U.S. Intelligence on South Asian Conflict Scenarios

During 1979 and 1980, U.S. intelligence was learning more about Pakistan’s secret program to produce fissile material for nuclear weapons, in part by using purloined gas centrifuge technology. That development amplified concern in Washington about nuclear proliferation and the implications for stability in South Asia. Necessarily India was monitoring the Pakistan nuclear situation even more closely because the potential threat it posed during the years ahead. By 1981, with Pakistani nuclear capabilities moving ahead, U.S. government officials wondered whether Pakistan’s nuclear activities would motivate preventive action by India. Evidently, that concern led to the commissioning of a national intelligence estimate assessing possible Indian reactions. The National Intelligence Council had approved a proposal for an intelligence estimate on the topic in the spring of 1981, although it had been under discussion for some months. By November 1981, the result was Special National Intelligence Estimate (SNIE) 31/31-81, “India’s Reaction to Nuclear Developments in Pakistan.”

The version of the SNIE published here was the result of a mandatory declassification review request by the National Security Archive. When the document appeared on the CIA’s FOIA website around 2015, it received much attention in the Indian mass media under dramatic headlines, for example, “Indira Gandhi planned attack on Pakistan’s nuclear facilities,” (India Today, 1 September 2015). In fact, no such claim appears in the estimate (although Document 3, below, indicates the preparation of contingency plans).

The SNIE’s exposition consisted largely of “ifs” and possibilities, although it drew on some factual reporting on internal discussions in Indira Ghandi’s government. India had staged a “peaceful nuclear explosion” in 1974 but it had not made decisions to produce nuclear weapons, much less stockpile them. Thus, a Pakistan armed with nuclear weapons could pose a significant security threat. The estimate pointed to recent detailed internal recommendations circulating within the government for “exercising the nuclear weapons option,” but analysts added that “we do not know whether the Prime Minister favors that course.”

With all the uncertainties, the estimate found that India would probably take a “wait and see” approach “based on whether Pakistan produced nuclear weapons or stockpiled them, which would then lead New Delhi to face a choice of accepting the high probability of a nuclear arms race or destroying Pakistan’s nuclear facilities.” If the Indian government was determined to uphold its non-weapons status, it would have to decide whether the threat of a nuclear-armed Pakistan was severe enough to make it necessary to destroy that country’s nuclear facilities. Yet, such action could risk Chinese intervention on behalf of Pakistan. Thus, a “final Indian decision is likely to depend critically on political and strategic circumstances prevailing at that time.”

The report closes with an annex on how Indian leaders might think about the development of a small nuclear arsenal, both the pros and the cons. On the positive side was to need to improve the military balance with China and to encourage the “superpowers to accord India greater respect.” On the negative side was that India did not yet have “unsafeguarded” stocks of plutonium, and the fact that a weapons stockpile would add to military expenditures, especially when considering the costs of delivery systems and “hidden costs,” such as for early warning systems.

Readers of SNIE 31/31-81 of 8 September 1981 (See Document 1) may have wondered, if India did attack one of Pakistan’s nuclear facilities, what would be the effects? That led to the commissioning by the Energy Department and the State Department’s Political-Military Affairs Division of a study on the radiological effects of an attack on the Karachi Nuclear Power Plan (KANUPP). Lawrence Livermore Laboratory’s Special Projects Division (later known as the “Z” Division), which specialized in nuclear proliferation topics, conducted the study, and Alden F. “Jerry” Mullins was its project leader and gave a briefing to State Department officials.

According to Mullins, the “most lethal consequences would occur only if the fuel in KANUPP’s core melted down,” the “worst-case scenario,” where a 2,000-pound bomb blew a “one square meter hole in the thick concrete and steel containment shell,” with one-sixth of the total iodine inventory of 800,000 curies of I-131 escaping through the hole. “Under these extreme conditions, Karachi residents would probably receive a 10 to 20 REM [Roentgen Equivalent Man] inhalation dose as far as 25 km down wind,” which would be “20 times as great as the serious escape of iodine during an accident at a UK reactor at Windscale in the 1960s.”

Under the worst-case circumstances, Karachi “could suffer heavy casualties”: “The worst result of iodine inhalation is latent thyroid cancer, which may take 15 to 30 years to develop.” With about ten percent of thyroid cancers being fatal, Mullins saw a “70% probability that there will be 10 additional latent thyroid cancers in Karachi as a result of the attack; a 60% chance there will be 100 cancers; a 50% chance there will be 200; a 30% chance there will be 1,000; and a 10% chance there will be 20,000.” The impact of iodine releases was predictable, but Mullin saw other potentially grave consequences: the release of other radioisotopes, “prompt radiation” effects (such as gamma radiation), and the impact of blast and heat in the immediate vicinity.

The study noted that, if an attack did occur, Pakistan could put before the world its “worst-case projections” of total casualties. “The final figure could be a huge one,” with significant “political and propaganda impact.” For its part, India could develop damage estimates that “would probably be higher than Livermore,” which could “dissuade the Indians from hitting KANUPP and, along with other factors, might encourage them to go after softer targets.” Not mentioned was the recent Israeli attack on Iraq’s nuclear reactor, which produced widespread condemnation and could have influenced the thinking of India’s leaders to the extent that they were considering a preventive attack of this sort.

International law may have also had an impact on Indian policymakers. Article 56 of the 1949 Geneva Convention’s Additional Protocol I (1977) proscribed attacks on installations such as “dykes, dams, and nuclear electrical generating stations” that contain “dangerous forces,” which could put “civilians at risk from the release of radioactive material.” India (as well as Pakistan) was not a signatory to the Protocol, but the restriction nevertheless stood.[4] Another complication was that the KANUPP reactor was under IAEA safeguards, and an attack would have incensed the Agency by casting doubt on the efficacy of its work.

Six years later, no preventive Indian attack on Pakistani nuclear facilities had occurred. Yet, the Reagan administration saw a continued risk and the possibility that such a strike could ignite a war. According to this CIA report on decision-making on the Indian nuclear program, intelligence analysts believed that India had contingency plans for such an attack, which some Indian strategists saw as highly dangerous because the “inevitable Pakistani retaliatory attack on Indian nuclear facilities would spread radioactive contamination over large portions of India.” Other U.S. intelligence analysts argued that India’s drive for “regional hegemony” could prompt preventive attacks on Pakistani facilities, such as the Kahuta enrichment plant, whose destruction would minimize radiological dangers. Other U.S. intelligence analysts then saw the probability of Indian action as “low,” not only because of “traditional indecisiveness and caution” but also because Prime Minister Rajiv Gandhi had said “publicly that such a move would not ensure the eradication of fissile material and components.” Moreover, a preventive Indian strike would “almost certainly begin a war between India and Pakistan.”

In December 1985, Gandhi and Pakistan’s dictator General Muhammad Zia-ul-Haq met in India, where they reached an understanding not to attack each other’s nuclear facilities. According to this CIA report, “We speculate that [the agreement] may have been prompted, at least in part, by US importunings in late 1985 that the nuclear issue be resolved between India and Pakistan,” although other intelligence experts believed that it was something that Gandhi had been seeking on his own, even though it was “congruent with U.S. interests.”

Whatever the sources of the Gandhi-Zia understanding were, it remained informal until December 1988.[5] By then, India and Pakistan had gone through the 1987 Brass Tacks crisis, instigated by Indian military exercises, which had raised fears of war and renewed concern among Pakistani leaders about the possibility of an Indian attack on their nuclear facilities. For his part, Gandhi wanted to find ways to reduce nuclear risks. Those interests and concerns provided context for the signing of the December 1988 Non-Attack Agreement when Gandhi traveled to Pakistan for meetings with Prime Minister Benazir Bhutto.[6]

This 1989 report by State Department intelligence treated an Indian preventive attack on Pakistan’s nuclear complex as “increasingly unlikely,” not least because the “optimum time” for such action had passed. INR analysts downplayed the possibility of military action or war in general because “neither India nor Pakistan want to fight a fourth war.” With the military balance decisively in India’s favor,

Pakistan “would have little chance of winning” and India would “suffer considerable costs” and have “to decide how to deal with a defeated Pakistan.”

The leaders of both countries, Prime Minister Benazir Bhutto of Pakistan and Prime Minister Rajiv Gandhi of India, had met several times and developed a “personal rapport,” but they could do little to improve the hostile relationship between their countries. Both were in weak positions politically in countries where “suspicions and prejudices … form the bedrock of each country’s world view and security policy.” Adding to the potential for hostilities were “points of friction,” such as rival claims for Kashmir, including the Siachen Glacier, “the highest and bleakest battlefield in the world.” Also adding to friction were “mutual charges of interference”: that each country was providing arms and assistance to “separatist extremists” in the other, for example, alleged Pakistani support for Sikhs.

INR analysts saw some areas of improvement, such as the confidence building agreements, notably the 1988 Nuclear Nonattack Pledge. Although Gandhi’s critics saw this as and other steps that he had supported as an example of his “being ‘soft’ on Pakistan,” INR’s analysts suggested that it could “improve regional security somewhat by strengthening, legitimizing, and publicizing the commitment of the political leaders on each side not to resort to preemptive strikes as a means of dealing with their neighbor’s nuclear program.” Yet, it “could encourage further procrastination in confronting” the dangers of the nuclear problem.

A detailed account of the military situation between the two countries demonstrated India’s superiority in terms of troops, equipment, and supplies, with the buildup expected to continue. The “capabilities gap” made Pakistan’s situation even worse. Acknowledging that India faced a “potential threat from China,” INR experts noted that “Indian forces are still primarily oriented toward Pakistan.” In that context, Pakistan treated its “covert nuclear program as the ultimate defense against India.”

Along with Kashmir and the claims about internal interference, the nuclear programs of each country were the “most serious” bilateral problem. Neither country had produced nuclear weapons, but the only “restraints” to doing so were “political not technical.” For example, Pakistan could test a device “within a few weeks of a decision to proceed.” Having not made decisions to produce nuclear weapons, India “now faces the prospect of a nuclear-armed Pakistan without having its own nuclear deterrent in place.” If “mounting evidence [demonstrated] that Pakistan has crossed or is about to cross the nuclear weapons threshold,” such as testing a device, Indian leaders would have to make decisions to produce their own nuclear weapons.

As “unlikely” as the prospect of India-Pakistan conflict was, there was a “high” risk that through a “miscalculation or irrational response,” a “conventional war … might tum into a nuclear exchange.” Such an outcome could draw the superpowers into the situation. INR even included an Annex with detailed scenarios for three situations, 1) Indian preventive action against Pakistan’s nuclear complex, 2) India initiating a conventional attack, or 3) Pakistan initiates a conventional attack. In all three scenarios, Pakistan “sues for peace” only a few weeks after war breaks out.

In 1990, months after INR prepared this estimate, a major crisis broke out. Triggered by military exercises and instability in Kashmir, the region was on edge with growing fear of a nuclear confrontation. U.S. diplomats, with Deputy National Security Adviser Robert Gates undertaking a special mission, worked to help defuse tensions. U.S. influence was probably not decisive but it may have helped.[7]

The late Mary Shoemaker, a Foreign Service Officer with rich experience in South Asian issues overseas and at the State Department, prepared this report when she worked at INR’s South Asia desk. In 1962 when she became married, Departmental regulations forbade her from taking an overseas assignment, and she resigned from the Foreign Service. When the State Department lost an anti-discrimination lawsuit, she restarted her Foreign Service career. Future member of Congress Rush Holt also contributed to the report.

Four years after INR prepared its report, the U.S. intelligence community conducted a major estimate of the India-Pakistan situation and prospects for conflict which considered developments since 1989. Against the backdrop of the 1990 crisis, significant violence by Hindu nationalists against Muslims in late 1992 and early 1993, growing nuclear capabilities on both sides, and a new U.S. president, William J. Clinton, who was greatly interested in India, the U.S. intelligence community prepared a National Intelligence Estimate on “India-Pakistan: Prospects for War in the 1990s.” According to the title page, the NIE was “prepared under the auspices” of Bruce Riedel, National Intelligence Officer for Near East/South Asia. A former CIA intelligence analyst, Riedel worked on the NSC staff during the George H.W. Bush administration and the first months of the Clinton administration and during 1993 rejoined the CIA to become the NIO.

After working on the NSC staff, Reidel rejoined the CIA. According to Reidel’s recollections, he worked on the NIE at the end of a process begun under his predecessor, Ellen Laipson.[8]

Who requested this NIE, whether it was from the White House or the State Department or senior CIA officials, remains to be learned. The editor approached former CIA and State Department officials who were in government during the period, but they did not recall or would not discuss the matter or did not respond to queries.

Despite the numerous excisions in this NIE, major points come through.

Looking at the probability of war between India and Pakistan during the years ahead, the drafters of this NIE found that there were relatively low odds: 20 percent. Like INR, the intelligence analysts found that the leaders of both countries wanted to avoid war, but that points of tension – “flashpoints” – notably Kashmir, “internal meddling,” and “communal incidents” could ignite crises and confrontations possibly involving the use of nuclear weapons. Yet even if war “was not on the horizon,” like INR, the analysts believed that the chances of war could increase through “miscalculations,” including intelligence failures that could trigger a “premature attack.” The analysts envisioned circumstances that could cause an “escalating confrontation,” including “protracted surges of violence in Kashmir,” “large scale military exercise in border areas,” or a “spectacular terrorist outrage” that one side could attribute to the other. The greatest miscalculation of all could involve “the breakdown of nuclear deterrence during a crisis,” especially if military leaders on either side developed “hair-trigger responses that ensure crisis escalation.”

The estimate included discussion of the possible contribution of Confidence Building Measures, such as the agreements not to attack nuclear facilities, and to provide advance notification of maneuvers, exercises, and troop movements, but did not take much stock in them because they “might prove irrelevant in a fast-moving crisis.” Moreover, verification for any of the agreements was “spotty.” Nevertheless, the U.S. had offered to help verify CBMs, and the analysts suggested that the hotline agreement could have the “best chance of success.”

For the most part, the analysts did not see “external factors,” such as hostile powers on their borders that constituted a dangerous threat environment for either India or Pakistan. But they postulated circumstances that could raise the chances of war. One was a change in the military balance in India’s favor to the point where Islamabad could feel “even more threatened by India’s growing military edge,” which could have such destabilizing consequences as Pakistan’s “open” deployment of nuclear weapons.

A change in the military balance in combination with the rise of an “extremist” government could raise the risk of war, although “extremists” holding state power in either country was “unlikely” during the 1990s. In Pakistan, it would take the form of an “Islamist” government and in India, an election that brought the Hindu-led right-wing Bhartiya Janata Party (BJP) [Indian People’s Party] to power. A “radical” BJP regime could “blunder into conflict with Pakistan” but also “polarize” Indian society by causing “more communal violence that would damage relations with Pakistan.” The NIE includes a photograph, with an excised caption, of the mob led by activists from the BJP and related groups that destroyed the Babri Masjid mosque in Ayodhya on December 6, 1992. An Islamist regime in Pakistan, triggered by an economic crisis “and/or a military dictator joining with militants,” could add to instability by “champion[ing] the cause of India’s Muslims” but also by “promot[ing] the liberation of Kashmir by openly arming and training anti-Indian militants.”

The NIE also includes a discussion of the implications of the India-Pakistan competition for the United States, with both countries seeking U.S. aid to support their positions but also reduce the danger of “inadvertent war.” This section is heavily excised as are the discussions of nuclear proliferation and of the possibility of greater levels of U.S. efforts at “peacebuilding.” Also fully excised is “Annex A,” describing Indian and Pakistani nuclear capabilities.