crisis During the Cold War, promises of support and and ultimately lead to war or possibly a nuclear intervention by the superpowers had major exchange. Some failures are common and apply implications for the outcome of crises and to interactions between all states, while others wars. The strategic alignments with outside are unique to the dynamics of the intense rival- powers have sometimes created perceptions ry between India and Pakistan. Understanding of advantage or security that enabled or how these failures apply to the emergence of encouraged one side or the other to take action. overt nuclear capabilities in South Asia will be Yet, the end of the Cold War demonstrated important for projecting future scenarios and that such support could quickly evaporate. As examining possible crisis prevention measures. Pakistan’s strategic importance receded in the The use of a “western” model has naturally early 1990s, American political and military drawn protestations from both sides that the backing waned. Islamabad was then forced to confl ict spectrum does not fi t neatly with the rely more on China for technical and political unique conditions of the competition. Some support. This was reversed yet again in the have argued that their common heritage and

Future Crises and Escalation Potential 85 mutual understanding from years of interac- and examined in terms of how they are unique- tion would enable each side to avoid the brink ly stressed in those situations. On all of of the precipice. Yet the history of the region these issues, India and Pakistan face steep clearly shows that both are and will still be sub- learning curves. In the United States and the ject and vulnerable to universal principles of Soviet Union policy makers, defense planners, confl ict. scientists, and military operators worked at Not only will these past trends and failures mastering these issues over two generations. likely persist in the future, but the addition The issues are so important that they are never of nuclear weaponry makes the avoidance of considered “solved,” and the United States is past failures during crisis an even more impor- constantly improving its technologies and pro- tant and dangerous task for both sides and the cedures in these areas. At best, India has faced international community. The following sec- these issues for a generation and Pakistan for a tion presents a series of alternative futures that little over a decade, but the nature and scope of involve all the potential elements of instability the challenges are changing signifi cantly as the examined above. two states move to weaponize their arsenals Before outlining specifi c scenarios and and then to deploy forces. attempting to identify the key variables with- The safety of a nuclear weapon, meaning in them that lead to escalation, it is important the ability to keep it from either detonating to note two additional areas of concern now inadvertently or being involved in an accident that India and Pakistan openly possess nuclear that leads to a release of radioactivity, will be weapons. The fi rst is the problem of the safety, more diffi cult to ensure in times of crisis or security, and reliability of nuclear weapons in war. First of all, in these situations, weapons the Indian and Pakistani arsenals. These are may actually be assembled. As noted in chapter serious concerns in peacetime, crisis, or war. two, in the current and some future force pos- However, they take on a heightened level of tures, particularly for India, the weapons may importance in crises or wars and may require not necessarily have the fi ssile material core different measures or procedures during those mated with the conventional explosives and times. As noted earlier, Pakistan upgraded and the rest of the bomb or warhead. This raises a changed its security procedures at the outset of question that might not be as critical in peace- the autumn 2001 war in Afghanistan, in time when weapons are disassembled: whether part in response to international pressure and there is a robust safety and arming mecha- in part as a result of its own assessment nism on the weapon ensuring that it cannot be of the threat. During crises, procedures and armed by accident. Second, because of the need technologies developed to address safety and either to disperse weapons to preserve them security concerns will be subjected to sig- against a possible fi rst strike or to ready them nifi cant strains, raising the chances of theft for use, they will be handled and moved much or accidental or unauthorized detonation or more than in peacetime, increasing the oppor- launch. Since this study is focused primarily tunity for accidents. Finally, if the weapons are on crisis situations, stability, safety, security, either brought together with delivery systems and reliability issues will be touched upon pri- or are actually mated to those missiles or air- marily in the context of future wars or crises craft, then the safety of those systems comes

86 Nuclear Stability in South Asia into question. For example, if either India or divergent doctrines on use. For Pakistan, reli- Pakistan is still using liquid-fueled missiles as ability is a much greater concern because of its nuclear weapons delivery vehicles, then the fi rst-strike doctrine and its belief that nuclear chances for accidents rise signifi cantly given weapons use may be necessary to ensure nation- the high volatility of that fuel type. al survival of an Indian conventional or nuclear Similarly, it will be more diffi cult to ensure attack. For its part, India is certain that it could the physical security of weapons (the ability prevail in either a conventional or a nuclear con- to keep them from falling into unauthorized fl ict with Pakistan, making it less worried if hands) in times of crisis or war simply because its nuclear weapons do not perform exactly as the weapons are likely to be moving around designed. However, in a future Indian crisis or outside of secure storage facilities. In addition, confl ict with China, this issue may be more criti- responsibility for their security is likely to be cal for New Delhi. handed over at various times (from the storage It is important to note that both Indian facility commander to a convoy commander and Pakistani offi cials have demonstrated pub- to a launch facility commander), opening up licly a keen awareness of and sensitivity to opportunities for miscommunication and laps- safety, security, and reliability issues. Indeed, es of responsibility. In the cases where India they have often taken offense at Western com- and Pakistan move to the high force posture mentary that the two sides would be unable to posited in chapter two, safety and security manage their nuclear capabilities. As a result, might actually be better since both forces will they have refl exively rejected unoffi cial Ameri- be deployed and more likely will have devel- can and other assistance.3 India’s NSAB report oped procedures and technologies to address placed great emphasis on these three areas as these questions during day-to-day operations priority tasks. Whether the declared objectives and exercises. If both sides remain at a low are merely rhetorical is unknown. Neverthe- force posture, they may only have to address less, both are no doubt taking some steps to these questions seriously for the fi rst time in address safety, security, and reliability issues. a crisis, whereupon peacetime measures and The relatively leisurely pace with which doc- procedures might simply prove inadequate. trine and force structure developments have The reliability of a nuclear weapon, the abil- proceeded in each country may refl ect, in part, ity to ensure that it performs as designed over the recognition that they do not yet possess time, is less of an issue for this study although all of the technologies and procedures to har- it is usually paired with safety and security in ness existing weaponry safely and securely. In U.S. discussions of nuclear weapons. It could other words, they recognize that crises and be argued that less reliable weapons, or weap- unanticipated shocks could overwhelm exist- ons where the users are unsure of their reliabil- ing measures and ongoing efforts to address ity, could actually be stabilizing. If either India safety and/or security issues. Whether they or Pakistan were uncertain that their weapons can devote suffi cient resources rapidly enough were reliable over time, they might be reluctant before the onset of another major crisis remains to take risks in a crisis that could lead to possible an open question. use. Much of this assessment, however, is situ- The second general area of concern may ation specifi c because of India’s and Pakistan’s seem obvious, but it is useful to note here since

Future Crises and Escalation Potential 87 it colors all future scenarios. This is that India ined in chapter 2 for India and Pakistan and Pakistan have policy objectives regarding demonstrate the range of the possible, and as the other that are fundamentally at odds with noted, elements of each of the three types of one another, and this is unlikely to change doctrines and force structures could exist at over the period of this study. India wants one time. With three distinct force structures to maintain security while achieving great- for each side, nine force posture combinations power status. These goals translate into the are at least theoretically possible. Combine this domination of South Asia, the pacifi cation of with various crisis or war types as seen over Pakistan, and the reincorporation of Pakistani- the past fi fty-plus years between India and and Chinese-occupied portions of Kashmir. Pakistan, and the number of potential sce- The crusading impulses of its liberal democrat- narios quickly gets beyond the bounds of the ic ideals, and its long history, giant landmass, reasonable. Therefore, this study reduces the impressive economic growth in the past decade, scenarios to three, based on the following cri- and military potential will all sustain and fuel teria. First, in order to ensure policy relevance, India’s hegemonic ambitions. the crises are within the realm of the plau- By contrast, Pakistan’s political objectives sible. As such, the force structures for each are seen by Islamabad as matters of national alternative future are timeframe sensitive. For survival. Their concerns are more structural example, domestic or foreign acquisition of sea- in nature and include territorial claims, ideo- launched nuclear weapons within the next fi ve logical disposition, and deep-rooted fears about years is implausible for either India or Pak- possible loss of autonomy.4 Having been defeat- istan. Therefore, a scenario involving those ed by India and having seen one of those capabilities is set further out into the future. defeats result in a signifi cant loss of territory Second, the scenarios all have nuclear escala- (East Pakistan), Pakistan wishes to resist Indi- tion potential based on analysis of previous an domination. As noted in chapter 1, the crises and wars. Third, each scenario highlights close linkages between national identity and distinct policy dilemmas, political problems, the unresolved Kashmir dispute have also and ambiguities in decision making for India, hardened Pakistani convictions that the reuni- Pakistan, and the international community. fi cation with the Indian-held territories is vital Most important, they encapsulate some of the to the nation’s well-being. Given how severe most destabilizing elements of each alternative Pakistan believes the consequences of failure doctrine and force structure that could push would be, it is far less comfortable with the sta- each side to escalate to nuclear use. Clearly, tus quo and more insecure than India. Hence these scenarios do not cover the full range of its future actions and reactions are likely to be what is possible. Nor does positing them mean more radical and less risk averse. that they are likely to happen or to unfold in exactly the manner described in this chapter. Future Scenarios Rather, they are meant to serve as analytical Before examining the specifi c scenarios, it is tools to help think about what the future may important briefl y to note the methodologies be like and where the most destabilizing inter- used to select and develop them. The various sections may occur between doctrines and doctrines and associated force structures exam- force structures on the one hand and future

88 Nuclear Stability in South Asia crisis situations on the other. Each scenario Buoyed by relative internal stability and begins with working assumptions and the external support, the leadership draws some proximate causes of the particular crisis. The misguided conclusions about the Kargil con- narrative then turns to the description of the fl ict. First, using a combination of regular and crisis. Each scenario concludes with an enu- irregular forces and better weapons and tac- meration and analysis of the specifi c escalation tics, Pakistan becomes convinced that it could potential that the mix of the scenario and achieve tactical and strategic surprise against doctrine/force structure elements highlight. India yet again. Second, Indian restraint dur- ing the 1999 crisis (e.g., refraining from Scenario 1 – Kashmir Yet Again crossing the line of control with its aircraft or The fi rst scenario involves the recurring Kash- on the ground in hot pursuit of infi ltrators) mir problem and is essentially a more severe reinforces perceptions that Pakistan can act on reprise of the 1999 Kargil crisis. The timeframe the low-intensity confl ict level with impunity for analysis is 2007. By that point, the inter- under the cover of its nuclear weapons. Third, national community has long since concluded Islamabad gambles that the fi rst two assump- that sanctions have done little to reverse the tions would enable it to force New Delhi to nuclear weapons programs in South Asia. Con- fi ght on Pakistani terms, limit New Delhi’s sequently, major supporters and donors such as escalation options, and provide Pakistan with Japan have followed the U.S. lead from the fall signifi cant leverage to force India either to cede of 2001 and granted broad sanctions relief to the territory or to negotiate a solution to Kash- Pakistan and India. This, combined with post- mir on Islamabad’s terms. Afghan war assistance from the United States, On the other side, India remains convinced in turn leads to a partial Pakistani economic that Kargil was a major military and political recovery and an associated rise in domestic victory and continues to consolidate its con- stability. The military leadership has lived up trol of Kashmir with little reference to popular to the promises it made in 2001, relinquished desires in the region or any of Islamabad’s its formal political authority, and returned the interests or wishes. A confi dent India does not country to nominal civilian rule in the autumn perceive any need to adjust conventional or of 2002. As time passes and because the eco- nuclear strategies to deal with future threats nomic and political consequences of the Kargil from Pakistan beyond those changes recom- crisis have faded relatively quickly, most pol- mended by the government-mandated Kargil iticians and military fi gures come to believe review committee. New Delhi believes that its that the 1999 operation was not only a tactical tactical military and organizational changes military success but that the political down- will suffi ce to guard against both strategic and sides were so minimal that it could warrant tactical surprise by Pakistan in the future, and a reprise. Washington’s political embrace of becomes confi dent that it can defeat any new, Islamabad in the fall of 2001 only bolsters the large-scale Pakistani-sponsored incursion by feeling among some in the political and mili- taking conventional military action that would tary elite that it has suffi cient political cover stop short of what it believes to be Islamabad’s from the relationship with the superpower to “red lines” that may lead to conventional and/ test India again. or nuclear escalation.

Future Crises and Escalation Potential 89 India’s relatively successful consolidation and begins planning for conventional escala- process in Kashmir triggers Pakistani fears tion. Moreover, the evidence of an unprovoked that it will soon be unable to halt or reverse Pakistani military action heightens domestic this trend either through its current levels of pressure in India not only to push out the infi l- support for the insurgency or by negotiations. trators but to punish Pakistan severely and Islamabad concludes that it must undertake resolve the Kashmir problem once and for all. military action whose type and scale at a mini- The military presents the civilian leadership mum match those of Kargil 1999 to ensure that with two broad sets of options that are not India does not achieve a fait accompli in Kash- mutually exclusive. First, Indian ground and mir. In an unexpected turn, renewed uprisings air forces could be allowed to cross the line of in Indian-controlled Kashmir provide Pakistan control with more relaxed rules of engagement. with the perfect opportunity to intervene. The This would enable them to interdict Pakistani civilian leadership hesitates. Fearing that the lines of communication, conduct hot-pursuit civilians might squander a dwindling chance operations, and target supply bases. Second, to wrest the land from India, the military plac- India could escalate horizontally by mobiliz- es tremendous pressure on the government to ing and moving armored forces forward in ensure that Pakistan takes decisive action. The the Rajasthan Desert near Lahore, threaten- military’s successful political maneuvers, cou- ing Pakistan at a vulnerable point as in 1971, pled with a groundswell of popular support for 1987, and 1990. This option avoids crossing action in Kashmir, force the civilian leader- of the line of control, which is perceived as a ship to order another military campaign across major escalatory threshold for Pakistan, but it the line of control. Regular forces, insurgents, does threaten Pakistani territorial integrity in and volunteers infi ltrate weak points across the another way. This type of move would send a line of control. They penetrate more broadly powerful, but potentially dangerous, signal of and deeply than during the 1999 Kargil crisis. Indian resolve. Moreover, it could function as Despite New Delhi’s efforts to improve intel- a strategic diversion. Either choice would rep- ligence capabilities, Pakistan’s sudden thrust resent a major step toward further escalation. surprises India again. While New Delhi belat- Decisions and Options edly plans for a response, the infi ltrators At this critical point both states are confronted entrench their positions, which further encour- with various diffi cult decisions and options. ages locally based insurgents. The Indian army The domestic political pressures on both sides and air force respond with massive counterat- are intense not to back down, to win, and in tacks using the techniques and restrained rules India’s case, to broaden the confl ict and “solve” of engagement that were successful in 1999. the Kashmir/Pakistan issue once and for all. However, this time they are unable to dislodge While outside powers may at this point attempt or drive back the enemy. to intervene to ease the crisis, history shows The Indian leadership then faces an acute that their infl uence, particularly when core policy dilemma: it must either escalate the war, issues like Kashmir are in play, is limited. or capitulate to Pakistan or negotiate on Islam- It is useful here to consider the elements abad’s terms, both unacceptable. New Delhi of escalation inherent in the scenario and how refuses to consider submission or negotiation they would play out with the assumed doctrine

90 Nuclear Stability in South Asia and force structure variables to see exactly take either of those conventional options, the under what circumstances the interaction effects may trigger other concerns about the could push either side to consider escalation to survivability of Pakistan’s nuclear forces that the nuclear level. In terms of doctrine and force force Islamabad to consider nuclear use. The structure issues, this scenario assumes that problem is twofold. First, India cannot know both sides have progressed modestly in nuclear clearly exactly where Pakistani red lines are. capabilities. By 2007, India and Pakistan will Therefore it may mistakenly choose a con- have likely accumulated enough nuclear weap- ventional option that it believes is safe but ons to fall somewhere between the low and is actually pushing Pakistan to the brink of medium ranges discussed in chapter 2, tend- nuclear use. Second, the small size of Paki- ing toward the low option. Both sides have a stan’s nuclear force structure in this scenario limited inventory of nuclear weapons and at makes it vulnerable to Indian conventional least India still relies heavily on aircraft as its preemption, which Pakistan may worry is also primary delivery platform. Each has a handful a military objective of Indian moves across the of short- to medium-range missiles that can be line of control or the international border far- armed with nuclear warheads. Given that Paki- ther south. Should New Delhi choose to take stan enjoys tested warhead designs, it is likely the war across the line of control, Indian con- to hold a numerical superiority over India in ventional air strikes could decimate Pakistani the ballistic missile area. Both sides have yet to air defenses and leave nuclear facilities and develop fully their command and control archi- weapons vulnerable to further air or ground tecture. Indeed, neither side feels compelled attacks. Also, a conventional ground invasion within the fi ve-year timeframe to incorporate could overrun nuclear facilities and forces or sophisticated systems to harness the small cut the north-south lines of communications, number of weapons that it has available. Nucle- which could prevent the Pakistani military ar forces remain for the most part undeployed, from mating warheads with the delivery sys- and each side must assemble warheads or tems or moving forces to predetermined launch bombs, move them to locations to meet up locations. with delivery vehicles, and then disperse the Regardless of the size of the Indian nuclear launchers or aircraft to predetermined launch- force, because of the small Pakistani force and ing points or designated airfi elds. the potential for conventional preemption, Paki- The escalation potential in this scenario stan is trapped in the classic “use it or lose it” is particularly high when both sides still dilemma. It can ill afford to refrain from nuclear have small, relatively unsophisticated nuclear use and risk a disarming conventional or nucle- arsenals. India’s two conventional escalation ar fi rst strike. Yet, nuclear use against India’s options (crossing the line of control or invad- soft targets would no doubt invite national anni- ing farther south) both may be considered by hilation. The small force structure and its vul- Pakistan as threats to its national survival or nerability to preemptive strikes would likely autonomy and therefore as warranting a nucle- force Pakistan to disperse and ready its arsenal ar response. Even if they are not considered during the early stages of the crisis if not before in and of themselves as crossing a red line for its onset. This very response could prove esca- Pakistan, once India begins actually to under- latory. Should India detect the movement of

Future Crises and Escalation Potential 91 the force, it could conclude that Pakistan was the international border in the vicinity of preparing to conduct a fi rst strike. This could Lahore. The problem with this option is that in turn spur the Indian leadership to consider the small number of weapons in Pakistan’s preparations for a conventional preemptive arsenal means that this is likely to be ineffec- strike on those assets, if they could be accurately tive. Moreover, it would be a useful option only targeted, or a punishing nuclear second strike. if India were attacking conventionally in the Should Pakistan detect that such deliberations south. Nuclear weapons would have little util- were occurring in New Delhi or discover the ity in Kashmir against Indian hot pursuit or actual movement of warheads or bombs to deliv- against air strikes across the line of control. ery platform locations, Islamabad may become India would also consider such a warfi ghting less constrained and more prepared to resort to use to be highly provocative. Furthermore, this the nuclear option. approach would use most of Pakistan’s ready Pakistan’s options under the belief that it weapons and invite an Indian second strike. may be in a “use it or lose it” situation are The fi nal, and most likely, option would three. First it could choose to take a risk (of los- be for Islamabad to use most of its force to ing assets before it could use them) and conduct strike New Delhi and other identifi ed nuclear a nuclear strike of very limited scope designed and military command nodes, hoping to wreak to demonstrate resolve to India and signal to enough damage and chaos to stop any Indian the international community that the situation conventional attack in its place and to cripple has gotten out of hand and requires interven- India’s ability to conduct a nuclear counter- tion. It would then reserve the remainder of its attack. In any of these situations, New Delhi arsenal for countervalue targets if India was would be under tremendous pressure to strike not dissuaded or chose to respond with its own back and, should the forces be available, to nuclear weapons. Whether the signal would annihilate Pakistan. be read correctly in either New Delhi or other Escalation Problems world capitals is an open question, and it could The escalation problems in this scenario then, provoke huge political and military pressure assuming the low force postures and associ- for India to strike back before the international ated doctrines, are six. First is the problem community fi gures out what its response would that Pakistan may initially believe that nuclear or should be. Politically, an Indian government deterrence will work as a shield under which may simply not be able to abide having its it can conduct low-intensity operations that forces under nuclear attack without respond- would force India to come to terms over Kash- ing in kind, and militarily, given intelligence mir. Second is the uncertainty on the part of shortfalls, India may not know that this was India as to what constitutes a Pakistani red line “only” a signal or demonstration shot. In many in terms of Indian use of conventional force to respects, this is a high-risk proposition for respond to incursions in Kashmir. Third, the Islamabad. small size of Pakistan’s force puts Islamabad Second, Pakistan could attempt to use its in a “use or lose” situation, almost regardless limited force in a tactical warfi ghting mode of India’s actions, if Pakistan believes the to strike Indian conventional capabilities, par- worst about New Delhi. Fourth, India’s con- ticularly ground units if Indian forces cross ventional escalation options may not in and

92 Nuclear Stability in South Asia Kashmir Crisis Summary Main Features of Crisis · Both Pakistan and India have the low force postures posited in chapter 2 · Pakistan attempts another Kargil and achieves tactical surprise · India unable to prevail using Kargil 1999 tactics and self-imposed limitations · India chooses to escalate conventionally · India’s conventional response crosses Pakistan red lines, leading to limited nuclear use · India responds with massive retaliation Pakistan Escalation Issues India Escalation Issues · Believes that it can conduct low-intensity confl ict with · Believes that conventional changes made since 1999 impunity thanks to nuclear capability Kargil confl ict suffi cient to deter Pakistan second · Fears Indian conventional superiority attempt or defeat it should deterrence fail · Small nuclear force structure leads to worries about · Believes that it knows Pakistani red lines well and can preemption and prompts “use it or lose it” mentality avoid crossing them · Small force size also prompts consideration of all-out · Unaware of effect of own conventional moves on use since unsure of signaling and insuffi cient for war- Pakistan’s considerations about nuclear use fi ghting mode · Any nuclear use by Pakistan will trigger tremendous pressure for all-out retaliation of themselves be suffi cient to threaten Paki- tional and nuclear forces to spiral into a crisis stan’s national survival and trigger a nuclear and introduce the risk of nuclear escalation. response, but they may incrementally begin The timeframe for this scenario is 2010 and to threaten the survivability of Pakistan’s each side has continued the development of its small nuclear force and its command and con- nuclear doctrine and force postures to the point trol structure. Fifth, even if Pakistan were where they approximate the medium option to choose a demonstration use of its nuclear outlined in chapter 2. capability, India would face both political and In the timeframe of this scenario, India has military pressure to respond with nuclear continued to modernize and expand its con- weapons, particularly considering the small ventional military capabilities with an eye to size of its own force and the potential vulner- dominating South Asia (including Pakistan) ability of its own civilian-dominated command and expanding its power projection capabilities and control system. Finally, of the three throughout the region. India is also keeping options available to Islamabad, its limited force a close eye on China’s ongoing military mod- size and the history of ineffective international ernization and attempts to keep pace although intervention on its side point toward an all-out at a lower level of overall expenditures. use of its nuclear capability. This modernization and expansion includes the acquisition of satellite-based intelligence Scenario 2 – Exercises in Nuclear Misperception capabilities, air-to-air refueling capability, lon- The second scenario examines the potential ger-range precision strike aircraft, aircraft for fi eld training exercises involving conven- carriers, and a signifi cantly enhanced ground

Future Crises and Escalation Potential 93 force capability for mobile combined-arms the survivability of its growing nuclear force. operations. As India modernizes, it also sees First, it raises the alert levels of a larger portion the need to expand the level and sophistication of its force by co-locating warheads and launch- of its command post and fi eld training exercises ers and pre-delegating authority to launch to to improve joint operations on the conven- lower levels in the chain of command. Second, tional level. As it watches Pakistan expand it attempts to raise the number of potential its nuclear force, New Delhi also begins to targets for India by announcing that its Hatf increase the pace and scope of exercises for its missile force, previously thought to have been own nuclear forces in order to be better pre- relegated to conventional missions because of pared to use them if necessary. Finally, with an its limited range and the coming on line of new eye to Pakistan’s nuclear force growing large missiles, is now formally dual-capable. Finally, enough that Islamabad may begin to consider Pakistan undertakes more regular and more tactical use against Indian conventional forc- extensive exercises with its nuclear forces, es, India expands the equipping and exercise including short-notice alert exercises, designed program for its conventional forces to be able to prepare it to act in case of a bolt-from- to operate in a nuclear environment. Despite the-blue attack by New Delhi. In addition, the growth of India’s conventional and nuclear Islamabad remains unconvinced of India’s stat- force structures, systems such as the early gen- ed policy of second strike for its nuclear forces, eration Prithvi ballistic missile remain in the believing instead that India would be prepared inventory, and it remains unclear to outside to strike fi rst with nuclear weapons if circum- observers, including Pakistan, whether it is a stances warranted. conventional or nuclear delivery system since In this scenario the geographic vulnerabili- it could carry either type of warhead. Finally, ties of India and Pakistan, evident in past crises the Agra summit process, begun in 2001, has (1965, 1971, 1987, and 1990), are exacerbated run its course over the next several years with in Pakistan’s case by the enhanced ability of few results. The process eventually grinds to a India to conduct rapid strikes on the ground halt, and tensions between India and Pakistan and precision strikes from the air against both at the time of this scenario are relatively high. its conventional and nuclear forces. Despite This scenario assumes that Pakistan con- the passage of time and investment, at least on tinues to modernize its conventional and India’s part, in new intelligence capabilities, nuclear forces with assistance from both China signifi cant gaps and shortfalls remain. One par- and North Korea, but that over time it falls fur- ticular shortfall relevant to this scenario is the ther and further behind India largely because lack of understanding of the other side’s nucle- of the mismatch in resources available to the ar doctrine, force structure, and, particularly, respective militaries. This not only concerns operational habits. This is due to the ongoing Islamabad from the perspective of a future con- secrecy that both sides have maintained about ventional clash, it also raises concerns that their nuclear forces and because each has only India may be acquiring the capability to con- recently begun expanding its exercises and duct a conventional preemptive strike against operations to correspond with a larger and Pakistan’s nuclear forces. In response to this more fully deployed nuclear force structure. concern, Pakistan takes three steps to improve The result is that neither side has settled

94 Nuclear Stability in South Asia into normal patterns of operation for its ing a surprise nuclear fi rst strike or dispersing expanded weaponized and deployed nuclear its nuclear forces to allow it to ride out any forces. Because of real shortfalls in intelligence Pakistani attempt to retaliate after an Indian capabilities, efforts to deny the other side intel- conventional assault. Assuming the worst case, ligence, and the lack of data points, neither side Pakistani planners also have in mind the possi- has a good picture of the other side’s regular bility that India is preparing for tactical nuclear operating habits and procedures for nuclear weapons use, retaining its strategic weapons forces, leading to an inability to distinguish to hold Pakistani cities hostage as its conven- between exercises and preparations for actual tional forces defeat the remainder of Pakistan’s use. army. Consequently, Pakistan’s response to all In something of a repeat of the Brasstacks of these Indian activities is to forward deploy exercise series in 1986-87, the Indian civilian its conventional forces and its Hatf missiles. and military leadership decides to hold a large This raises Indian fears that Pakistan would series of simulation, command post, and fi eld use its nuclear weapons in a tactical mode training exercises for both its conventional to overcome India’s conventional advantage, and nuclear forces, fi rst near the Chinese bor- keeping back its longer-range weapons to hold der and then near the Pakistani border. The Indian cities hostage. large conventional exercise is coordinated with As in 1987, attempts at communication a fi eld training exercise for India’s strategic to lower tensions fail to reassure either side. nuclear forces, including both aircraft and its Again, the problem is not that the means to Agni missile force. In a move intended both to communicate do not exist. It is either the lack test its own fl exibility and to keep Islamabad of will to use them or the inability to commu- off balance, India forward deploys Prithvi mis- nicate reassuring messages to the other side. siles with its conventional forces in the fi eld, in In their quests for more information, either for the desert training areas near Rajathsan and, reassurance or to prepare against a possible more importantly, near Lahore. attack, both sides engage in more aggressive Given Islamabad’s uncertainty about monitoring of the other, and intelligence gath- India’s capabilities (whether the Prithvi is ering efforts begin to look like preparations for nuclear, conventional, or armed with chemi- an attack. Efforts to ward off intelligence col- cal weapons), this element of the exercise in lection by aircraft and on the ground also raise particular is disturbing. The use of a large the prospects for clashes. Each side faces a set of number of Prithvis with either conventional decisions about whether to strike fi rst conven- or chemical warheads could slow a Pakistani tionally or even with nuclear weapons, fearing conventional response to an Indian attempt to that the other will take advantage of the tense cut the country in half. If armed with nuclear situation. warheads, such weapons could open a breach The elements of potential escalation in in Pakistani lines as well as target Lahore as this scenario derive in part from the relative Indian forces moved forward, should the exer- newness of the larger and more ready force cise turn into an attack. With the simultaneous structures assumed in the medium posture exercising of India’s strategic nuclear forces, postulated in chapter 2. Just as the Brasstacks Pakistan worries that India is either prepar- exercise series was a fi rst attempt by India to

Future Crises and Escalation Potential 95 test new theories and operational doctrines for its nuclear forces. Indeed, massed Indian forces conventional forces, the exercises in this sce- near the border would present a lucrative target nario are posited to be the fi rst, or among the for nuclear weapons used in a tactical or warf- fi rst, exercises of ready nuclear forces, includ- ighting mode, and Pakistani military offi cials ing those deployed with conventional forces. may press for such a use if they believe that In our medium posture model, we assume India is going to attack in any case. that India will retain a second-strike doctrine, Even if one assumes that neither side is so it is unlikely that New Delhi will change interested in conducting a nuclear or conven- course over the years and have its Prithvi mis- tional fi rst strike in such a scenario at least sile armed with nuclear warheads, particularly initially, escalation pressures exist because of because at the time of the scenario, 2010, the force structures, doctrines, and ongoing it is likely to have a signifi cant number of geographic vulnerabilities. Some aspects of the Agni-class missiles that would be both more medium force posture may be stabilizing in capable and more reliable. However, the level of this crisis, namely the elements of the force distrust likely to obtain during this type of sce- each side holds as an assured second-strike nario means that Pakistan would still assume force. Again, however, the issue is how the the worst case, that the Prithvis are armed with intentions of one side, combined with the dis- nuclear weapons and that India has adopted position of those forces, are viewed by the other a nuclear warfi ghting doctrine to augment its side and how the two sides see interplay with conventional superiority. potential tactical nuclear weapons use. In the medium force posture, Pakistan will Moreover, India’s and Pakistan’s ability to have a force just large enough to begin at least manage such a crisis and eventually de-escalate thinking about options for tactical use and it would remain questionable for the following warfi ghting with its nuclear weapons. India reasons. First, absent any signifi cant improve- will likely assume that this is the case and ments in general relations between the two also assume that the Hatf missile systems are sides, communications during such a crisis will the primary delivery vehicles for use of nucle- likely remain inadequate. Intelligence short- ar weapons in this mode. As the number of falls will likely remain, and analysis will still Pakistani missiles increases, and in a situation likely center on worst-case assumptions. In where Pakistani conventional forces are for- addition, if small-scale conventional clashes ward deployed with Hatf missiles, India will take place near the border, it is not obvious that worry that Islamabad would undertake a fi rst either side is equipped to quickly detect wheth- strike using these short-range missiles as a er an explosion was conventional or nuclear or way to counter India’s conventional superior- whether C3 is robust enough to pass informa- ity. Pakistani conventional forces would then tion up the chain quickly in the event of an be capable of striking east into the Punjab or accident involving a forward-deployed nuclear north into Kashmir, cutting India off from weapon. Unlike the Kashmir situation, domes- that province. In the meantime, Pakistan could tic politics may not play as big a role in such a reserve some nuclear weapons on dispersed scenario, so the pressures to escalate from this longer-range missiles to hold Indian cities hos- direction may not be as strong. In sum, how- tage, essentially daring India to retaliate with ever, such a situation could occur, and it could

96 Nuclear Stability in South Asia Exercise Crisis Summary Main Features of Crisis · Both India and Pakistan have the medium force postures described in chapter 2 · Field training exercises for conventional and nuclear forces by the two sides raise concerns of an attack out of an exercise · Both sides deploy short-range missiles that could be nuclear-armed with forward-deployed conventional forces · · Because of newness of medium force postures, neither side has a good intelligence picture of exercise patterns with nuclear or potentially dual-capable forces · Geographic vulnerabilities remain and exacerbated by threats of tactical nuclear weapons use India Escalation Issues Pakistan Escalation Issues · Prithvi missiles forward deployed with conventional · Hatf missiles have nuclear warheads and are forward forces deployed with conventional forces · Concern about Pakistan fi rst-strike doctrine and · Launch authority delegated to lower levels in chain of forward-deployed Hatf missiles – if nuclear could command allow Pakistani forces a breakthrough in the Punjab · Large portion of nuclear force on higher level of alert · Concerned that Pakistan has delegated launch · Lack of understanding of Indian exercise patterns and authority to lower-level commanders concern about preemption – either purely nuclear · Unaware or discounts that its conventional and/or or conventional with a tactical nuclear use for a nuclear exercises could be disconcerting to Pakistan breakthrough escalate to nuclear levels even though neither tive constraints about the use of force or the side was at fi rst even intending a clash at the possession of nuclear weapons become largely conventional or sub-conventional level. symbolic while realism becomes more pro- nounced in India’s foreign and security policy. Scenario 3 – Strategic Surprise Recognizing that a policy that actively shuns The third scenario involves the sudden open- India’s nuclear capabilities has proved fruitless, ing of a strategic capabilities gap between Washington abandons its sanctions complete- Pakistan and India that fuels instability and ly and, as part of a broader Asian balance of heightens the potential for crisis. This scenar- power strategy, engages India. With greater io is set in approximately 2015 and assumes economic and political ties, India enters a favorable domestic and international condi- strategic partnership or entente with the Unit- tions for India over the intervening fi fteen ed States. Washington and New Delhi slowly years. The success of India’s reforms spurs eco- develop a renewed military supply relationship nomic growth and helps sustain the necessary although India continues to purchase military resources to fund a signifi cantly larger nucle- items, particularly those that Washington is ar force along the lines of the high posture still reluctant to transfer, from Russia. described in chapter 2. A new strategic ori- At the same time, India perceives a rising entation, spearheaded by Jaswant Singh and Chinese threat. Beijing’s continued nuclear his followers, begins to take hold among the modernization and a shift toward limited elite as a guiding principle of statecraft. Norma- deterrence deeply alarms New Delhi. China is

Future Crises and Escalation Potential 97 reportedly developing a limited missile defense After years of secret and inconclusive nego- system with the technical help of Russia. As a tiations, the Indian government fi nally agrees result, India feels compelled to expand its own to an extensive arms deal with Russia. Despite forces to keep pace and to have the capability pressure from Washington, New Delhi pur- to defeat China’s missile defense. Relations chases one conventional submarine with SLBM with Pakistan remain tense. Periodic summits capabilities with an option to purchase more between the top leaders fail to achieve any in the future. After operational trials with an signifi cant breakthroughs on the Kashmir dis- Indian crew in Russia, the submarine is deliv- pute. India is also increasingly wary of the ered to India. While it will take a few more internal instability in Pakistan. Given these years to train the crew and achieve operational worrying trends, particularly China’s poten- profi ciency, the physical presence of an SLBM- tial military threat, India aggressively funds its capable submarine sends shock waves through nuclear program, with a particular emphasis Pakistan. Islamabad had assumed, based on past on sea-based capabilities. experience, that such a development was still In contrast, Pakistan remains in the grip of many more years away. An Indian announce- enduring challenges that it has yet to resolve. ment that at least two or more submarines will Despite reform efforts, the economy contin- be delivered over the next two years heightens ues to suffer from serious malaise. Domestic Pakistani fears further. With at least three subs, politics remains racked by instability. The India would be able to deploy one submarine civilian leadership is weak, corrupt, incom- armed with twelve ballistic missiles at all times. petent, and increasingly unpopular. Despite Pakistan worries that India will soon become having returned the civilians to power, the invulnerable to a fi rst strike designed to neutral- military maintains a fi rm grip over national ize India’s nuclear weapons and infrastructure security policy. It is also eyeing the civilian or command system. Despite India’s announced elites with increasing wariness. Rumors of a second-strike policy, Pakistan is convinced that new military coup abound on the streets. On India has never abandoned a fi rst-strike option. the surface relations with China seem strong, With India in possession of submarines carry- but strategic preoccupation with problems in ing SLBMs that can operate virtually undetect- North Korea and Taiwan diverts Beijing’s ed, Pakistan is extremely concerned about an attention from its southern front. China is Indian bolt from the blue to disarm or decapi- also genuinely constrained by nonproliferation tate Pakistan. Moreover, India’s growing land- agreements it signed or committed to in prin- and air-based nuclear arsenal would guarantee ciple with the United States. As a result, China New Delhi an assured second-strike capability is not as forthcoming as in the past with tech- to deter any remaining Pakistani response capa- nology or weapons transfers, and Pakistan feels bility. At the same time, Pakistan’s strategic posi- increasingly isolated and insecure in the region tion vis-à-vis India in the conventional military especially as Indo-American relations continue realm has deteriorated in relative and absolute to solidify. As India accelerates its convention- terms. Islamabad concludes that it can no longer al and nuclear capabilities, the asymmetry in close the strategic gap or compete with India power between the two rivals grows. much longer.

98 Nuclear Stability in South Asia Pakistan perceives the submarine delivery submarine attack against the Indian platform to be a harbinger of major changes in the stra- during sea trials. Such an attack would serve tegic balance that would severely undermine as a warning to India that Pakistan will not its security in the coming years. The leadership tolerate the fi elding of such a capability in the assumes that a fully operational submarine future. If done swiftly and effectively, the onus fl eet would enable India to exercise escala- would fall on India to escalate in response, tion dominance over Pakistan in any situation. and Pakistan would attempt to bring in out- Fearing that India’s nuclear and conventional side actors to stay New Delhi’s hand. Such an advances would irreparably unravel mutual option is not too far-fetched and would likely vulnerability, a desperate Islamabad begins to be painted by its advocates as the equivalent to consider riskier options. The military presents what it believed India almost tried in the 1980s, the civilian government in Islamabad with sev- striking against capabilities before they became eral policy alternatives. First, it could seek a problem. Moreover, some in Pakistan might to reverse the impending strategic imbalance see it as a long-overdue counter to the Israeli by seeking Chinese or North Korean assis- attack on the Osirak reactor in Iraq. tance in providing some sort of countervailing Clearly, the latter two policy options, if technology transfers such as MRV or MIRV chosen, would most likely escalate to a nucle- capabilities or missile defense capabilities. ar confrontation. Would Islamabad consider Since such a transaction might take several such drastic actions, particularly the preemp- years to complete, Islamabad must weigh the tive strike? It would depend largely on the benefi ts of new capabilities against the time quality of Pakistan’s intelligence capabilities, required to acquire and integrate them into its its military capability, and the confi dence arsenal. On a related point, there is uncertain- the leadership had in it to pull off such a ty about exactly how such capabilities could strike. Good intelligence might provide ade- be used to counter an SLBM, which many quate information with which to anticipate in Pakistan consider to be the magic bullet the arrival of the submarine. However, the that would unravel the nuclear balance.5 Sec- tactical and targeting intelligence, especially ond, confi dent that its fi rst-strike nuclear force if Pakistan were to attempt the strike while is still a useful deterrent against an Indian the submarine is at sea, would be a signifi cant nuclear or conventional attack, Pakistan could challenge. If Pakistan were to respond with a gamble that this would be the one last oppor- preemptive strike, India would be under severe tunity to take back Kashmir. This would entail domestic political pressure to retaliate in some undertaking the type of extensive support for manner, either with a devastating convention- a low-intensity confl ict in Kashmir detailed al attack on Pakistan’s nuclear infrastructure in the fi rst scenario. The third, and riskiest, (thereby raising the possibility of a Pakistani option would be a preemptive conventional or nuclear riposte) or a nuclear strike of its own. nuclear attack, presumably against the capa- In Pakistan, domestic political pressures bility that threatened to upset the strategic would be signifi cant for some sort of tangible balance. Such a strike could manifest itself reaction to India’s acquisition of the SLBM in an air or missile strike (conventional or capability. If Islamabad were not assuming nuclear) to destroy the submarine in dock or a the worst about New Delhi’s intentions, then

Future Crises and Escalation Potential 99 Strategic Surprise Crisis Summary Main Features of Crisis · Both India and Pakistan approaching the high force structure posited in chapter 2 · India pushing particularly hard to compete with China · Relations between the two remain tense · India’s relationship with the United States on the rise; Pakistan’s still relatively distant · With little warning, India acquires SLBM capability · Pakistan responds in one of three ways: 1) seeks countervailing capabilities from China/DPRK; 2) tries a Kashmir attack fearing that strategic window closing; 3) conducts preemptive strike against Indian submarine(s) Pakistan Escalation Issues India Escalation Issues · Strategic surprise victim · Capability acquisition is response to China; impact on · Increasing conventional and nuclear gap raises fears Pakistan downplayed that window of opportunity is closing · India could also be surprised if Pakistan acquired MIRV · SLBM seen as highly destabilizing because India has capability from China assured second strike or, in Pakistan’s mind, bolt-from- · If Pakistan conducted a preemptive strike against the the-blue fi rst strike capability Indian submarine, pressure for nuclear response or · High level of domestic pressure for short-term devastating conventional response very high, in turn response to the surprise risking Pakistani escalation to the nuclear level

acquisition of an SLBM capability should not, either side, such as missile defense (by either at least in theory, be seen as destabilizing. side), large numbers of missiles, or MIRV For a country with a declared second-strike technology. Most of these scenarios involve doctrine, the acquisition of an assured second- technology transfer, so outside actors play a role strike capability would be logical, and because in precipitating the crisis as well as potentially of the geographic proximity of the two states, having a role in ameliorating it. having a submarine-launched weapon provides The potential for escalation to the nuclear no new capability in terms of accuracy or fl ight level in this scenario has less to do with the time for India. However, Pakistan’s reaction actual force structures in place at the time it to India’s nuclear tests in 1998 provides evi- takes place (assuming medium moving to high dence that politics at times trumps reason (and postures for both India and Pakistan) than signifi cant pressures from the international with the fact that the levels of capability are community) in South Asia, as it does in many changing rapidly and unequally. Because one parts of the world. This scenario, which may side believes that it has been trumped strategi- unfold relatively slowly (because of the time cally, it may be willing to take high-risk actions between announcement and delivery of the in an attempt to recover. In this scenario this submarine), may allow time for international may involve attacks that could invite nuclear intervention to tamp down a potential crisis. retaliation or begin a confl ict spiral that could The scenario could also be played out using end in a nuclear exchange. Pakistan in par- acquisition of other strategic capabilities by ticular would be more open to risk taking if

100 Nuclear Stability in South Asia India continues to pull away in both con- that could be useful in thinking about the range ventional and nuclear capability over the of possibilities across a plausible spectrum in next decade-plus. It should be noted that the terms of both force structures and crises. assured second-strike capability that is part From a U.S. policy perspective, the point of an SLBM force is stabilizing theoretically then is not to attempt to persuade India or Paki- because it has deterrence value, but submarine- stan to adopt one or the other force posture (that launched capabilities raise other possibilities is, low versus medium versus high). Rather, U.S. for accident and incident. Sea-based assets policy should focus on making sure that regard- raise new and different command and control less of the force posture and doctrines that India issues, particularly if India wishes to keep tight and Pakistan adopt, chances of nuclear use are civilian control over its nuclear assets. Simi- as low as possible. This will be a signifi cant chal- larly, the possibility of accidents at sea raises lenge. Despite the fact that the United States has different opportunities for misunderstanding more infl uence with India and Pakistan than and possible miscalculation, particularly if does any other single country, that infl uence is Pakistan attempts to begin regular surveillance still relatively limited. The reasons for this are of an Indian SLBM capability. two. First, as described in chapter 2, India and Pakistan will choose their doctrines and asso- Conclusions ciated force structures based on a number of In some ways, this chapter outlines the worst variables, only some of which the United States possible cases and combinations of crises and can affect, and then possibly only marginally. future force postures. However, it has been Second, and as will be discussed more fully in designed to draw out the most troubling insta- chapter 5, the tools that the United States has to bilities in those combinations. As has been infl uence Indian and Pakistani choices on doc- shown above, each doctrine and associated force trines, force structures, and during crises are structure has the potential to be stabilizing or limited by competing priorities, lack of funds, destabilizing depending on a host of exogenous and – in some instances – international obliga- variables such as the perceptions of decision tions. makers at the time, the quality of intelligence, As is evident in the three scenarios above, and domestic political pressures. However, each the escalation issues vary with not only the sce- force structure and doctrine carries with it nario but with the force structure and doctrine unique qualities that could contribute to insta- of each of the two antagonists. Therefore a sup- bility in different ways. None of the posited porting issue for United States policy making is scenarios or associated force structures and doc- determining the doctrine and force structure of trines is necessarily more or less likely to come each side – not an easy feat, as noted in chapter to pass. The best that can be said is that crises 2 – and then determining if and when those involving the low force structure will occur ear- may change. This will not always be obvious ly in the study period and those involving a high given the opaqueness of the two sides’ nuclear force structure could not come to pass until near policies, but an overall increased engagement by the end of the study period, toward 2015. The the United States should improve Washington’s purpose of positing these scenarios is to provide insights and sources of information. Based on analysts and policy makers with information that information and on a renewed focus on sta-

Future Crises and Escalation Potential 101 bility issues, the United States should be able to craft policies that raise the nuclear threshold in South Asia. Specifi c suggestions for how the United States can craft and carry out such policies are examined in chapter 5.

102 Nuclear Stability in South Asia CHAPTER FIVE

Conclusions & Recommendations

The United States has a vital interest in keeping nuclear weapons from being detonated, either accidentally or intentionally, in South Asia. It also has a vital interest in preventing further proliferation of nuclear capabilities around the world – to states and non-state actors alike. The latter concern has only escalated in the aftermath of the September 11, 2001 terrorist attacks on the United States. India and Pakistan have been a potential source of nuclear proliferation since they fi rst acquired nuclear weapons technolo- gies, but their recent change to overt nuclear weapons states now means that the issue of further horizontal proliferation can be addressed openly for the fi rst time. These two vital U.S. interests – keeping nuclear weapons from being used and continuing to staunch their further horizontal proliferation – are not incompatible. As was argued in chapter 1, pursuing improved safety, security, and crisis stability in South Asia need not undermine U.S. global nonproliferation goals. Therefore the United States should craft new policies designed to enhance stability, particularly crisis stability, in the region and work to advance those policies in a systematic manner. Based on the analysis in the preceding chapters, this chapter provides policy recommendations for pursuing these policies and lays out implementation considerations.

Nuclear Stability in South Asia 103 Geopolitical Considerations control of the country to a democratic, civilian One of the most signifi cant challenges for the leadership by then. The terrorism issue is going United States in designing and implementing to be more diffi cult and more complex in large new policies to enhance nuclear stability in part because of the events of September 11 South Asia is how to integrate and prioritize and the tremendously diffi cult and fractious those policies with other interests the United situation in Afghanistan. Even if the Taliban States has with each country and in the region and the al-Qaeda network are vanquished in as a whole. In the case of India, the United Afghanistan, the situation will not be settled States has an interest in developing a strate- for years if not decades. The implications of gic relationship with a major democratic state this for U.S. relations with Pakistan will there- with a growing economy. This strategic rela- fore color the interactions between them for tionship will be supported by expanding trade years. In terms of nuclear security and stability and investment and broadening and deepening issues, the new closeness between Washington political-military relations as part of a policy and Islamabad could be either helpful or harm- of increasing stability in a critical arc extend- ful, depending on how it is handled. On the ing from the Gulf to the South China Sea. one hand, enhanced dialogue that results from India may play a role in U.S. policies regard- cooperation in the war on terrorism should ing China. U.S. interests in and with China increase trust, improve information gathering, will continue to be signifi cant and are likely and provide deeper insights into Pakistan’s to remain higher on Washington’s agenda than concerns about its general security situation. issues with India in the short to medium One would therefore expect that to result in term other than perhaps issues related direct- increased leverage on nuclear stability issues. ly to the war on terrorism. Therefore look for However, on the other, the need for ongoing relationships with India to play a supporting cooperation on terrorism between the two role, whatever Washington’s ultimate policy is states might push nuclear stability issues down toward China. on the list of priorities. In addition, if Islam- With Pakistan, the primary interest as abad feels that Washington needs its goodwill of fall 2001 is Washington’s declared war on terrorism issues, it might be less willing to on terrorism, particularly the al-Qaeda terror bend on nuclear stability considerations if the network and the ruling Taliban regime in United States asks Pakistan to take steps that Afghanistan. As part of this battle, the United might be uncomfortable or costly (such as cut- States also has an interest in seeing Pakistan ting off its technology import relationship with remain an ally and not turn into an extremist China). In other words, Washington’s leverage state. In the longer term, the United States has on nuclear stability may be hampered by its an interest in the restoration of democracy in need for Islamabad’s assistance on the terror- Pakistan and generally in building and main- ism front. taining good relations with a Muslim state that could be a moderating infl uence in the Mus- Realities and Constraints lim world. The democratic rule issue may be As noted in chapter 1, neither India nor Paki- dispensed with in October 2002 if the Mush- stan will be relinquishing its nuclear weapons arraf government keeps to its promise to return capabilities for the foreseeable future. There-

104 Nuclear Stability in South Asia fore U.S. policies must address the reality on States had to say about safety, security, stabili- the ground in the region and be fl exible and for- ty, or restraint measures. While understanding ward looking enough to anticipate likely future was expressed as to the importance of these developments. Even if both countries maintain issues during bilateral discussions, New Delhi doctrines and force structures that are along and Islamabad felt that any and all proposals in the lines of what was called the low posture these areas (minor though they were) during in chapter 2, the possibilities exist for acci- the Clinton administration were tied directly dents, incidents, and crises to lead to escalation to rollback and elimination. to a nuclear level. The United States, over U.S. policy on stability issues should be the course of two-plus generations, has built tailored to current force structures and doc- extensive experience and developed numer- trines in South Asia, but it also must have an ous technologies to address these issues, and anticipatory element to it. As noted in previous it needs to begin to apply this expertise in the chapters, doctrines and force structures will be South Asia region. driven by a number of variables, including the In order for such a policy to be viable, the occurrence of crises. The stability issues will United States must signifi cantly decrease its vary according to the actual force structure, policy emphasis on rolling back India’s and doctrine, and type of crisis, so U.S. policies can- Pakistan’s nuclear capabilities. Such an empha- not be of a “one size fi ts all” variety. As part sis, evident during the Clinton administration of a comprehensive policy on these issues, the although never expressed as such, had det- U.S. must continually monitor the various driv- rimental effects. First, even though “cap, ers that shape Indian and Pakistani decisions rollback, and eliminate” was a phrase not about doctrine and force structure and, where necessarily found in Clinton administration possible, anticipate when changes might occur. public pronouncements on the issue, these Policy changes can then be considered ahead were assumed by much of the U.S. bureaucracy of the event and possibly head off certain to be the core goals of the policy toward these decisions that may prove destabilizing in a two countries’ nuclear capabilities. Therefore future crisis. Of course, this is more easily policies that may have considered accepting the said than done, particularly when dealing current reality and focusing on stability were with issues that both states are likely to keep not explored in any detail. For example, it was very closely held. However, as will be detailed assumed that much of the possible technology below, improved dialogue with each state on sharing with either India or Pakistan related to these issues should supplement intelligence nuclear safety was illegal under article 1 of the and should bring a degree of transparency to NPT, but apparently no actual legal review was changes that could prove useful for the United conducted within the interagency. This may States and for India and Pakistan in pursuing be changing as the urgency of events following more stable relations. September 11 appears to be accelerating a vari- The United States should begin pursuing ety of reviews in this area.1 Just as harmful, a broad range of stability-related policy initia- the assumption that rollback was at the core of tives immediately. It should not wait for other America’s policies made both India and Paki- elements of policy to fall into place and cer- stan less willing to listen to what the United tainly should not wait until a crisis breaks out

Conclusions and Recommendations 105 before attempting to intervene on these types The fi rst key issue in heading off a crisis of issues. The Kargil crisis of 1999 shows that involving Kashmir that could escalate to nucle- crises can emerge quickly after changes occur ar use is the possibility that Pakistani decision in the nuclear status of India and Pakistan makers believe that their nuclear deterrent and, indeed, can arise even when it appears provides a cover for a low-intensity confl ict that relations between the two are improving. strategy against India. As outlined in chapter One recent simulation exercise on South Asia 4, Pakistan could continue this strategy and at involving regional participants and including some point attempt a reprise of the 1999 Kargil escalation to the nuclear level concluded that crisis. To help avert this possibility, the United engagement by the international community, States needs to address the dangers of this type including the United States, was of little use of strategy with Pakistani decision makers as in stabilizing the confl ict or preventing escala- part of a broader, regular dialogue on security tion if such engagement occurred only after the issues. In these discussions, which must be at crisis had begun.2 senior levels and regularly held, one point that must be emphasized is that the United States Issues and Policy Recommendations will not support Pakistan if such a strategy is The following discussions explore the escalation attempted. Basically the point must be made issues and risks at play in each of the scenarios that the United States will react as it did in the presented in chapter 4, along with recommenda- 1999 Kargil confl ict or indeed more strongly. tions for policy initiatives that the United States The United States should debunk any myths should take to head off similar crises. that the Pakistani civilian leadership “sold out” to Washington and that absent that, Kargil Kashmir and Low Level Force Postures would have been a great military and politi- The fi rst crisis scenario outlined in chapter 4 is cal success. This tough talk will not be easy, the closest to the situation on the ground today. but it is part of a dialogue that is necessary to As noted in chapters 2 and 4, there is reason to re-establish serious connections with the Pak- believe that both India and Pakistan might not istani government on both the military and move much beyond their current doctrines and civilian sides. It will be particularly diffi cult, force structures for at least part of the study as noted above, if Islamabad comes to believe period – through 2015. In addition, it is fairly that the United States needs it more for the war safe to assume that the Kashmir dispute will on terrorism than Islamabad needs Washing- not be settled any time soon. In fact, the fi ght- ton for overall political, economic, and military ing in Kashmir, and the rhetoric from both support. sides about it, has only escalated during the A second policy remedy that may help soft- U.S.-led war in Afghanistan, thereby showing en the strong no-support language if Islamabad that it is never clear what will trigger a crisis continues to pursue the insurgency option is or how outside intervention in the region will one of discussing with Pakistan the broader play out. Therefore, this scenario, and the esca- conventional balance between Pakistan and lation issues identifi ed in it, provides the ideal India. Again, in the context of regular, high- starting point for considering U.S. policy pre- level dialogues, the United States should begin scriptions in the near term. an exchange of views on the current and future

106 Nuclear Stability in South Asia conventional balance between Pakistan and a negotiated solution to the issue. While an India. In these dialogues, the United States agreed solution to this long-standing dispute should have three goals. The fi rst is to provide would seem an obvious way to head off the a forum for Pakistan essentially to set forth its entire problem, it should not be expected that perspectives on India and feel that it has the stepped-up U.S. engagement would lead to this ear of a strong friend. Second, the United States in the short or medium term. Rather, in terms should, through presentations of its own views of stability, the goal of the increased U.S. activ- and analyses, seek to knock down unreason- ity in seeking a solution to Kashmir is to keep able or unrealistic assessments by Pakistan of Pakistan from believing that it has no option either its own prowess or Indian capabilities. other than the use of force. In addition, the In particular, the United States should focus United States should press India not to create on countering assessments that have political what Pakistan may see as facts on the ground appeal among more radical factions within that preclude a negotiated solution. Such efforts Pakistan who advocate jihadi strategies or tac- are likely to make India somewhat uncomfort- tics against India. Even if some of these groups able, but countervailing dialogues with India are either rounded up or discredited as part both on its relations with Pakistan and on a of Musharraf’s policy of supporting Washing- broader range of strategic issues will likely soft- ton against terrorism, strong feelings about and en New Delhi’s stiffness. against India are likely to prevail in broad sec- A second key issue in heading off possible tors of Pakistani opinion. Finally, the United escalation, should a second Kargil crisis take States should use these dialogues to begin to place, involves Indian and Pakistani percep- identify an agreed set of defensive defi ciencies tions of the nuclear threshold. The paradox in in Pakistan that could potentially be remedied this type of scenario is that in order to pro- by a renewed military supply relationship with mote nuclear stability, the views held by each the United States. While the renewal of such a country should end up being the exact oppo- relationship has begun in the wake of Septem- site of the other’s. To have India refrain from ber 11, it needs to be carefully assessed on an actions that might cause Pakistan to cross the ongoing basis to ensure that it serves not just nuclear threshold, New Delhi must believe that the short-term purpose of rewarding Pakistani Pakistan’s threshold for nuclear use is fairly cooperation on terrorism but that it contributes low or at least it must be uncertain about exact- to raising the nuclear threshold between Islam- ly where the threshold lies. To keep Pakistan abad and New Delhi. In any dialogue and in from using nuclear weapons, Islamabad must the event of a renewed military supply relation- be confi dent enough in its capabilities (includ- ship, it should continually be made clear that ing their survivability) to have its threshold be the United States is in no way giving a green fairly high. light to Pakistan for escalation in Kashmir and, The policy solutions for this situation are in fact, that Washington expects just the oppo- not terribly clear cut, but again they begin site to occur. with a more serious dialogue between the Unit- A third policy remedy for a Kashmir-based ed States and the two antagonists – one that scenario is more active engagement on the part starts well before any future crisis erupts. In a of the United States in pushing both sides for U.S.-Indian bilateral dialogue aimed at promot-

Conclusions and Recommendations 107 ing stability and managing perceptions of the a “use it or lose it” situation in a crisis such nuclear threshold, the goal should be for India as this one. Therefore, in addition to more to believe that Pakistan’s threshold for nuclear general reassurances about the conventional use is low. In the Kargil II scenario, the point military balance, future U.S.-Pakistani dia- is to keep India from engaging in convention- logue on security matters should focus on al escalation that crosses Pakistani red lines, raising Pakistan’s nuclear threshold. Specifi - lines that Islamabad has established for itself in cally, the United States should walk Pakistan deciding whether to use its nuclear weapons. through the diffi culties that India is likely While some Indian government offi cials and to encounter in targeting Pakistan’s nuclear analysts claim to know Pakistan’s red lines, weapons, particularly those mounted on mobile they can never be sure that they are right or missiles. While not going too deeply into intel- that the situation may dictate a shifting of ligence matters on Indian capabilities, the those thresholds for Islamabad. The United United States can discuss its own diffi culties States should emphasize these uncertainties in in countering mobile missile systems, draw- dialogues with India and offer to play out sce- ing on its experiences hunting for SCUDs in narios with New Delhi in high-level simulation Operation Desert Storm. While both India exercises to test its assumptions and its levels and Pakistan are aware of these issues, they of certainty, focusing particularly on the inad- may not have suffi cient information to perform vertent crossing of thresholds. The United rational analyses and they each certainly over- States should also point out that two different estimate the capabilities of the other side. The paths could lead to Pakistani nuclear escala- point in this exercise is to reassure Islamabad tion. First, Pakistan’s nuclear threshold could that it has very little reason to panic in a crisis be lower than believed in New Delhi, and if its forces are mobile and dispersed. an Indian move such as crossing the line of Of course, Pakistani fears could have control in hot pursuit of insurgents may be another source – worry about a decapitating enough to trigger a nuclear response. Second, Indian strike at command and control assets. in conducting conventional operations against For this reason, and because of the impor- insurgency-related targets, India may inadver- tance of command and control in preventing tently degrade Pakistan’s command, control, unauthorized launches, another element of the and communications systems (C3), air defense, U.S.-Pakistani dialogue should focus on com- or the actual elements of its nuclear deterrent mand and control. The discussion should draw (e.g., aircraft, missiles co-located with logistics on the rich store of lessons and experience centers for the insurgency operations). This that the United States has had in handling could lead Pakistan in turn to escalate to the both its strategic systems and, more relevant nuclear level, fearing that its deterrent capabil- to Pakistan, in the command and control of ity was being rapidly eroded. tactical systems deployed during the Cold War On the Pakistani side, the concern is that in Europe as part of the U.S. commitment to Islamabad’s low nuclear threshold will be low NATO. Obviously, the United States cannot due to its small nuclear force and India’s con- pass on details of its own arrangements, and ventional superiority. In view of these two similarly Pakistan is not going reveal all of factors, Islamabad could believe that it is in its procedures and capabilities in the C2 area,

108 Nuclear Stability in South Asia but a discussion of principles, issues, concerns, istan about the security of its nuclear weapons generic solutions, and procedural and techni- and materials produced no agreements, but it cal fi xes is certainly possible. The dialogue can is not clear whether any carrots were offered. lay the groundwork for possible future techni- A fourth issue in this scenario is the deci- cal assistance with command and control if it sion about an Indian response if Pakistan used does not contravene existing U.S. laws or obli- its nuclear weapons in a very limited, dem- gations. If such assistance is deemed essential, onstration strike after Indian conventional then it may become necessary for the executive escalation. Indian analysts and policy makers branch to propose changes to domestic laws or have stated that any Pakistani nuclear use negotiate for exceptions to any international would be met with overwhelming retaliation, obligations. The goals of a C2 dialogue, then, essentially destroying Pakistan as a func- would be twofold. First, it should provide infor- tioning state. Such statements are sensible mation that would lower Pakistani anxieties declaratory parts of a deterrent policy as they about an Indian decapitating strike, thereby are both credible and extremely daunting. raising their nuclear threshold and lengthen- Indeed, if Pakistan were ever to use even a ing decision-making time in a crisis. Second, single nuclear weapon against an Indian tar- the dialogue would seek to strengthen tech- get (even if it were military forces in the fi eld), nical and procedural controls on weapons the political pressure on New Delhi for a large- to limit the possibilities of an unauthorized scale retaliation would be enormous. The time launch, particularly with dispersed weapons. to temper this pressure is well before any The details will be important here as the goals crisis breaks out and decision makers must are security and stability, not enabling Paki- contemplate their limited options in the glare stan to keep its forces on a hair-trigger alert of the public eye and under time constraints. (a possible perverse result of providing certain To do this, the United States should engage types of C2 assistance).3 This command and with India in a series of simulations that control dialogue may not necessarily be repli- walk through the consequences of nuclear cated with India, as Pakistan’s concerns are strikes, both by Pakistan and by India in retali- different because of its geographic situation, ation. These simulations should walk through its overall strategic inferiority to India, and its the political, economic, environmental, and consequent adoption of a fi rst-strike doctrine. humanitarian consequences of nuclear use In other words, this set of initiatives is probably for both India and Pakistan. While aware best approached as a set of bilateral talks, the of these consequences in general terms, such United States with Pakistan and, separately, information is not widely known either in pol- the United States with India. One set of quid icy-making circles or in the general public. The pro quo that the U.S. might require in order goal is to provide Indian decision makers with for higher-order technology release in this area a clear sense of the alternative consequences would be certain restraints on the buildup of of acting, not acting, or responding at differ- forces and the securing of forces and materi- ent levels of force. No option is without its als in ways that make them diffi cult either to costs, and the United States should not neces- steal or to use without authorization. Fall 2001 sarily seek to force more serious consideration discussions between the United States and Pak- of one or the other. It should, perhaps under the

Conclusions and Recommendations 109 Recommendations for Low Force Postures and Kashmir Scenarios India Pakistan · Politcal-Military Dialogue · Political-Military Dialogue Themes Themes - Uncertainty of Pakistani red lines - Debunk myths on Kargil 1999 outcome and blame - Broader strategic cooperation with the United - Nonsupport for a second Kargil-like crisis States – shifting focus from Pakistan - Regular exchange on Pakistani-Indian - Specifi c implications and downside risks for India conventional balance of nuclear use on Pakistan - Command and control dialogue (including safety, · Bilateral (U.S.-Indian) simulations and exercises to security issues) highlight escalation issues - Reassurance on survivability of arsenal - Specifi c implications and analysis of downside · Discussion and technology transfer to improve Indian risks of nuclear use detection and analysis of WMD use · Bilateral (U.S.-Pakistani) simulations and exercises to · Increased U.S. involvement in pursuing diplomatic highlight escalation issues solution to Kashmir · Regular joint defense assessments · Increased U.S. involvement in pursuing diplomatic solution to Kashmir · Provide technical assistance to bolster C2 capabilities consistent with safety and security goals · Deepen/renew understanding of Pakistani military through increased International Military Education and Training programs and exchanges

rubric of discussions about consequence man- could be based on its assumptions about the agement, walk through the various problems size, scope, and target of a Pakistani fi rst strike. that would come about for India from essen- Improving India’s ability to have accurate infor- tially destroying Pakistan. mation about such a strike could have a deter- A second policy proposal to lower the chance rent effect on its potential retaliation. that India would automatically retaliate, or The fi nal escalation issue in a Kashmir would do so on a massive scale, involves enhanc- scenario with low force structures is the pres- ing India’s ability to detect and analyze explo- sure on Pakistan, due to its relatively small and sions that occur during wars or crises. In other potentially vulnerable nuclear force structure, words, the United States should assist India in to choose an all-out attack rather than a lim- improving its capability to determine whether a ited demonstration use. Again, the issue here nuclear weapon has been detonated and if so, its is working on Pakistani decision makers’ cal- location and yield. During crises, it is possible culus of what would be a better outcome for that the use of conventional munitions mounted their country. Similar to the discussion that the on ballistic missiles could be mistaken for nucle- United States should have with India about its ar munitions. This could lead to preparations choices of retaliation type and scale, the United for a nuclear response. Additionally, the size States should engage in such discussions with and scope of a potential Indian nuclear response Pakistan, noting the signifi cant downsides of

110 Nuclear Stability in South Asia any nuclear use in terms of such factors as India and Pakistan in an effort to avert future the likelihood of Indian retaliation, the fallout crises that could arise out of exercises. This potential for Pakistan, and the international type of crisis, and the measures designed to stigma that would result (assuming Pakistan head it off, goes well beyond previous discus- survived as a functioning state). Similarly, sions about missile test notifi cations. the United States should conduct simulation At the very least, the political-military dia- exercises with Pakistan that sketch out the dif- logues that the United States should hold with fering outcomes of a war with India where each side should address the general utility of the two countries use nuclear weapons and nuclear weapons in a tactical role. Drawing on where they do not. A conventional defeat for U.S. experience in Europe during the height of Pakistan, as disastrous as it may appear, is the Cold War, Washington should make clear ultimately reversible and survivable whereas to each side that the number of tactical nuclear a nuclear exchange could have far more devas- weapons required to make a real difference on tating consequences for either side. While the the conventional battlefi eld is very large. Those situation of Kuwait after its occupation by Iraqi numbers are well beyond what either side forces in 1990 is not a perfect analogy, it should is either planning or could manufacture in be emphasized to Islamabad that conventional any reasonable timeframe. While each side is defeat does not equal annihilation in scenarios aware of this in terms of its own planning, such as those sketched out in chapter 4. the tendency is to worst-case the other side’s capabilities and intentions in this area. The Nuclear Exercises and Medium Force Postures message should be one of reassurance. If either This scenario, outlined in chapter 4, could side sees delivery vehicles on the other side occur as the force structures of the two sides during an exercise that could be nuclear capa- evolve over time. As soon as signifi cant mili- ble, they should not worry. Neither side has tary exercises are held involving short-range the numbers, the targeting information, or the missiles and as exercises (whether command tactical doctrine to use nuclear weapons in a post or fi eld exercises) begin to take place using militarily signifi cant manner on the conven- each country’s longer-range delivery systems, tional battlefi eld. the possibility exists for mistrust, misreading, Given that Pakistani and Indian exercises over-reaction, and escalation. Therefore, policy over time will involve new elements (ballistic solutions aimed at heading off such potential missiles – regardless of the warhead in question for escalation must be in place before such – different command and control schemes, and exercises become routine. Remedies for escala- enhanced reconnaissance capabilities), uncer- tion in this type of scenario involve traditional tainty and concern are likely. To attempt to CBMs designed to improve transparency about head off such worries, particularly as they conventional military exercises and more spe- could increasingly affect decisions about nuclear cifi c discussion about the capabilities of nuclear weapons use, the United States should active- weapons used in a tactical, warfi ghting mode. ly support the establishment of a more robust Again, the United States has signifi cant experi- mechanism for notifying, discussing, and moni- ence in both of these areas, and this experience toring exercises. As with all of the recommen- should be systematically shared with both dations in this chapter, the concept is to begin

Conclusions and Recommendations 111 small and build over time. Similarly, discussions security, and command and control issues for and mechanisms of this sort should build on systems of this type. current agreements, adding depth, complexity, In consultation with both India and Pak- and monitoring capabilities to those measures istan, the United States should also explore already agreed to but not yet implemented on a whether certain elements of their nuclear forc- regular basis. Many proposals for such CBMs es can be subjected to monitoring to lower the already exist, both in agreements signed by India chances for deployment races occurring during and Pakistan and in various nongovernmental exercises or other crises. Ironically, the chance and expert forums.4 What they have lacked in for competitive spirals of increasing readiness is the past has been systematic and high-level polit- particularly strong if each side keeps its force at ical support by the United States and a place for a relatively low level of readiness and has war- them as a centerpiece in discussions with India heads unassembled and geographically separat- and Pakistan. Even this type of pressure by the ed from delivery systems. Information exchange United States may not suffi ce to get agreements and monitoring must be done in ways that do implemented systematically, but without it, the not invite preemption. In other words, gathering political will in the region is likely to be found all warheads and bombs in a single place and lacking. providing for transparency as to its location and With U.S. diplomatic support and consis- status would provide a high degree of confi dence tent with agreements made at Lahore, the two that the weapons were not being moved or mated sides should fi rst be encouraged to meet reg- to delivery systems, but it would also be an ideal ularly and discuss what current and future target for preemption. Therefore, remote moni- elements of exercises would raise the most toring techniques and limitations on the data worries. The discussion could then move on shared (including exact location) would have to questions of whether those elements could to be adopted. Such types of systems have been either be eliminated or could be notifi ed in explored in the United States for use in U.S.- advance and monitored more comprehensive- Soviet/Russian arms control regimes and have ly to ensure that they did not mask attack also been explored for use in South Asia.6 What preparations. The United States should offer the United States has not done is to put its diplo- monitoring technology, beginning with demon- matic weight, as well as resources and technolo- strations and ending with offers of discounted gy, behind such an endeavor to ensure that it gets transfers if one or both sides are interested. off the ground and is used on a sustained basis. Given the ambivalence in the region about past Such a move would not only provide incentive CBM efforts, the U.S. should try to encourage for the two states to consider the measures seri- a competition between the two sides, seeing ously; it would also cushion efforts from the ups for example which would be the fi rst to invite and downs in relations between India and Paki- the U.S. to demonstrate its Open Skies aircraft stan, a phenomenon which has destroyed several capabilities on a national exercise.5 Finally, if promising starts at more systematic confi dence- either or both sides are actually contemplat- building and security measures. ing making some of its shorter-range missile One way to increase the chances that nucle- systems dual- or tri-capable, the United States ar stability measures are more systematically should step up its offers of discussions of safety, and rigorously applied is to have them linked to

112 Nuclear Stability in South Asia Recommendations for Medium Force Postures/Exercise Scenario India Pakistan · Politcal-Military Dialogue · Political-Military Dialogue Themes Themes - Same as low posture, plus discussion of low - Same as low posture, plus discussion of low utility of nuclear weapons in a tactical warfi ght- utility of nuclear weapons in a tactical warfi ght- ing mode ing mode - Regular discussion (bilateral between antagonists - Regular discussion (bilateral between antagonists or with U.S. participation) on specifi c exercise or with U.S. participation) on specifi c exercise concerns concerns · Enhanced exercise notifi cations and information · Enhanced exercise notifi cations and information exchange exchange · U.S. monitoring technology demonstration (e.g., Open · U.S. monitoring technology demonstration (e.g., Open Skies) Skies) · Exploration of remote monitoring of nuclear weapons · Exploration of remote monitoring of nuclear weapons sites sites · Establishment of permanent, joint risk-reduction · Establishment of permanent, joint risk- reduction center center a permanent joint center that acts as a monitor- India-Pakistan context, strategic surprise can ing center and clearinghouse for information. be particularly destabilizing. As shown in the Such a center, or centers (one in each country notional scenario in chapter 4 where strategic but linked and having joint manning), would surprise occurs at high force postures, the reac- be similar to the nuclear risk reduction cen- tion of the side that is surprised can range ters set up by the United States and the Soviet from ratcheting up the arms race yet another Union near the end of the Cold War but would notch to sudden preemptive action to counter handle conventional as well as nuclear infor- what it perceives to be a disastrous and poten- mation exchanges and monitoring tasks. Again, tially irreversible shift in the strategic balance sustained U.S. interest in such a venture com- of power. It is impossible by defi nition to antici- bined with a willingness to act as a lead fund pate strategic surprise, so policy remedies are raiser and provider of technology are key to mak- limited. However, it is clear in mapping out ing such an approach work and shielding it from possible trajectories of nuclear or strategic arse- the inevitable political strife in the region. This nals on both sides what some particularly is a considerable undertaking politically for the destabilizing acquisitions could look like. As two states and is likely not to be a realistic policy noted in chapter 4, Pakistan has already option until extensive discussions have taken strongly hinted at what it would consider place both bilaterally between the United States destabilizing when it proposed to India a mora- and each country and also between India and torium on the acquisition and deployment of Pakistan. SLBMs or ABM systems. What any policy remedy is looking to Strategic Surprise and High Force Postures achieve in this case is both to reduce the prob- At almost any level of force posture, it is pos- ability of strategic surprise and to tone down sible for strategic surprise to occur. In the any reactions to it. For the United States,

Conclusions and Recommendations 113 Recommendations for High Force Posture/Strategic Surprise India Pakistan · Politcal-Military Dialogue · Political-Military Dialogue Themes Themes - Same as lower force structure levels plus - Same as lower force structure levels plus transparency on future plans transparency on future plans - Offense/defense dialogue with emphasis on - Education on stabilizing qualities of certain missile defense capabilities · Regular India-Pakistan strategic discussions - Offense/defense dialogue with emphasis on missile defense · Provision of U.S. strategic early warning of major changes in the military balance · Regular India-Pakistan strategic discussions · Provision of U.S. strategic early warning of major changes in the military balance

this means promoting transparency between United States should be looking to debunk vari- India and Pakistan about their nuclear futures. ous concerns about certain types of capabili- Again, this should be done fi rst in bilateral dia- ties. For example, it is possible that acquisition of logues between Washington and either New SLBMs or missile defense systems on both sides Delhi or Islamabad. Then, the United States (depending on the confi guration and coverage of should push for serious dialogues at multiple the latter) could indeed be stabilizing in the long levels between the two antagonists. The goal run, reducing the chances for war in general is to preempt surprise so political pressures and for nuclear escalation should conventional for destabilizing responses can be countered. war break out. The issue in this type of sce- Ideally, if one side were to express a strong nario is how these capabilities are acquired and enough reservation about a particular capabil- whether the other side can accept them as stabi- ity, the other side would consider foregoing it lizing. Part of this discussion (between the Unit- or would consider adopting other measures to ed States and each of the two countries) should reduce concerns about its unsettling effect on be a series of exchanges on missile defense and the strategic balance. Of course, this will be the ways in which it may enable the two states diffi cult with India given that its nuclear force to pursue stability with relatively low levels of structure and doctrine are not just aimed at offensive weapons. Finally, the United States Pakistan but also have potential utility against should consider adopting a policy of warning China and are a tool for enhancing New Delhi’s India and Pakistan of possible changes in the world status more generally. strategic balance if the United States picks up In terms of toning down possible reactions intelligence that they are going to occur and if to certain acquisitions, the United States should Washington believes that the surprise could lead use its extensive knowledge of nuclear weapons to a particularly destabilizing reaction. and changes in strategic balances in discussions with both India and Pakistan about how per- Implementation Strategy ceived shifts can be remedied in ways that are Many of the recommendations above center on not themselves destabilizing. In addition, the dialogue and risk-reduction measures that have

114 Nuclear Stability in South Asia either been proposed but never adopted or have desire for a new policy approach and the accel- been formally adopted but have failed in their eration that occurred post-September 11, there implementation. Talk is cheap and CBMs are are limits to how fast and how far changes more often than not used as confi dence-busting can go. Many of the suggestions made above measures on the subcontinent. Therefore how require signifi cant detail work before they can the United States goes about pursuing these be implemented. This analytical work takes various initiatives is critical to the chances for time, and fi nding the necessary resources to their success. Five specifi c issues have to be implement certain policy decisions also takes addressed in implementing such a strategy to time and requires building political consensus improve chances for success. First, despite in the interagency, with Congress, and with the urgency of the issue – nuclear crisis sta- key international partners (see below). In par- bility – the United States is going to have to ticular, the focus on the war on terrorism move ahead slowly, building a consensus with- will drain ideas, resources, and manpower that in the U.S. interagency, with Congress, with will be necessary to develop a long-range and key international partners, and with India and comprehensive policy to address the nuclear Pakistan on how to address the question. Sec- stability part of new U.S. policies toward South ond, the United States must tailor its approach Asia. Moreover, the United States still has to India and Pakistan and recognize that the much work to do in turning around the nature two approaches will not be the same, because of its cooperation and dialogue with both Paki- of history, geography, political structure, and stan and India. A greater degree of trust has doctrine and force structure issues. Third, the to be built with each. While Islamabad is no United States must work to create and sustain doubt thrilled with the positive attention it has political, military, and economic leverage on been receiving from Washington after Septem- these issues with each country. Fourth, the ber 11, it retains the residual fear that it will be U.S. interagency must be organized around a abandoned by the United States yet again, as it new comprehensive plan for engagement with was in 1990, when its cooperation on Afghani- these two countries. Finally, the executive stan is no longer seen as being so urgent. At branch must construct a public diplomacy strat- the outset of the Bush administration, India egy that will help build broad support for was hoping to benefi t from a new and broad its new policies in Congress and with key relationship with Washington. While still hop- international partners. Each of these issues is ing to develop such a relationship, New Delhi examined in detail below. is more cautious about Washington because of what it sees as an overly eager courting of Paki- A Phased Approach stan as a partner in the war on terrorism. New Before September 11, the Bush administration Delhi continues to point out that it is the victim was signaling that it was going to take a signifi - of terrorism in Kashmir and that the Pakistani cantly different tack on policies toward South government is its sponsor. If, after the war in Asia and the nuclear question than the Clinton Afghanistan winds down, the United States administration. This change has been acceler- does not engage on this core issue for India, it ated because of the September attacks and the may prove diffi cult to build habits of consulta- subsequent war in Afghanistan. Despite this tion and cooperation with New Delhi.

Conclusions and Recommendations 115 Therefore new approaches to stability severely limit what the U.S. could accomplish measures will have to be taken slowly with dia- with either country. logue preceding ad hoc cooperation and ad hoc Similarly, the implementation of the mea- cooperation coming before institutionalized sures will require different approaches. For a or systematic measures. In addition, measures start, the interlocutors on these subjects will that require cooperation (and therefore some likely be different in each country. In Pakistan, degree of trust) between India and Pakistan civilian political leaders will have to be engaged will have to be introduced slowly and in once they are restored to power, but the focus sync with improvements in overall relations will be on the military as they will have the between the two sides. The Agra summit pro- predominant role in making national security cess provides some hope in this regard, but policy for the foreseeable future. In India, at it remains uncertain if it will stay on track. least in the short to medium term, the civil- Much has been thrown off by the upheaval in ian leadership is the primary target audience the region in the fall of 2001, and it is unclear for dialogue and exchange, and the military is whether and how the dialogue will get back secondary. In both countries the technical sci- on track. Finally, some of the more institution- entifi c community must be engaged as well, but alized measures, such as risk reduction/crisis again the channel through which the United management centers, will require discussion, States must go to link up with that group will design, funding, construction, and training for be different. staff, so a phased approach will be required. Creating and Sustaining Leverage Tailoring Strategies and Approaches A strategy focused on improving stability, par- As made evident throughout this study, the ticularly crisis stability, on the subcontinent nuclear doctrines and force postures of India requires the cooperation of India and Pakistan. and Pakistan have signifi cant differences As noted earlier, it is not necessarily clear that requiring varying measures and approaches. either side sees this as an urgent issue. New For example, Pakistan’s fi rst-strike doctrine Delhi may view discussion of these issues with makes a discussion and possible exchange of Washington, and indeed even with Islamabad, information or provision of technology on com- as a necessary evil while not being interested in mand and control more central than with pursuing any of the remedies seriously because India, where a more leisurely second-strike it simply views it as a problem. Islamabad, on posture and doctrine mean that simpler mea- the other hand, may be very eager to engage sures may suffi ce. What is offered to one on these issues as a way to ensure staying in country, either in terms of topics for discus- Washington’s good graces, but like New Delhi, sion or technical assistance, does not have to may have very little actual interest in pursuing be identical to what is offered to the other. stability measures seriously. Moreover, domes- For political reasons, it will have to be per- tic politics in both states may limit the realm ceived as having a rough equity in terms of of the possible in publicly visible cooperative the level of the dialogue and the value of what measures. Given this context, how is Wash- is being offered, but exact parallelism would ington going to create and sustain leverage on these issues, particularly given that it will be

116 Nuclear Stability in South Asia engaging with these countries on various other leverage and infl uence may apply in the Pak- issues at the same time (such as terrorism, istani situation if Islamabad begins to view trade, intellectual property rights, and democ- itself as an essential ally for Washington in the racy)? war on terrorism. Over time, as the relation- The Bush administration has already deter- ship grows, U.S. leverage on India may increase mined that sanctions, at least of the types that if only because more linkages will have been were applied to India and Pakistan after the established. nuclear tests, were not useful. Indeed more Not all sanctions possibilities should be generally the Bush administration is doubtful eliminated and indeed congressional interests that sanctions will produce many useful results or legal restrictions may prevent the formal at all. Indeed, in the area of nuclear stability, ruling out of some. For example, the nonpro- cooperation of the parties in question is essen- liferation sanctions will not be taken off the tial, so incentive packages are more sensible books, but the president will be able to waive than sanctions. These incentives should be tai- them on a regular basis, as he did in the fall lored to the country and issue in question and of 2001, as long as Congress is kept on board should be explicitly linked to actual adoption of with the policy. Once the immediate issue of certain measures to ensure implementation – a Afghanistan as a safe haven for terrorists is perennial problem in the region. For example, a ameliorated, Congress may begin to look hard- resumption and expansion of the Internation- er at waivers on such items absent progress al Military Education and Training program on issues such as nuclear safety, security, and (IMET) for Pakistan should include a require- stability. It may be useful to establish certain ment that a certain percentage of the courses clear red lines for the re-imposition of current or participants focus on nuclear safety or secu- sanctions or the development of new sanctions. rity issues or on crisis management. Similarly, These red lines would include the transfer of expanded military supply relationships with nuclear technology to a third state or other either side should hinge on at least some acqui- third party. Another red line would be the sition of technologies that enhance command explicit threat of nuclear use by either India or and control, crisis management, information Pakistan. distribution, or other elements that could con- The fi nal element of leverage is simply the tribute to stability in peacetime and in future political level and frequency of engagement. crises. As noted early in this study, U.S. engagement It should be recognized that the leverage in South Asia has been sporadic at best, and that the United States will have with the two high-level, sustained engagement has been non- states will be limited and uneven. The United existent. Regardless of the political system in States is likely to have more infl uence on Paki- place in India or Pakistan, much of the ques- stan, at least in the short run, as Islamabad tion of cooperation with the United States is has greater need for sustained, improved rela- political. For Pakistan, this means being seen tions with the United States and Washington to have connections to a great power that can is in a position to limit much-needed funds balance India. For India, it means being seen for Pakistan from the international fi nancial as being accepted in the circles of the elite in institutions. As noted above, however, reverse international politics. The United States must

Conclusions and Recommendations 117 capitalize on these facts of politics and ensure veneer of nuclear stability as an overlay and that both India and Pakistan are much more justifying rationale. regular destinations for high-level U.S. delega- Therefore, a new policy toward India and tions, that they are brought into our counsels Pakistan that emphasizes nuclear stability will on a variety of issues more frequently, and have to be led by a strong interagency process that they are subject to more attention than that provides clear guidance as to goals and they have been heretofore (other than, perhaps, objectives and regularly evaluates progress immediately after their nuclear tests). High- toward these goals. One way to ensure that level engagement with Pakistan has become the such a process reaches throughout various norm since September 11, but the issue will be parts of the bureaucracy is to conduct a com- regularizing it, sustaining it, and refocusing it prehensive review of all elements of policy – once the immediacy of issues in Afghanistan related to the two countries. Existing policies subsides – on nuclear stability issues. should be updated to refl ect top-level guid- ance as to the new U.S. strategic goals for the region and for each country individually. Good Organizing the U.S. Interagency examples of this are the various export control Because the United States does not have a guidelines and limitations on military con- long-standing and set policy toward the South tacts and interactions that exist for both India Asia region, it has historically not been well and Pakistan. While some of these restrictions organized to implement policies in this region. are the result of nuclear testing-related sanc- During times of no crisis when interest was tions legislation that may be loosened or lifted low, policy was driven primarily by the region- entirely, other restrictions and guidelines are al components of various elements of the a continuation of policies left over from the bureaucracy – the State Department, the time when the United States was not seeking to National Security Council (NSC) staff, and expand its relationship with India or when the the Offi ce of the Secretary of Defense (OSD). United States was seeking to keep both coun- When crises arose or certain functional issues tries from acquiring nuclear capabilities. What were deemed to be paramount (such as non- is needed is a sunset review of these restric- proliferation or counterterrorism) then policy tions to ensure that new initiatives are not making either rose to higher organizational stymied by old rules that no longer conform levels – usually with a broader view and to new overarching policy guidance or real- responsibility – or was handed over to the ities on the ground. The timing for such a functional entity responsible for the issue in review should be ideal in the wake of an even- question (for example, the nonproliferation tual wrapping up of the military phase of the directorate at the NSC or the counterterror- war on terrorism in Afghanistan. The United ism coordinator at the State Department). Left States will then be shifting to a longer-term pol- to their own devices with a new policy that icy of engagement in the region, with the goal focused on nuclear stability and that promoted of ensuring no resurgence of terrorist activity greater engagement, the elements of the nation- at the top of its agenda. A more comprehensive al security-making bureaucracy would each look at how best to shift policy gears from crisis pursue their own parochial interests with a to long-term management should include issues

118 Nuclear Stability in South Asia related directly to nuclear stability as well as additional resources and/or signifi cant backup terrorism. in Washington. Second, for India, the United The policy process should be under the States Pacifi c Command (PACOM) provides a direction of the National Security Council good start for engagement, but India’s civilian- staff, both to ensure that parochial interests dominated national security structure means of any of Washington’s bureaucratic players do that the civilian side of OSD, State, and the not dominate but also to show a high level of NSC must take a much larger role. For nuclear political interest. This in turn will assist in get- stability policy questions in India, currently, ting the best possible responses from both India the military simply is not the right address for and Pakistan. Under NSC direction, delegation dialogue. This of course does not mean that of implementing authority will have to vary the military should be ignored. They may mat- for Pakistan and India and specifi c guidance ter more in the future if India moves up the will have to be provided as various agencies ladder to a more fully deployed nuclear pos- and experts are engaged. One valid concern ture, and of course PACOM has other reasons expressed by nonproliferation supporters is to engage more fully with India. Finally, those that an engagement policy will let loose vari- other interests of CINCs may not necessarily ous elements of the bureaucracy like unguided be related to nuclear stability. They may be missiles. Each will engage, provide untailored interests related to warfi ghting considerations, set-piece briefi ngs, and advance its parochial counterterrorism, or other operational con- interests without necessarily advancing the cerns (port calls, access to facilities). The central policy interests. Again, specifi c guid- Washington-based bureaucracy will simply ance will have to be set out and regular reviews have to keep this in mind and regularly mon- conducted of programs to ensure that they are itor CINC-led activities to ensure that they advancing the higher-order goals set by the support both the CINC’s warfi ghting respon- president. sibilities and the broader policies on nuclear The Defense Department in particular stability set by the interagency. will have to adjust its usual temptation to del- egate most engagement activities to the regional Public Diplomacy – Building Support CINCs for three reasons. First, issues related in Congress and Internationally to nuclear stability in Pakistan, which is in In order to generate and sustain support for a the United States Central Command’s (CENT- policy shift that focuses on nuclear stability, COM) area of responsibility, may simply be too the U.S. administration will have to engage in much to add to the already full plate of the a fairly vigorous public diplomacy campaign, current CENTCOM commander. While inten- focused on the United States and key allies tions may be good and history has shown that and supporters. Congressional support will be intervention by the CENTCOM commander necessary for any legislative changes that the can be useful, the fairly full list of operational administration deems necessary. While most responsibilities in the war on terrorism and in sanctions legislation related to India and Paki- the Gulf means that CENTCOM simply may stan now has national security waiver clauses, not be up to taking on too many new responsi- if the administration fails to provide an ongo- bilities for nuclear stability in Pakistan without ing and cogent rationale for its new policy

Conclusions and Recommendations 119 as well as evidence that it is working, those will be seeking to modify the NPT to add new waiver clauses could be rescinded or Congress nuclear weapons states. Finally, the United could fi nd other ways to stymie policy. More- States will engage more vigorously in working over, congressional support will be necessary with the Indian and Pakistani governments to to fund new initiatives such as a risk- reduction remedy the conditions that caused them to seek center or the restarting of old standbys of for- nuclear weapons in the fi rst place. eign and national security policy such as IMET for Pakistan. Finally, congressional approval is Conclusions required for any signifi cant military sale, some- Nuclear weapons are not going to disappear thing that may eventually be desired for both from the subcontinent any time soon. India, Pakistan and India. Pakistan, China, the United States, and many On the international front, the United other states will have to adjust to this reality States will have to engage in some careful and adapt their national security policies to explaining to various friends and allies, par- it. For the United States this means focusing ticularly those who are ardent nonproliferation on what it can do, given this situation, to best supporters. On the one hand, the United States advance its national security interests. Such simply wants to avoid diplomatic fallout that a focus requires a re-balancing of priorities makes it more diffi cult to advance its policies. to favor enhancing crisis stability over global On the other, it may also need fi nancial or polit- nonproliferation. Actively pursuing policies ical support for some of its initiatives from key to enhance crisis stability in the region will foreign governments such as Japan, the UK, require a sustained and greatly deepened and/or France. Understanding about engage- engagement with both India and Pakistan on ment with Pakistan is high in the wake of a range of issues. It should be remembered, September 11, but that level of agreement may however, that despite a willingness, or even an ebb over time if nuclear and democracy issues eagerness, on the part of the United States to are not eventually addressed. re-engage with these two countries and to offer The key themes in such a campaign could all of the assistance it can to enhance stability be as follows. First, in addition to the obvious on the subcontinent, ultimately it is up to Paki- issue of terrorism, the United States is refocus- stan and India to establish a modus vivendi ing on the most immediate nuclear weapons that lowers the chances for a nuclear confron- dangers in South Asia – the chance that those tation. The United States, as the fi rst nuclear weapons may be used either accidentally or power, and as the world’s sole superpower, inadvertently during peacetime or a crisis. It however, has both the interest and responsibil- is seeking to engage the governments of India ity to do all that it can to assist them in reaching and Pakistan to reduce those chances. To this this situation. end, the United States is going to use all of its decades of experience with these issues and, where appropriate, share that knowledge with India and Pakistan. Second, in no way does this refocusing mean that the United States has changed its policy on nonproliferation or

120 Nuclear Stability in South Asia Endnotes for Chapter Five

1 Douglas Frantz, “U.S. and Pakistan Discuss Nuclear Security,” New York Times October 1, 2001. 2 Bradd C. Hayes, “International Game ’99: Crisis in South Asia, January 20-28, 1999,” Research Report Decision Support and Strategic Resource Departments of the Center for Naval Warfare Studies, United States Naval War College. 3 As noted in chapter 2, however, more and faster are better from the Pakistani perspective in terms of stability, and to some degree an ability to keep forces at a higher level of alert may actually raise the Pakistani threshold for use. 4 The most recent offi cial agreement on CBMs is the Memorandum of Understanding signed in Lahore in February 1999, which commits the countries to a range of dialogues and enhancements on existing CBMs. Because of the 1999 Kargil crisis, none of these measures has been implemented. Text of the Lahore documents available at http://www.stimson.org/cbm/sa/drs/lahore.htm. 5 The concept has been raised in the U.S. nongovernmental community, but it has yet to be adopted by the U.S. government. John H. Hawes and Teresita C. Schaffer, “Risk Reduction in South Asia: A Role for Cooperative Aerial Observation?” Nuclear Risk-Reduction Measures in Southern Asia, Stimson Center Report no. 26 (November 1998), http://www.stimson.org/pubs/cbm/cbmgen/saaerial.pdf. 6 Kent L. Biringer “Missile Reduction and Monitoring in South Asia,” Cooperative Monitoring Center, Sandia National Laboratories, June 2001. This and other papers on the subject available at http://www.cmc.sandia.gov/Links/regions/regions-mainframe.htm.

Conclusions and Recommendations 121 THE AUTHORS

Andrew C. Winner is a senior staff member at the Institute for Foreign Policy Analysis. A ten- year veteran of the U.S. Department of State, he has worked on political-military issues, crisis stability, and confi dence-building measures and arms control in a variety of regional settings. Mr. Winner holds an A.B. from Hamilton College and an M.A. from The Johns Hopkins University’s School of Advanced International Studies, and is completing a doctorate at the University of Maryland, College Park. His dissertation is on coalitions and coercion.

Toshi Yoshihara is a research fellow at the Institute for Foreign Policy Analysis. Mr. Yoshihara holds a B.S. from Georgetown University’s School of Foreign Service, and an M.A. from The Johns Hopkins University’s Paul H. Nitze School of Advanced International Studies. He is a Ph.D. student at The Fletcher School of Law and Diplomacy, Tufts University. His dissertation research focus is Chinese military innovation and information warfare.

122 Nuclear Stability in South Asia