India's Nuclear Force Structure

Gurmeet Kanwal, Senior Fellow, IDSA

 

Americans managed to convince themselves that thousands of strategic warheads and multiple means of delivering them were needed in order to deter the Soviet Union. If, however, one thinks politically instead of militarily, it becomes apparent that not much is needed to deter. What political leader would run the risk of losing even a city or two—and also his position of power—in military pursuit of problematic gains?

— Kenneth Waltz1

 

Background

Though the dust has settled over Pokhran and the Chagai Hills since the nuclear blasts of May 1998, their aftershocks are still reverberating round the world. The crossing of the nuclear Rubicon had propelled India and Pakistan into de facto membership of the most exclusive club in the world—that formed by the five Nuclear Weapons States (NWS). The NWS, of course, still refuse to accept the two nuclear upstarts as members. Paradoxical as it may seem, India has acquired nuclear weapons in order to renounce them. India's present status as a de facto nuclear power is already forcing the P-5 and the G-8 to take India's principled opposition to the discriminatory NPT, CTBT and the proposed FMCT seriously and to heed India's long-standing and consistent position that the P-5 must agree to implement a time bound programme for nuclear disarmament. In a world driven by Realpolitik, a soft voice is heard with respect only when it is backed by a big stick. However, at this stage it is more important to fine tune India's nuclear strategy and doctrine and draw up a balanced and overtly visible nuclear force structure with a viable command and control system in order to ensure that India's professed minimum deterrent is genuinely credible.

Above all else, India's nuclear force structure should flow out of the nuclear doctrine and the national security strategy. The nuclear force must be based on India's declared strategy of 'no first use' and should be capable of 'punitive retaliation' to inflict unacceptable damage for a nuclear strike on Indian civilians, soldiers and civil and military assets. Like any other military force structure, a nuclear force structure must be dynamic and flexible enough to evolve suitably in synchrony with the technology trajectory as, more than any other military force, nuclear weapons and their delivery systems are heavily dependent on emerging technologies. India's nuclear force structure should also be adaptable to changes in threat perceptions and be adjustable enough to give effect to bilateral or multilateral treaties that may be negotiated in future. Also, should the P-5 and other nuclear weapons states (NWS) be amenable to gradually moving towards the total elimination of nuclear weapons a few decades hence, India's nuclear force should be so structured that it permits easy dismantling. Total universal nuclear disarmament remains a key objective of India's foreign policy.

This article is a sequel to and based on the background work already published by the author as under:

l Command and Control of — Strategic Analysis, January 2000 India's Nuclear Weapons (Paper presented at IDSA Fellows Seminar on November 20, 1999.)

l Implementation of India's No — Strategic Analysis, April 2000. First Use Doctrine: (Commentary)

Need for some Inescapable Qualifications

l Does India Need Tactical — Strategic Analysis, May 2000. Nuclear Weapons?

l Nuclear Targeting Philosophy — Strategic Analysis, June 2000. for India

The salient deductions derived in the above mentioned articles are summarised below and form the backdrop for this article:

l Nuclear weapons are political weapons and not weapons of 'war fighting'. Their sole purpose is to deter the use and the threat of use of nuclear weapons.

l For India's doctrine of minimum deterrence and 'no first use' nuclear strategy to be credible, India's targeting philosophy must be based on a counter value strategy of massive punitive retaliation to inflict unacceptable damage against the adversary's population centres and industrial assets. The retaliatory strike should be massive regardless of the level (quantum, yield, type of target, location) of a first strike against India and its armed forces.

l Since India's targeting philosophy is not premised on 'proportionate deterrence' or 'flexible response', India does not need tactical or battlefield nuclear weapons. Also, the inherent disadvantages of tactical nuclear weapons (primarily, the lower threshold of use, the need for 'launch on warning' and 'launch through attack' strategies, complex command and control challenges, increased cost of manufacture and maintenance, the problems of storage, transportation and handling in the field and the greater risk of accidental and even unauthorised use) should preclude the use of these weapons for deterrence.

l A National Command Authority headed by the Prime Minister should be established for the command and control of India's nuclear weapons. A clear chain of succession should be laid down.

l Within the military, the Chief of Defence Staff (CDS—a new overall Commander-in-Chief who would provide 'single point military advice' to the Government) should be assisted by a tri-Service joint planning staff for threat assessment and the formulation and execution of a joint military strategy, including nuclear strategy.

l A tri-Service Strategic Forces Command should be raised under the CDS to exercise functional control over the nuclear weapons and to oversee the functioning of the surveillance, early warning, nuclear forces intelligence, targeting and attack and damage assessment systems.

l The actual nuclear force units, forming India's 'triad' of land-, sea- and air-based forces, should form part of the Army (all land-based ballistic and cruise missiles), the Navy (SSBNs with SLBMs) and the Air Force (strategic bombers with air-delivered nuclear bombs).

Is Minimum Deterrence a Numbers Game?

Writing in the early-1980s, Bhabani Sen Gupta had said that the entire basis for nuclear weapons is deterrence:2 "The entire purpose is to deter the enemy, not to fight him… the very existence (of nuclear weapons) is justified on a theoretical base that is gravely limited at best, and outright wrong at worst… it would be better for India to settle the doctrinal issues before going nuclear, instead of first going nuclear and then looking for doctrinal justification." However, India's nuclear policy evolved without major debate on the doctrinal issues and the nuclear weapons research and development programme was shrouded in secrecy. It is only after the Pokhran-II nuclear tests that Indian analysts have begun to wrestle with the complexities of nuclear theology and most of the home truths have had to be re-learnt. It is a universally accepted truism that deterrence is ultimately a mind game. It needs to be achieved during peace to ensure against the use or threat of use of nuclear weapons by one's adversaries and for the purposes of coercive diplomacy. The concept of nuclear deterrence first evolved in a US Joint Chiefs of Staff memorandum where it was stated that the "threat of the use of atomic bomb would be a great deterrent to any aggressors, which might be considering embarking on atomic war."3

Should deterrence fail and a nuclear strike be launched against India, the Indian nuclear force should be able to survive in adequate numbers to respond in a punitive manner to inflict unacceptable damage on the adversary who initiated the nuclear exchange. The key question is the determination of what the adversary considers unacceptable damage. If India's adversary for a retaliatory nuclear strike was a developed Western nation, for example one of the original members of NATO, it would be safe to assume that such a nation would be deterred if it was certain that nuclear warheads in single digit numbers could be effectively delivered to destroy a few of its cities. As McGeorge Bundy said:4

There is an enormous gulf between what political leaders think about nuclear weapons and what is assumed in complex calculations of relative advantage in simulated strategic warfare. Think tank analysts can set levels of acceptable damage well up in the tens of millions of lives. They can assume that the loss of dozens of great cities is a real choice for a sane man. They are in an unreal world. In the real world of real political thinkers… a decision that would bring even one hydrogen bomb over one city of one's own country would be recognised in advance as a catastrophic blunder; ten bombs on ten cities would be a disaster beyond history; and a hundred bombs on a hundred cities are unthinkable.

However, what the Pakistanis consider acceptable damage is likely to be several orders of magnitude greater than what, for example, the Americans might consider acceptable. And, what the Chinese consider acceptable is likely to be much more than what may be acceptable to the Pakistanis. It would also be far more than what the Indian Government may consider acceptable. "In his conversation with Jawaharlal Nehru, Chairman Mao Dze Dung told him that even if 300 million Chinese perished in a nuclear war, the remaining 300 million Chinese would build a new glorious civilisation."5 It is well appreciated that totalitarian regimes tend to be less mindful of the value of the lives of their citizens than democracies. On the other hand, these perceptions vary with the level of development and "What the Chinese considered acceptable damage in 1950 would be way above what they consider acceptable today. With greater affluence and more democracy, the level of acceptable damage will go down further. Even today, India and Pakistan are softer states in this regard than China."6

During the Kennedy era, the Soviet arsenal was estimated to be only one-seventeenth of the US stockpile.7 Yet, it is well known that during the Cuban missile crisis the US was deterred because the Chiefs of Staff could not assure the government that a few Soviet warheads would not hit American cities even if the US launched a massive disarming first strike. The enduring lesson of the Cuban missile crisis is that even gross asymmetry in the number of nuclear warheads in one's adversary's arsenal provides no guarantee that the adversary would not be deterred if he was convinced that even a few warheads would get through and cause unacceptable damage. K. Subrahmanyam has written in the Indo-Pak-Chinese context:8

"When we talk of deterrence between Pakistan and India, is Kashmir worth the loss of Lahore for the Pakistanis? Even if they are in a position to hit Delhi, will that compensate for the loss of Lahore and Karachi? Surely not. Will the Chinese risk Kunming and Chengdu at present and even Shanghai and Guangzu later – when India has an operational Agni —for any conceivable political, military and strategic objective?…

"Minimum deterrence is not a numerical definition but a strategic approach. If a country is in a position to have a survivable arsenal, which is capable of exacting an unacceptable penalty in retaliation, it has minimum deterrence as opposed to an open-ended one aimed at matching the adversary's arsenal in numerical terms. Those arsenals in thousands were produced in an era when the strategic establishments believed in nuclear war fighting and did not understand its ecological consequences. Today, sections of the US strategic community argue that the US can discharge its global responsibilities with an arsenal of 200 warheads."

China has approximately 20 ICBMs to over 1500 of the US. However, it is nobody's case that the US is not deterred by the Chinese arsenal. Had China's ICBMs not deterred the US, it would not have been so vigorously engaged in developing a national missile defence (NMD). It clearly emerges that the nuclear force levels necessary for a retaliatory strike are independent of the quantum of the adversary's nuclear force and depend only on the numbers that are needed to inflict unacceptable damage. The side that can cause greater damage does not necessarily achieve greater deterrence. Quite obviously, the required number of nuclear warheads and their delivery systems must survive a first strike and there should be adequate redundancy. Hence, for a retaliatory strategy, attempts at maintaining a numerical parity with the adversary are neither necessary nor desirable and, as a corollary, there is no substance in the bogey of a nuclear arms race on the Indian Subcontinent. Kenneth Waltz has written:9 "Those who foresee intense arms racing among new nuclear states, fail to make the distinction between war fighting and war deterring capabilities. Forces designed for war fighting have to be compared with each other. Forces designed for war deterring need not be compared. The question is not whether one country has less than another, but whether it can do unacceptable damage to another, with unacceptable damage sensibly defined." Among other senior government functionaries, Brajesh Mishra, the National Security Advisor and the Principal Secretary to the Prime Minister, has emphatically stated that India has no plans to enter into an arms race with China.10

Nuclear Capabilities of India and its Neighbours

Due to the thick veil of secrecy surrounding the nuclear weapons research and development programmes of almost all the nuclear weapons states (NWS), it is extremely difficult for an analyst to estimate the number of nuclear warheads possessed by a NWS with any degree of precision. Estimates are generally based on the quantity of un-safeguarded enriched uranium or plutonium that the NWS may have accumulated over the years. Since these figures are only rough estimates, they are bound to be unreliable. Also, there is considerable difficulty in trying to compute the total yield of fission and thermonuclear warheads that a NWS may possess because, for the same amount of fissionable material, it is possible to produce almost the same number of thermonuclear bombs but with much greater yields than pure fission or even boosted fission bombs. However, it is possible to arrive at rough guesstimates with a similar margin of error by using the same method of calculation. Such estimates should suffice to serve the purpose of computing relative nuclear capabilities.

India

In an interview after the Pokhran-II nuclear tests of May 1998, Dr. Rajagopala Chidambaram, Chairman of India's Atomic Energy Commission (AEC), said that India has the capacity to design and make nuclear weapons of any size or type, including neutron bombs.11 Bharat Karnad has reported that "according to the data collected by the US Congressional Research Service, by the year 2000, India will have an annual production rate of 127 kg of un-safeguarded fissile material and an accrued total of some 1,607 kg of the same, which is sufficient to fashion 400 warheads… were all this material to be turned into fission (nuclear) weapons, the cumulative yield of the prospective Indian arsenal by the end of the century would be 3,095 Kt or a little over three megatons, while the total destructive power if this amount of material were exclusively fusioned, would be over three times as much at… 10.317 megatons of TNT."12 In a report entitled "Repairing the Regime: Preventing the Spread of Weapons of Mass Destruction," the Institute for Science and International Security (ISIS), Washington, has estimated that the median value (which is midway between the smallest and the largest estimates) of India's stockpile of weapons-grade plutonium (in which the concentration of Plutonium-239 isotope is greater than 93 percent) was about 290 kg at the end of 1998.13 "The report claims that India's inventory of weapons-grade plutonium is derived by estimating total production in its reactors and by subtracting 'drawdowns' from nuclear testing, processing losses and civil uses." Since 290 kg is a median value, the fissile material stockpile could actually range from 200 to 400 kg. Earlier, in July 1998, Dr. David Albright of ISIS had estimated that India's stockpile of weapons-grade plutonium was around 370 kg and was enough to make about 74 nuclear weapons.14 These estimates are at variance with the figures quoted by Bharat Karnad.

Table 1. Nuclear Tests at a Glance

US Russia France UK China India

Number of Tests 1,032 715 210 45 45 6

First Explosion 1945 1946 1961 1952 1964 1974

Last Explosion 1992 1990 1996 1991 1996 1998

Nuclear Warheads 12,070* 22,500* 500* 380* 450* 65*

* Estimates.

(Source: The Bulletin of Atomic Scientists.)15

A report by W. P. S. Sidhu in the Jane's Intelligence Review soon after the May 1998 nuclear tests estimated that India's nuclear stockpile contained between 20 to 60 warheads assembled from the weapons-grade plutonium re-processed from the fuel taken from the research reactors located at Trombay.16 However, according to Sidhu, if the plutonium produced in India's commercial reactors is also taken into account, India would possess adequate fissile material to produce "at least 390 nuclear weapons and as many as 470 weapons."17 R. Ramachandran writes that it does not make sense to use reactor-grade plutonium "which has a high content of spontaneously fissionable Pu-240 and makes only 'dirty' bombs as against weapons-grade Pu-239 from research reactors."18 He has calculated that India is likely to have adequate stocks of plutonium for about 30 bombs and that "A good upper band would… be 35."19 It also needs to be noted that India's fast breeder programme requires reactor-grade plutonium and if it were to be used for making nuclear warheads, it would not be available for the purpose for which it is actually intended.20 Estimates of the nuclear stockpile in the Indian media have ranged from 25-6521 warheads to 50-6422 warheads.

George Perkovich has written, "knowledgeable Indian sources confirmed" that reactor-grade plutonium was used in a low-yield device tested on May 11, 1998 at Pokhran.23 He believes that India's stockpile of weapons-grade plutonium was "probably around 250 to 300 kg by 1998… this stockpile could 'fuel' perhaps fifty weapons." On the other hand, if approximately 600 kg of reactor-grade plutonium was to be added to the stockpile and actual tests in May 1998 had confirmed that the difficulties associated with designing nuclear warheads using such fissile material had been overcome, India's capability to produce and stockpile nuclear warheads would be considerably enhanced and so would the quantitative content of its deterrent. Given the lack of transparency in these matters, it will be a long time before this controversy is resolved. The Kargil Review Committee has recommended that a White Paper should be published by the government on India's nuclear weapons programme.24 Such a step would be a positive one.

Pakistan

After its stunning military defeat at India's hands in 1971, Pakistan's Prime Minister, Zulfiqar Ali Bhutto had said that the Pakistanis would eat grass if need be but would spare no effort to produce an 'Islamic' (nuclear) bomb. Pakistan started producing weapons-grade uranium at the Kahuta Research Laboratories near Islamabad some time between 1980 to 1985. During the mid-1980s, Pakistan is now known to have conducted 20 'cold' tests. In 1987, Richard Barlow, the CIA operative in Islamabad, reported to Washington that Pakistan had assembled a nuclear weapon. Pakistan acquired this capability with generous help from the Chinese, who found in Pakistan a strategic ally willing to countervail India. By now it is well established that there was nothing Pakistani about the Islamic bombs detonated in Chagai in May 1998. "The Chinese help to Pakistan in supplying ring magnets (5,000—for gas centrifuges), the Khushab plutonium production reactor and setting up a missile factory, is now part of history."25 The Heritage Foundation, a Washington based think tank that used to wield considerable influence during the Reagan and Bush presidencies, has chronicled how the Chinese helped Pakistan to build its nuclear bomb and delivery systems. "China's role in helping Pakistan to acquire nuclear weapons has raised serious concerns about China's part in fostering instability in South Asia."26

China is also known to have provided a complete nuclear weapon design to Pakistan along with sufficient weapons-grade uranium for two tests, established a special industrial furnace at the Khushab facility to produce plutonium, transferred enough tritium gas for triggers for ten nuclear weapons, trained Pakistani technicians and guided Pakistani scientists in propellant and warhead technologies.27 It is also generally believed that China permitted Pakistan to carry out a nuclear test at its Lop Nor range. During the early to mid-1990s, China transferred fully assembled M-11 missiles to Pakistan. In February 2000, the CIA reported to the US Congress that despite its promise to stop ballistic missile and nuclear assistance to Pakistan in May 1996, China continued such assistance during the first half of 1999.28 The CIA also reported that Pakistan's flight test of the Ghauri missile in April 1998, was based on North Korea's No Dong missile.29 Earlier US reports had attributed the Ghauri design to China.

Pakistan's nuclear scientists were apparently quite satisfied with the May 1998 nuclear tests and Dr. Abdul Qadeer Khan even boasted that they could test a thermonuclear bomb "within days" if the political leadership wanted them to do so.30 The Pakistani scientists believe that the country needs between 60 to 70 nuclear warheads to achieve credible deterrent capability against India. Pakistan has between 350 to 400 kg of weapons-grade (highly enriched) uranium which is sufficient for about 20 warheads of 15 to 20 Kt yield.31 The ISIS estimated after the Chagai nuclear tests that Pakistan possessed 335 to 400 kg of weapons-grade uranium that is sufficient to assemble 16 to 20 nuclear weapons and that by end-1998, it could increase the production to 600 kg—enough for 30 nuclear weapons. "The ISIS report predicted that in future Pakistan's stock of weapons-grade uranium would grow at an annual rate of about 110 kg, or enough for five weapons per year."32 According to W. P. S. Sidhu, "Pakistan's arsenal is estimated at between six and 12 nuclear weapons."33 In the Indian media, estimates of Pakistan's stockpile have varied from 15 to 2034 warheads to 15 to 2435 warheads. R. Ramachandran has estimated that Pakistan is likely to have "about 10-15 bombs if it had capped (production of highly enriched uranium in 1991, as has been widely reported) and 25-40 bombs if it had not."36 In Brigadier Vijay K. Nair's view, "Pakistan's nuclear potential… could be 1998-99: 2 to 4 first generation, free fall, aircraft delivered nuclear explosive devices, 1999-2001: 10 to 15 nuclear weapons having first generation guidance systems, delivered by aircraft or short-range missiles, 2002-2010: 25 to 35 nuclear weapons delivered by a mix of aircraft, short-range missiles and IRBMs."37

An NBC TV report on June 7, 2000, ostensibly based on information leaked by a US intelligence agency, stated that Pakistan has 25 to 100 nuclear weapons, "five times" India's arsenal.38 It also gave Pakistan a major lead in ballistic missile capability based on Chinese and North Korean designs and the supply of even fully assembled missiles by these two nations and claimed that India was trailing way behind.39 Pakistan, of course, formally denied the assessment. A Foreign Office spokesperson said that "it was removed from reality" and was an "extraordinary assertion".40 Since Pakistan's nuclear weapons are based primarily on highly enriched uranium and each warhead requires a minimum of 15 kg of fissile material (assuming a sophisticated design), with a production capacity between 45 to 120 kg per year since 1986 at Kahuta, Pakistan could have produced a maximum of between 30 to 50 warheads, according to Dr. Sanjay Badri-Maharaj.41 He estimates the Indian capability at 40-plus warheads from plutonium produced at the CIRUS and Dhruva reactors and affirms that India may have produced enough reactor-grade plutonium for 300-plus weapons at its nuclear power reactors. The NBC report is obviously a motivated report and its accuracy could be discounted, even though the purpose of the concerned US intelligence agency in releasing such information to the media shall remain the subject of debate. R. Prasannan has quoted the Carnegie Endowment for International Peace as having estimated India's stockpile at between 85 to 90 warheads and Pakistan's at 15.42 He quotes the estimate of Gregory S. Jones at 80 warheads for India and 20 for Pakistan.43 It could be concluded from the foregoing analysis of the nuclear capabilities of India and Pakistan that India possibly possesses between 50 to 70 warheads (excluding fissile material from commercial nuclear power plants) and Pakistan is likely to possess between 15 to 30 warheads.

China

China is believed to have developed an arsenal of about 400 to 450 warheads of various types—for land-and sea-based ballistic missiles, for delivery by bomber aircraft and tactical nuclear weapons for tube and rocket artillery and short-range missiles. The Stockholm International Peace Research Institute (SIPRI) assessment of Chinese nuclear forces is given at Table 2:44

Table 2. Chinese Nuclear Forces (January 1999)

Type NATO Range (km) Warheads Warheads in

Designation x yield stockpile

Aircraft

H-6 B-6 3,100 1-3 bombs 120

Q-5 A-5 400 1 bomb 30

Land-based missiles

DF-3A CSS-2 2,800 1 x 3.3 Mt 40

DF-4 CSS-3 4,750 1 x 3.3 Mt 20

DF-5A CSS-4 13,000+ 1 x 4-5 Mt 20

DF-21A CSS-6 1,800 1 x 200-300 Kt 48

SLBMs

Julang-1 CSS-N-3 1,700 1 x 200-300 Kt 12

Julang-2 CSS-N-4 8,000 1 x 200-300 Kt ?

Tactical weapons

Artillery/ADMs, short-range missiles (< 1,000 km) Low Kt 120

Total 410

China is known to be seriously engaged in efforts to modernise its nuclear forces. The focus appears to be primarily on developing "new land-based, solid-fuelled, road-mobile missiles such as DF-21, DF-31 and DF-41 to replace older liquid-fuelled missiles such as the DF-5A as well as producing a new class of warheads thought to be either miniaturised or of smaller yield to increase targeting flexibility and launcher mobility. Other developments include developing a new second-generation replacement sea-launched ballistic missile, the solid-fuelled JL-2, and possibly a small fleet of four to six more advanced ballistic missile submarines, as well as a new bomber, the FB-7, as a replacement for its antiquated H-5 and H6 fleet."46 China is also engaged in developing new warheads for its short-range missiles47 and further improving its tactical nuclear warheads.

Should China ever decide to launch a disarming first strike against India (an improbable but not an impossible scenario), it will be able to employ up to 60 percent of its IRBM force and air delivered warheads, keeping the remainder, as well as the entire ICBM and SLBM force, as a strategic reserve. Hence, in a worst case scenario, approximately 240 warheads (60 percent of 400) would be available for a first strike. With these, using four warheads per target, it will be possible for the PLA's Second Artillery to effectively destroy about 50 to 60 targets, with an assurance level of almost 80 to 90 percent if warheads in the megaton class are used. The strikes are likely to include primarily counter force targets and may include some counter value targets. All air bases, missile silos, hardened shelters for ballistic missiles, rail-mobile missile marshalling railway yards, military headquarters and communications centres and strategic choke points (such as the rail and road bridges across the Brahmaputra River) in eastern and northeastern India will be targeted and destroyed. Only the most skilfully and innovatively concealed nuclear force installations are likely to survive. However, the strikes will not be limited only to eastern India. Similar counter force targets and population and industrial centres in other parts of the country will also be hit and will suffer varying degrees of damage. India's ability to retaliate will depend on how widely dispersed and well concealed over Peninsular India the nuclear force assets are, how efficient the command and control structure is, how failsafe the communications system is and how well integrated the armed forces are in India's nuclear deterrence programme.

Strategic Delivery Systems

The number and variety of nuclear warheads and their level of technological sophistication are ultimately of value only to the extent that accurate strategic delivery systems are available for reaching them to the intended targets. Since it is now possible for almost any technologically advanced country to design simple nuclear weapons (in fact, several designs can be downloaded from the Internet!), the real test of genuine deterrence capability lies in the possession of strategic delivery systems, particularly ballistic missiles. Missiles are increasingly being touted as the new currency of power.48 K. Subrahmanyam is of the view that while nuclear weapons are largely unusable as weapons of war, missiles are usable and that the development of Agni and other such missiles will be useful to keep missiles like the Tomahawks away at safe distances.49 He has observed that "The NATO campaign (in Yugoslavia) is proof that missiles would be standard equipment for war and deterrence in future and, therefore, they are as much part of a country's defences as aircraft, tanks and guns."50 It is in this field that China is well ahead of India and, since India has carried out only limited tests of the Agni-I and Agni-II,51 the 'missile gap' appears to be growing. "Today, Chinese CSS-2 missiles in Saudi Arabia, nuclear warheads from Tibet and mainland China… all affect Indian security interests."52 Despite some reports that the serial production of 20 Agni missiles is likely to be completed by 200053 some analysts have estimated that "the Agni will not be fielded with nuclear warheads for another ten years."54

Another area in which India is lagging is in the development of cruise missiles. Commenting on India's plans for cruise missiles, Pravin Sawhney has written that the Sagarika missile, currently under development, is an anti-ship cruise missile. Quoting Dr. A.P.J. Abdul Kalam, he states that the 'pilotless' target aircraft (PTA) Lakshya "can be converted, without much alteration, to a cruise missile with a range of 500 km and a 300 kg payload."55 The export of missile technology and even fully assembled missiles by China and North Korea to Pakistan has been well documented. In fact, confirmatory information has often come from former CIA officials.56 It is well known that between 30 to 84 x M-11 Silkworm missiles of 280 km range and an unknown number of M-9 missiles were supplied by China to Pakistan during the early 1990s. North Korea has supplied No Dong and Taepo Dong missiles to Pakistan. The US policy has been ambivalent towards the proliferation of nuclear weapons and missile technology by China and North Korea to Pakistan, presumably to safeguard US trade interests. However, this is now being increasingly criticised in the US itself. Paul Bracken has written that "… India, the world's largest democracy, is punished for testing nuclear weapons, while China, hardly a democracy, is courted as a strategic partner."57

Agni-II58

l Length: 20 metres; diameter: 1 metre; weight: 16 tons; achieved range: 2,200 km; achievable range: up to 2,500 km; booster engine: two-stage, solid-solid propellant; time of flight: 11 minutes for 2,200 km, of which powered phase is two minutes.

l Agni-II is an Intermediate Range Ballistic Missile (IRBM), implying that the reentry vehicle, which has the payload and other accessories, leaves the atmosphere, coasts in an elliptical orbit in vacuum and then re-enters the atmosphere at temperatures up to 18,000 degrees Celsius.

l Agni-II is a mobile system with strap-down inertial navigation system and indigenous on-board microcomputers that will provide greater accuracy with the aid of Global Positioning System.

l The two stage-solid booster engine of Agni-II is an improvement over the solid-liquid engine of Agni-I. A complete solid propellant, as compared with a solid-liquid configuration, has quicker reaction time, longer shelf life and minimal logistics problems in the field. Otherwise, basic capabilities of Agni-I and Agni-II are the same. Agni-I—length: 19 metres, diameter: 1 metre, weight: 14.5 tons.

l Approximate cost of a single Agni-II weapon system is Rs. 335 crores. It can be cost-effective only with nuclear warheads. Payload is about 1,000 kg, implying that the actual warhead weight would be 700 to 800 kg. The remaining weight is of support systems inside the re-entry vehicle.

l Agni-II can be enhanced to Agni-II with a range of up to 5,000 km with a new design: the diameter will change from the present 1 metre to about 1.8 metres (as suggested in Dr. A.P.J. Abdul Kalam's book Wings of Fire).

A 1999 RAND report stated that India lacks the capability to launch effective missile strikes against China and is incapable of withstanding a nuclear first strike by China. The report, entitled "From Testing to Deploying Nuclear Forces: The Hard Choices Facing India and Pakistan", observed that "India has an 'unready force', is vulnerable to a first strike, does not have the means to detect enemy ballistic missiles in flight, and does not have a command, control and intelligence structure that is resistant to an attack aimed at decapitating India's civilian leadership."59 Around the same time, the New York Times had reported in a front-page article that "China possessed roughly 20 missiles that could reach American shores and perhaps 300 nuclear weapons aboard medium-range missiles or bombers that could hit India, Russia and Japan."60 While China's aim is clearly to build up Pakistan as a strategic ally to countervail India in Southern Asia, North Korea's interests are primarily monetary in nature. Pakistan's testing of North Korean missiles also enables North Korea to validate its partially tested systems at Pakistan's cost without inviting international opprobrium and the pressure of sanctions.61 Using designs of foreign origin, Pakistan is engaged in pursuing both the liquid and solid fuel routes to develop its indigenous ballistic missile capability. Clearly, India is at a major disadvantage and it is imperative that the foremost national security priority should currently be the development and operational fielding of Agni-I and Agni-II ballistic missiles (to a range of 5,000 km to cover all likely targets in China), followed by SLBMs and the Surya ICBM with a global reach to cater for future threats. Till such time as these missiles become fully operational, India's nuclear deterrent will continue to lack credibility, especially against China.

Force Structure Proposals of Indian Analysts

Throughout the 1980s and the 1990s (up to mid-1998), the Indian nuclear debate concentrated mainly on striving for universal nuclear disarmament, nuclear free zones, whether India should join the Non-Proliferation Treaty (NPT) and the Comprehensive Test Ban Treaty (CTBT), the advantages and disadvantages of exercising the 'nuclear option', the finer nuances of 'opaque', 'non-weaponised', 'recessed' and 'existential' deterrence, and the need to work for a global 'no first use' commitment. Very little was either discussed or published even in academic journals and by security studies think tanks on a nuclear strategy for India and a viable force structure to give effect to that strategy. The anti-nuclear lobby, the so-called 'doves', was so strong that few editors would have dared to publish an article specifying the nuts and bolts of a nuclear arsenal for India. As such, few Indian analysts have written on the subject and there is not much to go by.

K. Subrahmanyam has written that "… if a country can project an image of having around 500 nuclear warheads, which India can build in twelve to fifteen years time if it were to set out on the programme and disperse them on its vast area, the country will have a credible deterrent."62 Even after the Pokhran-II tests, while explaining that minimum deterrence is not a numbers game, he wrote: "Whether it is 150, 250 or 300, the Indian deterrent will still be a minimum one compared to others except Pakistan."63 However, he is known to believe that "… a force of around 60 deliverable warheads could meet adequately India's need for a minimum deterrent."64 For delivering these 60 warheads, Subrahmanyam advocates the development of 20 Prithvi missiles and 20 Agni missiles; the remaining 20 warheads are to be delivered by air force bomber/fighter-bomber aircraft. Subrahmanyam argues that "If India were to develop a modest force of 20 Agni missiles, the India-China ratio in deterrence capability will still be higher than the present China-US ratio."65 He does not visualise the need for SSBNs armed with SLBMs. He does not make a major distinction between low-yield fission weapons and those in the thermonuclear class, but instead emphasises the criticality of solid-fuel missiles: "Whether the warheads are of 15 kilotons-fission or 120-150 kilotons (thermonuclear warheads), both are bound to have a deterrent effect… What is absolutely crucial for credible deterrence is the solid-fuelled missile of appropriate ranges. That is what India needs to concentrate on."66

Jasjit Singh also advocates a minimalist approach and a time period of 15 to 20 years for the Indian arsenal to stabilise. He writes.67 "The exact size of the arsenal needed at the end-point will need to be worked out by defence planners based on a series of factors. But at this point it is difficult to visualise an arsenal with anything more than a double-digit quantum of warheads. It may be prudent to even plan on the basis of a lower end figure of say 2-3 dozen (survivable) nuclear warheads by the end of 10-15 years. It is necessary to keep in mind the fact that with the passage of time, deterrence decay factors will lead to the requirement of a smaller arsenal rather than a larger one." Though he bases his force structure on a triad for delivery, he feels that "It would indeed be infructuous to pursue the development or possession of an intercontinental capability to do so," and that intermediate range ballistic missiles (IRBMs – with a range between 500 and 5,000 km) would be adequate for India's requirements. Maharajkrishna Rasgotra is of the view that "Some 30 bombs of Hiroshima strength committed against five major targets in Pakistan, 60 deployed against eight to ten targets in China, and another 30 held in reserve for contingencies and deployment at sea, should adequately meet the needs of minimum deterrence. This number (120 warheads in all) allows for possible losses in an enemy first strike and leaves enough for a devastating counterattack."68

General K. Sundarji, a former Indian Chief of the Army Staff and a perceptive military thinker, was perhaps the first analyst in India to write about the military aspects of India's nuclear deterrence. He advocated a nuclear force structure of approximately 150 warheads mounted almost entirely on a Prithvi-Agni missile force.69 Brigadier Vijay K. Nair has suggested a force level of 132 nuclear warheads of different types, including weapons in the megaton range.70 For delivery, besides bomber/fighter-bomber aircraft, he recommends five SSBNs (each with 16 SLBMs) and 48 ballistic missiles (12 SRBMs and 36 MRBMs). He writes: "India must ensure adequate reserves to provide fail safe assurance of her strategy and yet maintain an adequate force structure after hostilities cease. An additional reserve of two weapon systems is required for each planed autonomous strike and a minimum of 20 percent of the entire force structure should be available for post-strike security imperatives." Out of a total requirement of 111 nuclear warheads for retaliatory strikes against Pakistan (17 targets) and China (eight targets), he feels that 37 warheads are required for strikes and an additional 74 as "65 percent reserve for reliability". He adds another 22 as a "post-war reserve", taking the total to 132 warheads.

Rear Admiral Raja Menon (Retd.) recommends that India's nuclear arsenal should be based primarily on SSBNs from about 2020 onwards.71 Till then, he feels that India's nuclear deterrent should be based only on ballistic missiles. He excludes bombers/fighter-bombers from the deterrence calculus because of the destabilising impact of the short flying time (20 minutes) between Pakistan and India's forward airfields and the inducement for the early use of nuclear weapons even in a conventional conflict. He supports the elimination of air delivered nuclear weapons from the arsenals of both India and Pakistan under mutually agreed nuclear risk reduction measures (NRRMs). He writes: "If both India and Pakistan can resist the temptation to field some kind of nuclear weapons urgently and postpone their arsenals to a less vulnerable set of missiles, the total number of nuclear weapons would be reduced; and, the inducement to a first strike would also vanish." While the logic may be impeccable, practical realities preclude the adoption of such an option. Till the Agni-I and Agni-II IRBMs become fully operational and the Indian Navy begins to acquire SSBNs, India has no option but to base its deterrence against China solely on air delivered nuclear warheads. Indian bombers/fighter-bombers such as SU-30s, Jaguars and Mirages can effectively reach value targets in the Chinese hinterland. However, it would be wishful thinking to assume that Pakistan would accept that India might retain nuclear-capable/nuclear-armed aircraft in the eastern theatre while removing them from its western borders under mutually agreed NRRMs.

Admiral Menon has estimated that the modernised Chinese arsenal would comprise 596 warheads after 2010. Up to 2030, he suggests that an all-missile, land-based force should comprise five regiments of 12 missiles each (with survivability being ensured by concealment and rail-garrison mobility) and fifty percent of them should have up to four independently targetable warheads each. He feels that these would suffice to withstand a first strike by China with the maximum number of warheads that China may decide to launch and yet have enough missiles remaining to inflict unacceptable damage. He feels that some hardened silos may need to be provided "if the rate of degradation of the rail garrison missile force is judged to be too rapid." Against Pakistan, he proposes a force of 200 cruise missiles, 36 of them nuclear tipped, as cruise missiles are the least provocative. He visualises the "handing over of Indian deterrence from the land-based force to the sea-based force… over a ten year period… (to be) completed by 2030" and suggests a nuclear force of six SSBNs, each armed with 12 SLBMs. Each SSBN will carry at least 12 missiles and, in his view, as India has MIRV (multiple, independently targetable re-entry vehicles) ambitions, each missile could carry up to ten 250 to 400 Kt nuclear warheads. "Such a force would give India a warhead strength of 216 (6 x MIRV) in a pre-launch scenario and probably 380 warheads in a scenario with adequate strategic warning and with five boats deployed. This could be the entire Indian deterrence till the middle of the 21st century."72

It is indeed intriguing that Admiral Menon does not visualise a need for air-delivered nuclear warheads for the Indian deterrent. He writes:73 "With opposing airfields separated by barely 20 minutes flying time, it would be a case of use-them-or-lose-them for Pakistan, a fear reinforced by the threat of capture by armoured forces, in a country handicapped by lack of strategic depth. An airdropped bomb is perhaps the farthest from a second strike weapon on the Subcontinent. It contributes nothing to deterrence stability and, if at all the weapon is discussed in a worsening crisis, it can only be in reference to a first strike." Besides their dual-use capability and the "sunk costs" already incurred in the acquisition of nuclear capable bombers/fighter-bombers, it must be remembered that unlike ballistic missiles, aircraft are recallable even after they have taken off with nuclear weapons on board. They present a cost-effective solution to India's immediate deterrence requirements till the Agni-I and Agni-II series of ballistic missiles can be made fully operational. As for the proclivity to use-them-or-lose-them, the analogy is suggestive of flippant nuclear decision-making. There is no reason to believe that the Pakistani leadership, military or civilian, will act irrationally and set off nuclear exchanges merely because they would be worried about their forward airfields being put temporarily out of commission by the Indian Air Force. India will need air-delivered nuclear warheads in its arsenal not only against Pakistan, but also against China for a long time to come as they offer a here-and-now solution and are akin to a bird in hand.

To base the entire Indian nuclear force only on SSBNs would not be appropriate for the following reasons:

l Flexibility of targeting options for targeting individual targets with a variety of weapons platforms to achieve better strike assurance would not be available.

l The problems of communicating with submerged submarines are well known.

l It is becoming increasingly easier to locate and track submarines and by 2020-30 advanced navies such as the US Navy may be able to map and track the position of every submarine in any of the oceans.

l The Indian hinterland provides adequate depth and area to disperse widely India's ballistic missiles and the rail and road networks provide ample opportunity to keep moving the missiles at random, though with attendant problems of command and control and missile launch inaccuracies.

l The force structure would lack inter-Services balance, which in itself is not desirable.

Bharat Karnad follows what has been dubbed a 'maximalist' approach to nuclear deterrence and strongly advocates the need for megaton-class thermonuclear weapons in the Indian arsenal. He assumes that the primary and secondary target lists could contain about 60 locations that need to be hit. In order to ensure that each of these targets can be destroyed with an acceptable assurance level so that deterrence is credible, he recommends the targeting of each with four nuclear weapons, each of which has a two mile (approximately three km) CEP (circular error probable—a measure of the accuracy of delivery; it denotes the distance from the point of impact to the centre of the target as the radius of the circle within which, on average, 50 percent of the missiles aimed at the target will fall). Bharat Karnad suggests that India's nuclear arsenal be gradually built up over a period of three decades to a total of 328 nuclear warheads, as given in Table 3:74

Table 3. Requirement of Nuclear Warheads

Timeframe Maximally Strategic* Minimally Tactical** Total

(Warheads) (Warheads)

l 2000-2010 57 30 87

l 2010-2020 131 40 171

l 2020-2030 268 60 328

Delivery systems:

* ICBMs, IRBMs, SLBMs, SU-30s armed with N-gravity bombs (NGBs) and N-air-to-surface missiles (N-ASMs) and atomic demolition munitions (ADMs).

** Jaguars and Mirage-2000s armed with NGBs and N-ASMs up to 2010 and SU-30s thereafter.

The breakdown of the final figure of 328 nuclear warheads and the proposed delivery systems suggested by Bharat Karnad is as under:

l 4 x SSBNs with 48 SLBMs (presumably with a single warhead each).

l 40 x SU-30s with 40 x NGBs and 40 x N-ASMs (maximally strategic) and 30 x SU-30s with 30 x NGBs and 30 x N-ASMs (minimally tactical).

l 25 x ICBMs.

l 40 x IRBMs.

l 25 x ADMs.

l 50 reserve warheads.

Leaving out the 50 reserve warheads, Bharat Karnad has suggested that, out of the remaining 278 warheads, 253 should be thermonuclear. Hence, in his view, all the warheads in the Indian arsenal, with the exception of 25 ADMs, should be thermonuclear. He writes: "If a counter-cities or counter-value nuclear bombardment strategy is the only one that makes sense, then thermonuclear bombs, with megaton yields, are the most convincing instruments of this strategy."75 It is difficult to dispute the logic of this statement. However, India's 45 Kt, two-stage thermonuclear explosion during the tests of May 1998 has the potential to be upgraded to a practical capability of about 200 Kt only.76 For weapons in the megaton class to be developed, further refinement of the design and additional physical testing would be necessary. As India has unilaterally renounced further nuclear tests, it is doubtful if such a capability can be created. Also, as missile accuracies continue to improve and CEPs drop to hundreds of metres and lesser, the trend among the P-5 is to limit warhead yields to between 200 to 300 Kt. Yields in this range should be adequate for India's purposes also for a counter-value strategy provided India can develop ballistic missiles with a CEP of less than 500 metres. As for ADMs, their use in Tibet, as suggested by Bharat Karnad to stop a major Chinese offensive, would prove to be counter-productive as the waters of almost all the rivers in Tibet flow into India and it would hardly be prudent to pollute those waters with long-lasting nuclear radiation by creating landslides on the snow covered mountainsides and endanger life in the northeastern states. Besides, it would be a major environmental disaster.

R.R. Subramanian, a senior analyst at the Institute for Defence Studies and Analyses (IDSA), New Delhi, and a physicist by training, is of the opinion that India needs at least 425 warheads if the combined efficiency (accuracy, reliability, in-flight interception) of the delivery systems is taken to be 0.3. He feels that with an efficiency of 0.3, approximately 125 to 130 warheads could be counted upon to destroy their intended targets. Of these, he is of the view that 25 to 30 are needed to target Pakistan and about 100 are required against China.77 To assume that India's nuclear force would deliver an overall efficiency of only 0.3 would appear to be overly pessimistic. However, in the absence of accurate factual information about individual components of the nuclear force, primarily the targeting and delivery systems, it is possible to neither prove nor disprove this contention.

Lieutenant General Pran Pahwa (Retd.) recommends that India's deterrence be based on 182 warheads.78 He bases his calculations on the assumption that China is likely to employ two warheads each to destroy every Indian warhead and that 20 percent Indian warheads would survive a Chinese first strike which would be essentially a counter force one. He feels that if India had 182 warheads, China would need to fire 364 warheads and, given a Chinese arsenal of 400 warheads, it would be left with 36 to India's surviving 36 warheads. Since the numbers remaining would be matched, China would be deterred from launching a first strike. This argument gives the Chinese an excessive 80 percent success rate and does not take into account the possibility that a Chinese first strike is likely to combine counter value with counter force targets.

It emerges that Indian analysts have widely varying views on the number of nuclear warheads that India needs for its minimum deterrent. The figures vary from the low double digits ("two to three dozen") at the lower end to just over 400 at the upper end. Suggestions for weapons yield range from 15 to 20 Kt fission weapons to thermonuclear weapons in the megaton range. The recommended delivery vehicles embrace the entire range of the triad including ICBMs and cruise missiles. As discussed earlier, the sole purpose of India's nuclear weapons is to deter the use and the threat of use of nuclear weapons. Minimum deterrence is not a numbers game. Its ends are served if the adversary is deterred from crossing the nuclear threshold and from threatening to do so. As Kenneth Waltz famously said, "More is not better if less is enough."79 What matters is to find a rational way to determine how much is enough.

Practical Considerations Underpinning the Force Structure

It has been emphasised by the Indian government that India's nuclear weapons are not country specific. However, a military force structure can only be configured to neutralise specific current threats and must be made flexible enough to be adaptable to meet emerging challenges. Without going into specific details of the nuclear threats faced by India, an aspect that requires a separate analysis, it could be stated that the nuclear weapons possessed by both China and Pakistan, when placed in the context of their unresolved territorial and boundary disputes with India, are definitely a threat in being while those possessed by the other NWS that are distant from India are far less hazardous. Hence, a realistic force structuring option would be to base India's deterrent on the nuclear weapons stockpile of China as a planning parameter.

How Many Cities should be Targeted?

The key decision to be made in working out a nuclear force structure for India is regarding the number of nuclear warheads necessary for a retaliatory strike with a counter value strategy that emphasises attacks on large cities and industrial centres. At the heart of the planning process is the question: how many such targets must Indian nuclear forces plan to destroy in order to ensure that deterrence works? Would it be adequate to target one large, metropolitan-cum-industrial centre with thermonuclear weapons, or would it be necessary to plan to destroy ten, twenty or more such complexes? Is the level of deterrence proportionately dependent on the number of targets in the adversary country that one's nuclear force is capable of destroying or, does the law of diminishing returns apply after the capability to destroy a certain number of targets has been acquired? These are complex challenges to resolve and several reputed nuclear strategists have attempted to find the answers through intuitive as well as mathematical solutions based on econometric models and the "innovative use of statistics". There is, perhaps, something to be said for the theory of diminishing returns. Quoting Bernard Brodie, Bharat Karnad has written:80

Brodie, in one of his most memorable series of insights, noted that in the nuclear age, "the potential deterrence value of an admittedly inferior force may be sharply greater than it has ever been before," in the event, that "a menaced small nation could threaten (a big one) with only a single thermonuclear bomb" directed at its premier city, which he contended "would be a retaliatory capability sufficient to give the (aggressor) government pause." Further, that ten thermonuclear tipped ("city busting") missiles aimed at the same number of the enemy's cities "would no doubt work still greater deterrent effect." But, increase in the deterrent effect, he surmised, would be "less than proportional to the increase in magnitude of potential destruction" in part because the deter-er would run out of high-value enemy metropolitan centres to hit. It amounted to saying that too large a nuclear arsenal would only fetch diminished deterrence.

However, Herman Kahn, another prima donna of nuclear theology, disputed Brodie's logic. He explained that, as the efficiency of a ballistic missile is the square of its accuracy as represented by its CEP,81 ten ballistic missiles aimed at ten cities would not deter an adversary who could be reasonably sure that given a CEP of one to two km, a large number of them are bound to fall substantially away from the city centre and may be only partially effective or completely ineffective. Hence, for credible deterrence to be achieved, the adversary must be certain that India would be in a position to ensure that three to four nuclear warheads can be delivered with the required accuracy on every selected target. The comparatively greater density of population per square kilometre in Asian cities is another factor that must be taken into account while deciding the number of cities that must be targeted for effective deterrence.

Despite Mao's assertion that "300 million Chinese would survive" nuclear war, it could be argued that the fear of losing some of its modern showpieces on the eastern coast, combined with the certainty of horrendous civilian casualties due to extremely high population densities, would be adequate to deter China from being the first to begin nuclear exchanges that are bound to escalate to city-busting strikes. The China scholars at IDSA hold sharply divergent views on the number of Chinese cities that need to be targeted to ensure deterrence. Sujit Dutta is of the opinion that China would be deterred if its leadership were convinced that its adversary could destroy even three major cities.82 M.V. Rappai concurs with this view and argues that the Chinese are taking their economic development very seriously and would not do anything to jeopardise the future of their thriving population and industrial centres.83 Swaran Singh advocates the targeting of five cities for effective deterrence but feels that rather than the ability to target a number of cities, India's overall nuclear capability should be built up for effective deterrence.84 However, Srikanth Kondapalli holds the view that perhaps even the credible targeting of 15 to 20 Chinese cities may not be adequate for deterrence as the Chinese would not hesitate to take whatever military action they might consider necessary if, in their view, their national security interests were to be seriously threatened.85

In the context of China, India's major nuclear adversary, the targeting of only three to five cities may be inadequate as totalitarian regimes are known to have high tolerance levels. In the absence of hard intelligence about what would deter the Chinese, it would be appropriate for a nuclear planner to err on the side of caution and plan on overkill to ensure that deterrence does not fail. Hence, it could be argued that India should plan on the assured destruction of about 10 to 12 major population and industrial centres (including two to three high priority military or counter force targets such as SSBN bases and nuclear command and control centres) in a retaliatory strike. This should definitely be adequate for the purposes of credible deterrence in the 2000-2010 timeframe. As China develops further into a major industrialised nation, the possible destruction of even three to five major population-cum-industrial centres would be adequate for deterrence. In the case of Pakistan, the destruction of an even lesser number of carefully selected targets would mean that Pakistan would ultimately cease to exist as a nation. An assured destruction capability in respect of eight to ten major value targets would be recognised in advance by India's adversaries as a catastrophic disaster of unmanageable proportions. Such a capability would constitute adequate deterrent threat to qualify the likely damage as unacceptable.

Are Megaton Monsters Necessary?86

The next major issue that needs to be resolved is whether the Indian deterrent should be built with 'megaton monsters' or be limited to boosted fission kiloton weapons with yields ranging from 20 to 30 Kt with current technology. It can hardly be disputed that the deterrence value of thermonuclear weapons is much greater than that of fission weapons due to their far greater destruction potential. However, the key question is whether such destructive potential is militarily necessary. Solly Zuckerman has stated that "There is built into nuclear weapons greater destructive power than is necessary for military purposes and their secondary, non-military effects overshadow those which relate specifically to their military use."87A nation with a limited availability of fissile material, that has opted to suffer what may be a disarming first strike and must, for that reason, stockpile at least twice the number of warheads than what it may actually need for deterrence, has no choice but to opt for thermonuclear weapons in the megaton class. Another compelling reason for thermonuclear weapons in the megaton range is that their larger lethal radius can achieve the desired results with lesser numbers even when the accuracy of delivery is low, that is, the CEP is one to two km. A megaton warhead-tipped ballistic missile would cause horrendous damage even if it detonated well off a city centre and would, therefore, contribute far more to deterrence than fission or even boosted fission bombs. Dr. G. Balachandran argues succinctly for India to base its deterrence on high-yield thermonuclear weapons:88

"… India cannot rely on a nuclear deterrent based on weapons of the types tested, that is, fission weapons in the range of 20 Kt. Even with accurate long-range missiles, with a CEP of the order of 200 metres, the (fissile) material requirements are far in excess of the current inventory… It should be understood that increasing the yields of pure fission type of weapons would not solve the problem. On the contrary, the use of fission devices of higher yield will require larger stockpile of W-Pu (weapons-grade plutonium). For instance, for a soft point target at large distances, with a CEP of 1,000 metres, one would require either 10 weapons with a yield of 20 Kt (each) or one with a yield of one megaton. 10 x 20 Kt weapons would require 30 kg of W-Pu… one pure fission weapon with a yield of one megaton would require 150 kg of W-Pu. A thermonuclear weapon, of course, requires far less fissile material for high yields with resultant reduction in the weight of the weapon. Therefore:

l The first requirement for an effective and credible nuclear deterrent is the need for the Indian nuclear arsenal to be based on high yield thermonuclear weapons.

l The second requirement… is to accelerate the missile development programme, especially the development of ICBMs… India cannot be said to have a truly effective nuclear delivery system against China. Its delivery systems are not yet adequately developed.

Citing a Harvard Nuclear Study Group report on the politics of nuclear weapons, Bharat Karnad writes about a 'nuclear paradox':89 "Nuclear weapons can prevent aggression only if there is a possibility that they will be used, but they should not be made so usable that anyone is tempted to use them… The decision to launch city-busting hydrogen bombs and thermonuclear warheaded ICBMs in a direct attack on the enemy homeland would, on the face of it, be more onerous and far-reaching and, hence, will be more difficult to make than, say, a decision to loose off a theatre nuclear weapon or a tactical nuclear salvo on a peripheral target. And for this reason, thermonuclear weapons, theoretically speaking, would be under tighter command and control and would more easily help stabilise the security situation vis a vis a bigger nuclear power." Also, as thermonuclear weapons do not cost substantially more than fission weapons, it would make sense to optimise India's meager fissile material stockpile by producing sufficient thermonuclear weapons in the megaton class to equip at least all ballistic missiles with them. However, though the requirement is apparently justified, since the demonstrated thermonuclear capability is limited to only 200 Kt warheads, India will have to make do with 200 Kt weapons till (and if) further testing enables the development of megaton class warheads. On the other hand, if in future it is possible to improve the accuracy of India's IRBMs to a CEP of less than 0.01 percent (50 metres at 5,000 km; the Minuteman-III ICBM of the US is reported to have a CEP of 120 metres at a range of 13,000 km), though they would still not be ideal, 200 Kt thermonuclear warheads would be adequate to meet the requirements of deterrence.

The Need for a Triad

Many analysts, particularly those in the West, have interpreted the National Security Advisory Board's reliance on a triad in the draft Nuclear Doctrine as one of the main areas of concern for a likely arms race in Southern Asia. Criticism has centred around the view that a doctrine of 'minimum deterrence' does not need a triad of delivery systems. Their main bone of contention is that SSBNs with SLBMs are not essential for India's nuclear force. All such criticism is obviously ill informed and without objective analytical basis. It would be unrealistic to base India's retaliatory strike force mainly on bomber/fighter-bomber aircraft and land-based ballistic missiles. While all possible targets inside Pakistan can be fully covered by the deep penetration aircraft in service with the Indian Air Force (IAF), they lack the range necessary to hit high-value targets deep inside 'Han' China, including Beijing, and the rapidly modernising cities on the east coast. Air-to-air re-fuelling capability is also not available.

It is often said that a SU-30 on a one way mission to Beijing can reach its target and that there would be no dearth of pilots in the IAF to volunteer for such a suicide mission. While that may be true, it would be foolhardy to base deterrence calculations on such a course of action. Though the SU-30 is a multi-role aircraft, due to the paucity of numbers presently available and those on order, a SU-30 deep strike mission would require a large number of specialised air defence and electronic warfare-capable aircraft as escorts to negotiate vast stretches of the increasingly well defended Chinese air space. The escort aircraft with the IAF are mostly capable of only tactical ranges and during the final and the most crucial portion of the flight, the SU-30s would have to fly very much on their own. If the Su-30s were launched un-escorted, their chances of survival would be rather slim during war and their contribution to furthering the cause of deterrence would remain completely doubtful during peace. In view of the fact that an Indian strike would be a retaliatory one, the risk of SU-30 airbases in eastern India being rendered unusable due to the preceding nuclear as well as conventional strikes, would also have to be vectored into the planning parameters.

A substantial proportion of the land-based ballistic missile force is also likely to be destroyed in a disarming first strike or even a conventional strike during war before the nuclear threshold is crossed. Missiles of the Agni-I class can be made road-and rail-mobile and can be moved around over large areas in a random, un-predictable manner without a discernible pattern, though they would be vulnerable to attacks by terrorists and insurgent groups. They can also be housed in hardened, over-ground shelters and moved frequently from one shelter to the other or emplaced in fixed silos designed to withstand the overpressures likely to be generated by nuclear explosions of 20 to 30 Kt. Hardened shelters can be easily spotted by modern satellites and can all be destroyed in a disarming first strike. To ensure that not all the missiles are so destroyed, such shelters will need to be constructed at the scale of at least two to three per missile, including some realistic dummy shelters that may not be hardened but must be equipped with dummy missiles mounted on actual tractor erector launchers (TELs) being moved into and out of them so as to enhance their credibility as real missile shelters. Though mobile missiles are harder to locate and track and are less vulnerable at present, surveillance and target acquisition technologies are improving rapidly and adversaries planning first strike strategies may be expected to make the required investments in them.

Fixed silos capable of withstanding a nuclear attack are extremely costly to construct and maintain. So far it has not been possible even for advanced Western countries to construct shelters capable of withstanding the blast of thermonuclear weapons in the megaton range, because such silos are a technological challenge and are prohibitively costly. Should India's adversaries choose to employ such weapons in future, fixed silos would be virtually useless. Hardened shelters for mobile missiles and fixed silos for the Agni-II and larger class of missile will have to be constructed over huge areas with wide gaps between them to ensure that an attack on one shelter or silo does not destroy more than one missile. They also have to be secured against stand off and commando type ground attacks by terrorists and insurgent groups. Despite all these measures, it would be fair to assume that up to 40 to 50 percent of the strategic missile force would still be destroyed in a disarming first strike as, despite the US failure to destroy Iraqi Scud launchers during the 1991 Gulf War, modern surveillance capabilities are continuously improving and new military satellites will enable the constant tracking of an adversary's missile force. There is apparently a need to hedge against the destruction of land-based ballistic missiles by distributing them between static silos, hardened shelters and rail-and road-mobile storage. At the same time, there is a need to build into the force structure sufficient reserves to cater for pre-launch losses. Nuclear deterrence does not come cheap.

Since air bases are susceptible to destruction in a disarming first strike and India's strategic aircraft themselves have limited range and are vulnerable to in-flight interception by the enemy's air defences, and a large proportion of the land-based ballistic missiles may be destroyed before they can be launched, there is no option for India but to go in for submarine launched ballistic missiles (SLBMs) on nuclear-powered submarines (SSBNs) as soon as both SSBNs and SLBMs can be developed. As is well known, SSBNs are fairly safe from detection even with state-of-the-art reconnaissance, surveillance and target acquisition (RSTA) means, not vulnerable to a 'zero warning' surprise attack, and are not affected by increases in the adversaries' missile accuracy except those that may be anchored in a submarine base at the time of attack. They also have their limitations, not the least of which include the technological complexity of achieving the desired missile launch accuracy from a mobile platform that can never be as sure of its location (and depth) as a fixed silo and the difficulty of communicating targeting information and executive orders to them once they are submerged. Also, in the Indian context, SLBMs would provide value for money only if they could be developed to reach ranges in excess of 5,000 km so that they can reach all likely targets from their patrolling bases well away from the shores. The SSBNs should be capable of operating from patrolling bases in the Indian as well as the Pacific Oceans and should be equipped for spending at least 30 to 40 days continuously at sea. As stated earlier, Brigadier Vijay Nair is of the view that India needs four SSBNs, each with 16 SLBMs; Bharat Karnad wants five SSBNs in India's nuclear force, each with 12 SLBMs and Rear Admiral Raja Menon has recommended a force of six SSBNs (with two 'on station' at all times), each equipped with 12 SLBMs. Six SSBNs would obviously be preferable to four or five as they would provide greater redundancy and flexibility. However, the difference in capital costs would be considerable and four SSBNs (with a minimum of at least one on patrol at all times) should meet India's requirements of deterrence well into the first few decades of the 21st century.

Recommended Nuclear Force Structure

It is now acknowledged in almost all quarters that successful deterrence does not demand qualitative or quantitative parity in force structures—the ability to inflict unacceptable damage is adequate. However, an adversary confronted with having to worry only about a retaliatory strike, would be deterred only if he was convinced that the nuclear warheads aimed at his cities, military and industrial complexes would, firstly, survive his own first strike in adequate numbers; secondly, they are powerful enough to destroy vital targets and, thirdly, they can be delivered with the required accuracies. The problem of survival can be overcome by building in sufficient redundancies into the force structure, besides dispersion, hardening and concealment. The remaining two, accuracy of delivery and the warhead yield, are directly dependent on each other – greater the CEP of a missile, larger the warhead yield required to cause the same damage for a given assurance level. G. Balachandran's calculations for an assurance level of 90 percent are given at Table 4:90

Table 4. Number of Missiles Required to be Launched to Destroy a Point Target with a 90 Percent Assurance Level

Type of Target Soft Semi-hard Hard

(Overpressure) (20 psi) (100 psi) (300 psi)

Yield Missile CEP Missile CEP Missile CEP

200 m 1,000 m 200 m 1,000 m 200 m 1,000 m

20 Kt 1 10 2 39 4 84

1 Mt 1 1 1 3 1 6

It can be seen from Table 4 that to destroy a soft point target, of strength up to 20 psi, 10 ballistic missiles with a CEP of 1,000 metres and a warhead yield of 20 Kt are required to achieve an assurance level of 90 percent. This is because the single shot kill probability (SSKP) of such a missile is only about 0.2. If the CEP of the missile could be improved to 200 metres, only one missile would be sufficient. On the other hand, if the missile could be tipped with a one-megaton warhead, one missile would still be adequate even if its CEP was as high as 1,000 metres. Soft area targets like population and industrial centres (which may have an area in excess of 100 km square), that are primarily likely to be targeted as part of India's counter value targeting philosophy, would require a much larger number of missiles to destroy with a 90 percent assurance level. However, in their case, the Commander, Strategic Forces Command (a tri-Service command to be created under the CDS) may accept a lesser assurance level (possibly 70 to 80 percent) because the damage would be horrendous in any case.

For 10 counter value targets in China, a total of 40 nuclear warheads (at the scale of four warheads per target, at least three of which should be of 1 megaton each) would be adequate to cause unacceptable damage if the CEP of the delivery systems was 1,000 metres and an assurance level of about 70 percent was acceptable.91 If the efficiency or overall reliability of the whole system was taken as between 0.5 to 0.6, a reasonable assumption for a modern nuclear force, then 80 warheads must actually be launched for about 40 warheads to be effectively delivered and explode over their targets. Hence, 80 warheads and, of course, their delivery systems must survive a first strike. If maximum possible concealment and dispersion measures have been taken, including the emplacement of dummy warhead storage sites and dummy mobile missiles, in the worst case, approximately 50 percent of the land-based nuclear warheads and delivery systems may be destroyed in a first strike.92 Of the SLBMs carried by SSBNs, 80 to 90 percent may be expected to survive. It would, therefore, be necessary to plan a warhead stocking level of twice the numbers of land-based warheads and delivery systems required to be launched and cater for the loss of some sea-based warheads. If approximately 25 percent to one-third of India's deterrence is sea-based, a total of about 150 warheads must be stocked. That is, 120 land-based warheads and about 30 warheads on SLBMs. The last aspect to be catered for is a prudent level of reserves for larger than anticipated damage in a first strike and unforeseen eventualities. Escalation control/domination and war termination strategies would also be dependent on the ability to launch counter-recovery strikes if necessary and some fresh strikes. Adding one-third the required number of warheads should be adequate. Hence, the requirement works out to 200 nuclear warheads for a minimum deterrence policy with a no first use strategy against China if 10 major population and industrial centres are to be attacked in a retaliatory strike to achieve a 70 to 80 percent assurance level of destruction.

The question naturally arises: what about deterrence against Pakistan? Clearly, India's 200 warheads with the necessary delivery systems would also be more than adequate for deterrence against Pakistan. Doomsayers will, of course, worry about a collusive Sino-Pak first strike. If such an incredible eventuality actually transpired, it would be an extraordinary failure not only of deterrence, but also of the entire diplomatic process and India's nuclear arsenal would prove to be inadequate even if it was as large as that of China. What could be considered is that, as and when the fissile material stocks position permits, 20 to 30 x 20 Kt boosted fission warheads could be produced for short-range, Pakistan-specific, ballistic missiles like Prithvi-plus (400 to 450 km range with strap-on boosters; 500 kg warhead) to reduce the costs of delivery systems. These missiles could form part of suitably enhanced reserves.

India is confronted with a unique missile quantity versus warhead quality paradox. Counter value targets, being massive area targets, require very large numbers of accurate missiles to destroy if 20 to 30 Kt warheads are employed—India can ill-afford the manufacture of a large number of missiles with CEP less than 200 metres.93 Also, the fissile material stockpile does not permit the luxury of producing a large number of warheads. On the other hand, if tried and tested 1 megaton warheads were available, a much smaller number of missiles (with a relatively larger CEP of up to 1,000 metres) would be adequate for meeting the requirements of deterrence. These twin weaknesses make the composition of India's present nuclear arsenal unbalanced. After the Pokhran-II tests, at best it is possible to produce 200 Kt thermonuclear warheads, for which technological capability has been claimed by Dr. Chidambaram. Also, the non-availability of operationally proven Agni-I and Agni-II missiles and the fact that SLBMs are still at least a decade away from being inducted into the Indian Navy,94 dictate that India's nuclear force structure will need to evolve incrementally to keep pace with technological advancements. In the initial years there would be no option but to base deterrence capabilities on air-delivered nuclear warheads and those that can be delivered by available Prithvi-plus and Agni-I missiles. Similarly, the number of warheads that can be produced will be limited by the availability of weapons-grade plutonium. For India to produce a total of about 200 warheads, either the production of weapons-grade fissile material will have to be continued, with corresponding repercussions for joining the Fissile Material Cut-off Treaty (FMCT) negotiations or reactor-grade plutonium stocks will need to be used, for which additional testing may be necessary.

Taking into account the requirement and the likely availability of nuclear warheads and delivery systems, it would be advisable that India's nuclear force be raised in a phased manner over a period of three decades. Mid-course corrections can be applied based on the availability of new technologies and developments in the diplomatic field. For example, depending on the pace of development in China and whether that country graduates to a democratic form of government, the need to plan to target ten cities and industrial complexes for a counter value strategy, could be reviewed around 2010. In the nuclear era, strategy has never been the sole determinant of force architecture. This, according to Rajesh Rajagopalan is exemplified by the US decision to opt for the MIRV programme as the technology for it was available and it would help them to circumvent nuclear arms reduction negotiations.95 The technology trajectory will continue to drive nuclear force structures that should therefore be flexible and adaptable. The recommended nuclear force structure is given at Table 5.

Table 5. Recommended Nuclear Force Structure: 2000-2030

Delivery System Number(s) Warheads

(Including reserves)

Phase I: 2000-2010

l Prithvi-plus unit 1 (8 launchers) 16 (20 to 30 Kt fission)

l Agni-I unit 1 (8 launchers) 24 (200 Kt thermonuclear)

l Dhanush 4 launchers 8 (20 to 30 Kt fission)

l SU-30s, Mirage-2000s -- 32 (200 Kt thermonuclear)

Jaguars

Total 80 warheads

Phase II: 2011-2020

l Prithvi-plus units 2 (16 launchers) 16 (20 to 30 Kt fission)

l Agni-I unit 1 (8 launchers) 24 (200 Kt thermonuclear)

l Agni-II units 2 (16 launchers) 36 (1 Mt)

l 2 x SSBNs 24 (SLBM launchers) 26 (1 Mt)

l SU-30s, Mirage-2000s -- 48 (200 Kt thermonuclear)

Total 150 warheads

Phase III: 2021-2030

l Prithvi-plus units 2 (16 launchers) 16 (20 to 30 Kt fission)

l Agni-I units 1 (8 launchers) 24 (200 Kt thermonuclear)

l Agni-II units 3 (24 launchers) 54 (1 Mt)

l Surya ICBM ? ?

l 4 x SSBNs 48 (SLBM launchers) 50 (1 Mt)

l SU-30s -- 56 (200 Kt thermonuclear)

Total 200

Thumb rule planning ratios in force structuring are almost invariably wrong and one nation's force structure details cannot be readily compared with another. However, ratios can sometimes be indicative of general trends. In the US, at the height of the Cold War, during the mid-1980s, the distribution of nuclear warheads was as under:96

l Land-based missiles 4,070.

l Submarine-launched ballistic and cruise missiles 8,712.

l Air-delivered bombs and missiles 1,600.

The ratio of US warheads worked out to approximately 2.5 (land):5.5 (sea):1 (air). In the Indian force structure recommended above, the ratio is about 2 (land): 1 (sea): 1 (air). There is no doubt that deterrence based on SLBMs is qualitatively superior. However, SSBNs require very high capital investments and it is unlikely that India's defence budget, pegged as it is at less than three percent of the GDP, will be able to support the development and acquisition costs of more than four SSBNs.

There is undoubtedly much more to a nuclear force structure than working out the number of warheads and delivery systems. Decisions regarding deployment, the custody of nuclear warheads, alert status, command and control and the need to ensure deterrence stability, present complex challenges. Jasjit Singh is of the view that India's nuclear arsenal "should be built up gradually and not deployed as a full-fledged weapon system. This implies keeping delivery systems and warheads separate, with the latter themselves stored in disassembled form."97 Others have expressed a view that there cannot be any credibility without deployment. These issues are not dealt with in this article as they require more detailed examination. However, there can be no doubt that for India's no first use doctrine to be credible, India's strategy should be to target high value population and industrial centres in adversary countries with a high level of assurance after absorbing the full weight of what would in all probability be a disarming first strike. Only then would the adversaries be sufficiently deterred to avoid launching nuclear strikes against India. While India's present capabilities may be limited, these must be gradually built up to a level India considers adequate for its national security requirements.

 

Notes

1. Kenneth Waltz, "Does India need the Bomb?" The Times of India, January 26, 2000.

2. Bhabani Sen Gupta, Nuclear Weapons: Policy Options for India (New Delhi: Sage Publications, 1983), pp. 88-89.

3. John Lewis Gaddis, The Long Peace: Inquiries into the History of the Cold War (New York: Oxford University Press, 1987), p. 109.

4. Cited by General K. Sundarji, "Nuclear Deterrence Doctrine for India", Trishul, vol. 5, no. 2, December 1992, pp. 43-60.

5. K. Subrahmanyam, "No More Hibakushas', The Economic Times, June 18, 1998.

6. Gen. Sundarji, n. 4.

7. V. N. Khanna, India's Nuclear Doctrine (New Delhi: Samskriti, 2000), p. 151.

8. K. Subrahmanyam, "Not a Numbers Game: Minimum Cost of N-Deterrence", The Times of India, December 7, 1998.

9. Gen. Sundarji n. 4.

10. "No Plans for N-Arms Race China: Brajesh", The Times of India, February 7, 2000.

11. "India Capable of Making Neutron Bomb", The Hindustan Times, August 17, 1998.

12. Bharat Karnad, "Going Thermonuclear: Why, With What Forces, at What Cost", U.S.I. Journal, July-September 1998, p. 315.

13. Ramesh Chandran, "India, Pakistan Increasing Stocks of Weapons Grade Plutonium, Enriched Uranium: Report", The Times of India, March 8, 2000.

14. Rajendra Prabhu, "US Publication Explodes Western Scientists' Myth", Observer of Business and Politics, July 25, 1998. (Source: The Bulletin of Atomic Scientists, July-August 1998.)

15. Cited in India Today, June 1, 1998.

16. W. P. S. Sidhu, "India Sees Safety in Nuclear Triad and Second Strike Potential", Jane's Intelligence Review, July 1998, pp. 22-25.

17. While the plutonium from commercial reactors producing electricity is not considered ideal for producing nuclear weapons, the United Kingdom is known to have used it to make nuclear weapons and some scientists are of the view that India attempted to do the same in the second lot of tests on May 13, 1998.

18. R. Ramachandran, "Pokhran II: The Scientific Dimensions", in Amitabh Mattoo (ed.), India's Nuclear Deterrent: Pokhran II and Beyond (New Delhi: Har-Anand Publications Pvt. Ltd., 1999), pp. 35-36.

19. n. 18. Ramachandran has based his calculations on an average plutonium production rate of 12 kg annually for the Dhruva reactor (at an average of 60 percent of rated capacity) and 4 kg per annum for CIRUS (at 50 percent of rated capacity). Together, the total fissile material stockpile is 120 kg from Dhruva (12 kg per year for ten years) and a nearly similar amount from CIRUS (4 kg per year for almost 40 years). As approximately 8 kg of plutonium is required for each fission bomb, the total weapon stockpile would be limited to about 30 bombs.

20. Ramachandran, n. 18.

21. Raj Chengappa and Manoj Joshi, "Hawkish India", India Today, June 1, 1998.

22. "Stockpile", The Times of India, May 31, 1998.

23. George Perkovich, India's Nuclear Bomb (New Delhi: Oxford University Press, 1999), p. 428.

24. From Surprise to Reckoning: The Kargil Review Committee Report (New Delhi: Sage Publications. 2000), p. 260.

25. K. Subrahmanyam, "India's Nuclear Truth", The Times of India, January 26, 2000.

26. Aziz Haniffa, "Nothing Pak About it, Bomb Purely China Baby: Report", The Economic Times, June 19, 1998.

27. K. Subrahmanyam, "Gospel According to Lucifer", The Economic Times, July 10, 1998.

28. "China, N Korea Aided Pakistan's Missile Programme, Says CIA", The Times of India, February 4, 2000.

29. "Pak N-Plan Depends on Foreign Aid: CIA", The Hindustan Times, February 4, 2000.

30. "H-Bomb Whenever Govt Wishes: Khan", The Economic Times, June 9, 1998.

31. Frank Barnaby, "Discrepancies Claimed in Islamabad's Nuclear Tests", Jane's Defence Weekly, June 10, 1998.

32. Aziz Haniffa, "Pak Can Make 16-20 Nukes Even After Tests: ISIS", Observer of Business and Politics, June 5, 1998.

33. Sidhu, n. 16.

34. Chengappa and Joshi, n. 21.

35. n. 22.

36. Ramchandran, n. 18.

37. Brigadier Vijay K. Nair (Retd.), "The Structure of an Indian Nuclear Deterrent", n.18.

38. Ramesh Chandran, "Pak N-Might Bigger Than India's, Says US", The Times of India, June 8, 2000.

39. Manoj Joshi, "The Nuclear Maharaja has no Clothes", The Times of India, June 9, 2000.

40. "Pakistan Calls its Nuclear Arsenal 'Modest", The Times of India, June 9, 2000.

41. Dr. Sanjay Badri-Maharaj, "Nuclear India's Status: Examination of the Claims in the NBC Report", Indian Defence Review, April-June 2000.

42. R. Prasannan, "Stealing a March", The Week, July 9, 2000.

43. Ibid.

44. SIPRI Yearbook 1999: Armaments, Disarmament and International Security (Oxford: Oxford University Press, 1999), p. 555.

45. Source: Extracts from a Paper by Dr. Karl Z. Morgan, The Hindustan Times, June 8, 1998.

46. Michael D. Swaine and Ashley J. Tellis, Interpreting China's Grand Strategy: Past, Present and Future (Santa Monica, California: RAND, 2000), p. 123.

47. China has 100 x DF-11 and 300 x DF-15 short-range missiles. The Military Balance 1999-2000 (Oxford: Oxford University Press, 2000), pp. 175-177.

48. Brahma Chellaney, "Missile Muscle is the New Definition of World Power", Asian Age, April 16, 1999.

49. K. Subrahmanyam, "Agni-II: Modern Divya Astra", The Economic Times, April 15, 1999.

50. K. Subrahmanyam, "Agni-Ghauri Tests no Cause for Alarm", The Times of India, April 14, 1999.

51. Jasjit Singh is of the view that "nearly three dozen tests would be necessary before a missile can be considered operationally reliable… our aim should be to undertake at least two dozen tests of the Agni class missile over the next five years…" India's Defence Spending: Assessing Future Needs (New Delhi: Knowledge World, in association with Institute for Defence Studies and Analyses, p. 191.

52. Lieutenant General V. R. Raghavan (Retd.), India's Need for Strategic Balance (New Delhi: Delhi Policy Group), p. 22.

53. Srinand Jha, "Serial Production of IRBMs by 2000", Observer of Business and Politics, August 12, 1998.

54. N.C. Menon, "Pak N-Arms Superior to India: Report", The Hindustan Times, June 9, 2000.

55. Pravin Sawhney, "Missile Control in South Asia and the Role of Cooperative Monitoring Technology", CMC Occasional Papers (Albuquerque, New Mexico: Cooperative Monitoring Centre, Sandia National Laboratories, 1998), p. 36.

56. While Richard Barlow, a CIA operative in Pakistan, first reported Pakistan's nuclear weaponisation in 1987, in 1997, Gordon Oehler resigned as Director of the CIA's non-proliferation centre and reported that Pakistan had developed the Ghauri missile with Chinese help. See K. Subrahmanyam, "Ghauri Against Prithvi", The Economic Times, December 22, 1997.

57. Paul Bracken, "The Second Nuclear Age", Foreign Affairs, Vol. 79, No. 1, January-February 2000.

58. Vijay Kumar, "What is Agni-II All About", The Times of India, April 13, 1999.

59. Dinesh Kumar, "India Can't Thwart Chinese Nuclear Strike: Report", The Times of India, March 18, 1999.

60. Ramesh Chandran, "India Under Chinese N-Shadow: US Daily", The Times of India, March 16, 1999.

61. Group Captain A. K. Sachdev, "Pakistani Missiles: Their Pertinence to the Indo-Pak Conflict". Quoted with the author's permission from the Paper presented at IDSA Fellows Seminar on July 1, 2000.

62. K. Subrahmanyam, "India's Response", in K. Subrahmanyam (ed.), India and the Nuclear Challenge (New Delhi: Lancer International in association with Institute for Defence Studies and Analyses, 1986), p. 276.

63. Subrahmanyam, n. 8.

64. K. Subrahmanyam, "Nuclear Force Design and Minimum Deterrence Strategy for India", in Bharat Karnad (ed.), Future Imperilled: India's Security in the 1990s and Beyond (New Delhi: Viking Penguin India, 1994), pp. 189 and 193.

65. K. Subrahmanyam, "China and Nuclear Rationale", The Economic Times, July 26, 1997.

66. Subrahmanyam, n. 8.

67. Jasjit Singh, "A Nuclear Strategy for India", in Jasjit Singh (ed.), Nuclear India (New Delhi: Knowledge World, in association with Institute for Defence Studies and Analyses, 1998), p. 315.

68. Maharajkrishna Rasgotra, "Countering Nuclear Threats", in Brahma Chellaney (ed.), Securing India's Future in the New Millennium (New Delhi: Orient Longman, 1999), pp. 238-239.

69. General K. Sundarji, "Imperatives of Indian Minimum Deterrence", Agni, May 1996, p. 21.

70. Brigadier Vijay K. Nair (Retd.), Nuclear India (New Delhi: Lancer International, 1992), pp. 170-182.

71. Rear Admiral Raja Menon, A Nuclear Strategy for India (New Delhi: Sage Publications, 2000), pp. 177-234.

72. Ibid.

73. Rear Admiral Raja Menon, "The Nuclear Doctrine: Yoking a Horse and Camel Together", The Times of India, August 26, 1999.

74. Karnad, n. 12.

75. Karnad, n. 12.

76. "India Can Produce N-Bomb of 200 Kiloton: Chidambaram", The Times of India, May 23, 1998.

77. Interview with the author, July 1, 2000.

78. Lieutenant General Pran Pahwa (Retd.), "Organisation and Employment of Strategic Rocket Forces", USI National Security Series 1998 (New Delhi: United Service Institution of India, 1999), pp .294-296.

79. Kenneth N. Waltz, "What will the Spread of Nuclear Weapons do to the World?" in John K. King (ed.), International Political Effects of the Spread of Nuclear Weapons (Washington, D.C.: United States Government publication, 1979), p. 188. Cited by General K. Sundarji, n. 4.

80. Bernard Brodie, Strategy in the Missile Age (Princeton, New Jersey: Princeton University Press, 1959), pp. 275-276. Cited by Bharat Karnad, "A Thermonuclear Deterrent", n. 18.

81. Herman Kahn, On Thermonuclear War (New York: The Free Press, 1969), pp. 482-483.

82. Interview with the author, June 30, 2000.

83. Interview with the author, July 20, 2000.

84. Interview with the author, July 20, 2000.

85. Interview with the author, July 10, 2000.

86. For a detailed analysis of the efficacy of megaton warheads, see Herman Kahn, n. 81.

87. Solly Zuckerman, Nuclear Illusion and Reality, p. 69. Cited by Bharat Karnad, n. 18.

88. G. Balachandran, "Nuclear Weaponisation in India", Agni, vol. V, no. 1, January-April 2000, p. 37-50.

89. The Harvard Nuclear Study Group: Albert Carnesale, Paul Doty, Stanley Hofmann, Samuel P. Huntington, Joseph S. Nye, Jr., and Scott Sagan, Living With Nuclear Weapons (New York: Bantam Books, 1983), p. 34. Cited by Bharat Karnad, n. 18.

90. Balachandran, n. 88.

91. G. Balachandran has estimated that to destroy eight soft area targets a total of 800 missiles would be necessary if the missiles were armed with a 20 Kt fission warhead and had a CEP of 1,000 metres (100 missiles per target). The plutonium required for such an attack would be nearly 2,500 kg. n. 88.

92. Balachandran, n. 88.

93. One method of reducing the number of missiles necessary to achieve the required assurance levels would be to develop multiple, independently targetable reentry vehicle (MIRV) technology. Each missile could then carry four to six warheads. However, at present India lacks the technological sophistication necessary to develop MIRV warheads; nor is a suitable missile on the way.

94. Raj Chengappa has quoted Defence Minister George Fernandes to have stated in 1999 that "the (nuclear) submarine project would take another four years to build". Even if that turns out to be a realistic timeframe, SLBMs having ranges in excess of 5,000 km are still likely to take much longer to develop. Raj Chengappa, Weapons of Peace (New Delhi: Harper Collins, 2000), p. 436.

95. Rajesh Rajagopalan, "The Advocates of Armageddon: Nuclear War and the Victory Theorists", Strategic Analysis, vol. XI, no. 8, November 1987, pp. 935-936.

96. Donald R. Carter, "Peacetime Operations: Safety and Security", in Ashton B. Carter, John D. Steinbruner and Charles A. Zracket (eds.), Managing Nuclear Operations (Washington, D.C.: The Brookings Institution, 1987), p. 26.

97. Jasjit Singh, "Ensuring Peace and Development with N-Capability", Tribune, August 15, 1998.