"In my home there are few publications that we actually get hard copies of, but [Arms Control Today] is one and it's the only one my husband and I fight over who gets to read it first."

– Suzanne DiMaggio
Senior Fellow, Carnegie Endowment for International Peace
April 15, 2019
Nuclear Stability and Arms Sales to India: Implications for U.S. Policy
Share this

Eric Arnett

Eric Arnett is leader of the Stockholm International Peace Research Institute's Project on Military Technology and editor of Nuclear Weapons and Arms control in South Asia after the Test Ban, forthcoming, by Oxford University Press.

The news from South Asia has been surprisingly good in this, the 50th anniversary year of India's and Pakistan's independence. Despite earlier signs that both countries would continue to be governed by perennially weak governments, 1997 has seen the strengthening of democratic institutions in Pakistan and the emergence of surprisingly decisive and conciliatory leaders on both sides: Inder Kumar Gujral, the new Indian prime minister, and Nawaz Sharif, the Pakistani prime minister re elected with a confidence inspiring majority after Benazir Bhutto's government was dissolved by President Farooq Leghari. Sharif's position was further strengthened when the provision allowing the president to dissolve the government was eliminated. Gujral and Sharif appear to have a better chance of bringing a sturdy peace to the region than any of their predecessors. This is as much a result of their personal abilities and a creeping recognition of the status quo after more than 25 years without a war as the expectations of those who believe in nuclear deterrence or the inevitability of peace among democracies.

Still, not everything is going the right way at this hopeful moment. Politically, both prime ministers seem to derive their strength in part from a willingness to play to the consensus in their respective polities against meaningful nuclear arms control. Militarily, ballistic missile programs are being reinvigorated in both countries due as much to the spirit of nuclear populism as to any realistic threat to the security of either state.

For its part, the United States, understandably eager to engage both countries as they pursue a renewed bilateral dialogue, has apparently decided that the transfer of military technology is a useful inducement to offer India. While technology transfer is a prize which the United States can offer in a quantity and quality unmatched by most of its economic competitors, greater care must be taken than has been the case up until now in offering military technology to New Delhi. Transfers should take into account India's nuclear weapon capabilities, as well as the region's strategic dynamics, but so far signs are that U.S. policy makers and their advisors tend to think the details of the South Asian conventional balance do not matter much.

In fact, they matter quite a bit, both for the sake of regional stability, should Indian Pakistani relations deteriorate, and for the success of the U.S. goal of capping the nuclear weapon and ballistic missile capabilities of the two states. The Clinton administration has already allowed the transfer to India of potentially destabilizing conventional weaponry—guidance kits for "smart" bombs that could be used in conventional counterforce attacks on Pakistan's nuclear delivery systems. With other transfers of concern possibly in the works, it is vital that guidelines covering transfers of military technology to South Asia be considered more carefully.


Nuclear Stability in South Asia

With the end of the Cold War and the opportunity for improved U.S. relations with both India and Pakistan, some proponents of increased arms sales to New Delhi have advocated accepting the purported stability of nuclear deterrence between India and Pakistan and using arms and technology transfers as a means to create a "strategic partnership" with India.1 Unfortunately, there is good reason to believe nuclear deterrence is not stable in South Asia and that U.S. arms transfers have already made the problem worse.

The claim that a stable deterrent is emerging in South Asia and can be managed with judicious transfers of technology is rooted in an oversimplified model of Indian and Pakistani strategic planning, viewed through the distorting lens of the Cold War relationship between the United States and the Soviet Union. In the later years of the Cold War, the superpowers enjoyed a sort of nuclear parity and pursued their rivalry through proxy conflicts on the territories of other states with little risk of direct conventional war. In South Asia today, the relationship between India and Pakistan is still one between a regional hegemon and a weaker neighboring state dissatisfied with the status quo. Their rivalry is pursued in part through proxy wars on each other's territory which could escalate into another conventional war despite the presence of nuclear weapon capabilities and improving democratic conditions.

Significantly, Pakistan's nuclear weapon capability may actually contribute to the risk of war. The Pakistani leadership shows indications of believing that the risk of nuclear war makes even small conventional conflicts extremely unlikely if not impossible, and that South Asia's nuclear stalemate gives both sides the cover to continue the struggle in Kashmir by other means. India has the capability to cut off Pakistani support to Kashmiri insurgents if it becomes too blatant or too successful, but Pakistani officials have been claiming since the late 1980s that India would not dare take such a step because of Pakistan's nuclear option. Speaking in the context of a nuclear umbrella over Pakistani support to the Kashmir insurgency, Pakistan's former chief of army staff, Mirza Aslam Beg, declared in 1993, "There is no danger of even a conventional war between India and Pakistan.... There is no possibility of an Indian Pakistan war now."2

While some Western observers have been taken in by this logic, there is little evidence that the Indian armed forces have accepted it and every reason to believe that an escalation in Pakistani support to insurgents on Indian territory will be dealt with by military means if deemed necessary. The insurgency can only succeed if it inflicts intolerable costs on India, but New Delhi is less likely to capitulate when it still has the options of "hot pursuit" and interdiction, both legitimate actions under international law. The resulting mismatch of perceptions makes war more likely.


Conventional War Planning

While the current trend in South Asia toward strengthening democratic institutions and re establishing a bilateral dialogue on conflict resolution is cause for hope, U.S. policy cannot be—and is not—based on the assumption that there is no longer a danger of war between India and Pakistan. Nevertheless, the Clinton administration has allowed transfers of military technology to both India and Pakistan. Transfers of conventional arms are not always or inherently destabilizing, nor are the technologies of greatest concern necessarily those with the highest profile or price tag. To relate arms transfers more coherently to nuclear stability in South Asia, it is important to understand how the Indian military thinks a fourth Indo Pakistani war might unfold.

While Pakistani leaders might believe that their nuclear weapon option gives them a free hand to pursue measures short of war in Kashmir, Indian military planners are not buying it. There are strong indications that Indian military planners do not take the Pakistani nuclear capability seriously and continue to plan for conventional war. The top priority of the Indian Air Force (IAF) in such a war would be destroying Pakistani air bases, which is where Islamabad's nuclear delivery systems (though probably not nuclear warheads) are stored. Air bases are a legitimate target in conventional war, and have always been the first object of attack in wars of the air age, including the 1965 and 1971 Indo Pakistani wars. If India and Pakistan went to war over Kashmir as they did in 1965, there is a good chance that the conflict would begin (as it did in 1965) with immediate air base attacks by both sides.

The presence of nuclear weapon capabilities apparently has not changed this calculus for the IAF. Quite the contrary. As early as 1979, D. K. Palit, the commandant of the Indian Military Academy, was already outlining a doctrine of damage limitation through conventional counterforce:

India's defensive policy against a likely nuclear conventional attack by Pakistan must aim, at first priority, to minimise the nuclear threat. In this case, Pakistan's weak point will be its delivery system, because for a considerable time to come its only recourse will be the fighter bomber.3

P. R. Chari, a former Ministry of Defense official with responsibility for the IAF who drafted many of the ministry's annual reports, suggests that India is relatively unconcerned by the Pakistani nuclear option because India's nuclear preponderance, among other factors, would deter Pakistan from first use. The result, according to Chari, is a need for conventional forces to fight below the nuclear threshold.4

Preparations for that contingency are being pursued by the IAF with the blessing and backing of the Ministry of Defense. Like its American counterpart, the IAF has long preferred the battle for air superiority—fought primarily over the enemy's air bases—to the nitty gritty of direct support for ground forces. Strike aircraft have been the highest priority in India's tight military modernization budget, not as strategic counterforce systems, per se, but as the most effective tools for gaining air superiority as part of India's doctrine of "offensive defense."

While most of the Indian armed forces struggle in the face of funding constraints imposed by the budget crisis, the IAF's 22 strike squadrons have fared well. During the past 25 years, the IAF has received more than 700 MiG 21, MiG 23, MiG 27, Mirage 2000 and Jaguar aircraft used primarily for air to ground missions; approximately 300 of these strike aircraft remain in active service. In a conflict with Pakistan, most of these aircraft would be used to attack Pakistan's 17 air bases (of which about half are primary bases and half are "dispersal" bases), and nine other airports capable of handling jets from the first day of a war.

In addition to smart bombs, IAF attack aircraft have been armed with runway cratering munitions and area denial bombs (which prevent runways from being repaired), as well as unguided gravity bombs and cluster weapons. If recent writings are to be believed, even the Prithvi short range ballistic missile is seen by the IAF primarily as a conventionally armed, air base attack weapon that could destroy aircraft in the open and hamper runway repairs.5

During the 1991 Gulf War, IAF officials watched with particular interest as U.S. strike aircraft destroyed Iraqi hardened aircraft shelters with 2,000 pound smart bombs. Because destroying Pakistani aircraft on the ground is so important to the IAF, Indian military planners had been frustrated by the practically invulnerable sanctuary created for the Pakistan Air Force (PAF) in the form of hardened aircraft shelters, built with U.S. assistance in the 1980s. IAF planners quickly set out to buy heavy, shelter busting smart bombs like the ones demonstrated in the Gulf War. In 1992, India closed its first arms deal with Russia that included an unknown number of 2,200 pound smart bombs. These were probably copied from the U.S. Paveway I, deployed in Vietnam in the 1970s, and could only be used with India's aging fleet of MiG 27s, only 36 of which are still operational according to a recent report.6

The IAF, eager to acquire better technology that would be compatible with its 88 Jaguars and 35 Mirages—which would bear the main responsibility for defeating the PAF—did not have to wait long. In 1994, 315 Texas Instruments Paveway II guidance kits—one of the types used by the U.S. Air Force in 1991—were delivered for installation on 2,000 pound British bombs. French and Israeli contractors assisted the IAF in its initial experience with the laser guided weapons. Then, in April 1997, U.S. manufactured smart bombs were delivered to the IAF for the first time. They are now a standard part of IAF air base attack exercises.

Indian attacks on Pakistani air bases would no doubt target the building at the Sargodha air base in which Pakistan's ballistic missiles are believed to be stored. Because the attacks would be conventional attacks on legitimate military targets, the Indian leadership would not expect Pakistan to escalate to nuclear first use, especially given India's capability of retaliating "in the range of ten megatons for one," in the words of a retired Indian chief of army staff.7


Nuclear Escalation

If a conventional war continued long enough, most of the PAF's strike force could be destroyed—even if the IAF was not intentionally targeting Pakistan's nuclear capability—a danger of which Pakistani military planners are becoming increasingly aware. At an April 1997 briefing to journalists, a PAF official expressed doubts that the armed forces could hold up for more than six to eight weeks under IAF plans "to neutralize [Pakistani] radars and [surface to air missiles] and destroy the Pakistan air force on the ground and in the air."8 According to another official, the arguably better trained PAF pilots could not compensate for the "expected high attrition [on] the ground and [in the] air...buzzing with advanced guided [weapons].... The IAF has a tremendous edge in numbers and in the quality of weapons."9

If a war started in the near future, the IAF's poor state of readiness might limit the extent of the damage it could actually inflict. On the other hand, the PAF has in the past fled from air battles with the IAF, granting India air superiority by default. This makes the success of any Indian air campaign against PAF bases difficult to predict. But even if the IAF did not entirely eliminate Pakistan's nuclear delivery capabilities (a possibility that reduces India's incentive to risk a preventive war to destroy Islamabad's nuclear option without some other provocation), these capabilities would still be eroded. Furthermore, the Pakistani leadership's understanding of the unfolding battle would be undermined by IAF attacks on radar sites carried out with lighter smart bombs and Armat anti radar missiles, which were acquired from France in the late 1980s. In 1971, even without the Armat, the IAF did considerable damage to the Pakistani radar network with unguided bombs.

India's capability to erode Pakistan's nuclear option is particularly troubling because of a penchant for early first use of nuclear weapons that appears to be taking hold in Islamabad. For example, in 1990, then President Ghulam Ishaq Khan told a U.S. official that "in the event of war with India, Pakistan would use nuclear weapons at an early stage," despite the risk of Indian retaliation and regardless of the war aims.10 In seeking to promote the deterrent benefits of the country's nuclear option, Pakistani leaders have recognized that they must be perceived to be prepared to use nuclear weapons even if such action seems irrational. For example, Asad Durrani, the former director of Inter Services Intelligence, said Pakistan can only hope to deter wars for limited aims if Indian planners believe "we are primed, almost desperate, to use our nuclear capabilities when our national objectives are threatened, for example, a major crackdown on [the] freedom movement in Kashmir."11

While Indian planners obviously do not believe these Pakistani claims regarding nuclear use, the danger is that their Pakistani counterparts might talk themselves into such a strategy. In particular, even without attacking any of Pakistan's nuclear installations covered by their 1989 bilateral agreement (which obligates both states to foreswear such attacks), India could so erode Pakistan's nuclear delivery capabilities that Islamabad would face a "use it or lose it" dilemma. Worse still, with the destruction of parts of the country's warning and command and control networks, the Pakistani leadership might fear it had reached that point, even if it had not. Obviously, it is difficult to imagine what Pakistan could hope to gain by launching a nuclear attack in a war for limited aims, other than the world's scorn for being the first state to use nuclear weapons in more than 50 years and the possibility of Indian nuclear retaliation. Nonetheless, recent statements by spokesmen on both sides suggesting that Pakistan believes the nuclear threshold in South Asia is much lower than does India are a cause for concern.


Controlling Military Transfers

The Clinton administration's understated response to India's emergent conventional counterforce capability is not only a product of the relative secrecy of the Paveway transfer; it also stems from a U.S. tendency to see the specifics of conventional arms transfers as unimportant because of the nuclear factor. When asked why the administration approved the Paveway transfer, given its recognition of the risk of war in South Asia and its policy of not introducing destabilizing new capabilities into the region, one U.S. official said the go ahead had been given because India already possessed similar technology. But this U.S. technology is certainly more advanced than that of its Russian competitor, and although France has supplied similar technology, it is associated with lighter weapons and is not as advanced.

Proponents of the Texas Instruments sale also argued that the Paveway kits were simply components supplied to Britain for the bombs it was delivering to India. In fact, the kits were the main reason why New Delhi bought the British system, which otherwise is simply a "dumb" bomb that India could have made itself. Because the value of Texas Instruments' deal was less than $50 million and the Paveway kit is not considered major military equipment, the Clinton administration was not obliged to notify Congress of the proposed transfer, suggesting that notification requirements are not strong enough for some types of high tech military hardware and for certain recipients.

Another rationale for the Paveway transfer—not cited by any U.S. official—is popular among some military analysts but, ultimately, is not persuasive. Their claim is that the IAF is simply not up to the job of using military technologies like the Paveway effectively, so there is no reason to worry. Aside from the troubling implications of basing policy decisions on the assumption that any weapon can legitimately be sold to states if their armed forces are poorly maintained or incompetent—that the content of conventional arms transfers does not really matter—this line of argument overlooks two important considerations.

First, Pakistani planners may not be as sanguine about Indian capabilities, and the fears and perceptions of Pakistanis are the central issue. From Islamabad's perspective, it increasingly appears that most of the major arms suppliers are cooperating with India, even those that have already sold systems to Pakistan. Some of the systems that have been supplied, especially the Armat anti radar missile, can be very effective even in the hands of less skilled pilots. The quality and likely effectiveness of other systems are very difficult to judge without much greater transparency on the part of the IAF.

Second, Indian society is set to progress at an unprecedented rate, potentially enjoying economic growth and modernization that are likely to dramatically increase the budget and competence of the country's armed forces. As relations with China improve, even more of India's military potential can be focused on Pakistani contingencies. Even if IAF strike squadrons have some weaknesses now, U.S. policy makers should not be betting against substantial improvements in Indian technological proficiency in the near future, especially since U.S. firms are contributing to the technological base of both the civilian and military sectors.

Indeed, despite the common perception in India and the United States that there is little cooperation between the two countries on military technology, there has been an effort to increase it since the Reagan administration. In 1983, India launched its Light Combat Aircraft (LCA) program to develop a fighter to replace its aging fleet of MiG interceptors. The LCA is touted as a completely indigenous design, but it is actually an odd amalgamation of subsystems from Britain, France, Italy, Sweden and, primarily, the United States. U.S. firms are not only supplying the engines, flight control system and several other subsystems, they are providing expertise in systems integration and a flying testbed for subsystems designed in India. U.S. participation in the LCA project stands in stark contrast to export control policy in the 1970s, when the United States actually refused to allow Saab to sell its Viggen fighter to India because of its U.S. supplied engine. U.S. firms have also contributed technology to India's land and naval forces.


Toward New Guidelines

U.S. participation in the LCA project and the sale of the Paveway kits to the IAF should not only put to rest the idea that there is no arms trade between the United States and India, but they point up the dangers of expanding U.S. arms sales—even under a policy that nominally forbids transfers of destabilizing technologies—unless the South Asian strategic context is re evaluated. A policy of expanded transfers—based on the assumption that nuclear deterrence makes war unlikely—is likely to lead to more reckless policy implementation rather than the restraint and responsibility needed now in the South Asian security environment.

In seeking to convince the Indian and Pakistani governments to cap their nuclear programs, some observers have been tempted to congratulate them for not falling under the spell of the logic that drove the superpowers' Cold War buildups. While some problems for the triangular relationship among Islamabad, New Delhi and Washington have arisen from false analogies, a case can be made for applying certain concepts developed by arms controllers during the Cold War more systematically to the case of South Asia.

One example is the principle that counterforce systems—both nuclear and conventional—have an inherently destabilizing aspect and should therefore be treated with care. While the Clinton administration has been sensitive to the counterforce problem, it seems to have mistakenly concluded that only nuclear counterforce capabilities—in the specific form of ballistic missiles—are of concern. In fact, conventional counterforce technologies should be of greater concern because they need not be perfectly effective in a first strike. A counterforce attack with nuclear weapons must succeed in an almost perfect first strike—something far beyond India's capabilities or those of any country—because it would remove the onus of first nuclear use from the adversary. Conventional counterforce leaves the onus on the other side, while increasing the pressure under which choices between surrender and first use of nuclear weapons must be made. Even if ballistic missiles were much more effective than they are, short range missiles are less of a counterforce problem than longer range combat aircraft, because retaliatory forces can simply be moved out of range. India's short range Prithvi missile cannot reach large areas of Pakistan, especially if launched from secure areas some distance from the border.

The main risk posed by nuclear armed ballistic missiles is that they entail a risky devolution of launch authority in cases where a robust communications link cannot be counted on. This risk is greater if the command and control system is gradually being destroyed by conventional attacks or fails catastrophically. (There is evidence that Pakistan does not take the survivability issue seriously enough.) Aircraft, on the other hand, have greater ranges and are more effective delivery systems. In South Asia, as in most parts of the world, they cannot be defended against any more than ballistic missiles and warning times are short to non existent, especially after radars have been attacked.

That leads to a second arms control principle developed during the Cold War: strategic defenses can also be destabilizing, especially if there is some uncertainty about their effectiveness. Again, in the South Asian context the main concern is not ballistic missile defense but air defense. In the near future, India is expecting to buy or build a full range of air defense and air to air capabilities, including the Light Combat Aircraft, 40 multi role Su 30 fighters on order from Russia, a variety of missiles, and airborne warning and control system (AWACS) aircraft. These would not only create the possibility—or at least the potentially destabilizing perception—of a highly effective defense against a nuclear strike, but better air to air capabilities actually make India's conventional strikes more effective. An AWACS capability would make it possible for India to achieve air superiority all the more quickly and completely. With air superiority, aircraft can drop their smart bombs with less fear of coming under attack while guiding them to their target, thereby offering a greater probability of a successful strike.

Furthermore, an AWACS system can enable Indian planners to keep track of which Pakistani aircraft are at which bases, helping them to prioritize targets. To make matters worse, the actual capabilities of an AWACS are very difficult to judge, so Pakistani planners will have reason to err on the side of worst case scenarios. PAF officials seem to have an exaggerated regard for AWACS following the U.S. denial of their request for E 2 aircraft in the 1980s. An important new goal of export control for the region is therefore to inhibit India's ability to achieve an air defense system with strategic consequences. The reported deal that would supply India with four AWACS aircraft co manufactured by Israel and Russia is of particular concern.

The goal of such a measure is not to hobble India's ability to defend itself and wage conventional war if the Kashmir situation deteriorates. The ground forces central to a plausible war scenario are not of comparable concern. Rather, the aim is to avoid new sources of instability and to save the initiative to cap Pakistan's nuclear option somewhere short of the deployment of warheads on missiles. If Indian pressure on the survivability of the Pakistani nuclear delivery systems intensifies—as has occurred with the cooperation of almost every major arms supplier—the result could be not only open deployment of nuclear tipped ballistic missiles, but a resumption by Islamabad of the production of fissile materials to increase the number of potential weapons and perhaps a nuclear test to certify the performance of a missile warhead. Such a development would surely have serious implications for full implementation of the Comprehensive Test Ban Treaty and the initiation of negotiations leading to a fissile material cutoff treaty. Clearly, the stakes reach beyond the stability of South Asia to the web of treaties and norms that seek to regulate nuclear behavior globally.


1. See for example, Richard N. Haass, et al., A New Policy Toward India and Pakistan, Washington, D.C.: Council on Foreign Relations, 1997.

2. Quoted in G.F. Giles and J.E. Doyle's, "Indian and Pakistani Views on Nuclear Deterrence," Comparative Strategy, April/June 1996, p. 146.

3. D.K. Palit and P.K.S. Namboodiri, Pakistan's Islamic Bomb, New Delhi: Vikas, 1979, p. 117.

4. P.R. Chari, Indo Pak Nuclear Standoff: The Role of the United States, New Delhi: Manohar, 1995.

5. J.P. Joshi, "Employment of Prithvi Missiles," Journal of the United Services Institution of India, October/December 1996.

6. "India Postpones MiG 27 Upgradations," The Hindu, June 11, 1997.

7. S. Gupta and W.P.S. Sidhu, "The End Game Option," India Today, April 30, 1993.

8. Air Marshal Ayaz Ahmed Khan (Ret.), "Challenge of the Indian Air Threat," The Nation, April 7, 1997.

9. Ibid.

10. Hamish McDonald, "Destroyer of Worlds," Far East Economic Review, April 30, 1992, p. 24.

11. A. Durrani, Pakistan's Security and the Nuclear Option, Islamabad: Institute for Policy Studies, 1995, p. 92.