by Andrew Small
In 1982, a C-130 Hercules transport aircraft belonging to the Pakistani military left Urumqi, capital of the north-western Chinese province of Xinjiang, headed for Islamabad carrying five lead-lined, stainless steel boxes, inside each of which were 10 single-kilogram ingots of highly enriched uranium (HEU), enough for two atomic bombs.43 It is likely that this was the only time a nuclear weapon state transferred HEU to a non-nuclear country for military use. China had already sent 15 tons of uranium hexaflouride to Pakistan—somewhat more than a bomb’s worth—to ensure that the nuclear project continued on schedule: “China’s gas was most likely used in Pakistan’s first round of enrichment while the Pakistan Atomic Energy Commission was still struggling with UF6 production,” according to one Pakistani nuclear expert’s account.44 Their scientists had also been closely involved in technical cooperation, as a regular visitor to Khan Research Laboratories explains: “The Chinese were working on triggering mechanisms, the centrifuges, vacuum systems. They brought rocket propellant and super-hard metals like maraging steel…. They brought in fissile material and Khan gave them the data on enrichment and metallurgy. They helped Pakistan import and experiment with high explosives and Khan gave them his work on the centrifuge rotors.”45 Chinese officials stayed at Khan’s guesthouse at Kahuta, which was done up in the style of a Chinese hall.46
But by 1982, General Zia was nervous about the slow pace of Pakistan’s progress. The Israeli strike on Osirak, destroying Iraq’s latent nuclear programme, drew fears that India could do the same thing—or even the Israelis themselves.47 Five days before the operation, the Israeli ambassador to the UN had warned that “there is abundant evidence indicating that [Pakistan] is producing nuclear weapons”.48 Israel had made plans for a pre-emptive attack. As had India.49 Even Moscow was now a potential threat—Pakistan had already embarked on its programme of support for the mujahideen’s anti-Soviet campaign in Afghanistan, with the obvious risk of retaliation. Zia sent his military aide, Lieutenant-General Syed Ali Zamin Naqvi, to request weapons-grade fissile material and the bomb design from China, in an effort to speed Pakistan’s efforts along.50 Deng Xiaoping agreed. In each area where the Soviets had pulled the plug on Beijing, the Chinese would prove to be far more obliging to the Pakistanis.
The scope of this cooperation was ascertained relatively quickly by Western intelligence agencies. The papers that eventually turned up in Tripoli in plastic carrier bags had even been in the hands of US agents before. Until he was told to stop by Zia, A.Q. Khan had the habit of carrying weapons designs in his briefcase.51 During one trip abroad in the early 1980s, US intelligence officers gained access to his luggage in a hotel room and found drawings of a bomb and the instructions to make it, the very documents that would later be sold to Libya and possibly other customers too.52 Their Chinese provenance was as clear then as it would be three decades later. In fact, so thoroughly had US intelligence penetrated Pakistan’s nuclear programme that American weapons experts were even able to create a detailed model of the bomb, which they showed to Pakistan’s foreign minister in 1987 as a demonstration of just how much they knew.53 The drawings themselves were shown to Zia by Vernon Walters, former deputy director of the CIA, as early as 1982.54 In spite of this, the political pressure on Pakistan and China from the United States could at best be described as modest. The three sides had been working as a virtual alliance against the Soviet Union in Afghanistan since the 1979 invasion, and proliferation issues were of lower salience than the opportunity to deal the Soviets a fatal blow. The US National Security Advisor, Zbigniew Brzezinski, articulated the stance most pithily in his argument to President Carter, “Our security policy cannot be dictated by our nonproliferation policy.”55 The Chinese were informed of this stance in the course of bilateral meetings in 1980:
Secretary Brown: “Our big problem with Pakistan was their attempts to get a nuclear program. Although we still object to their doing so, we will now set that aside for the time being and concentrate on strengthening Pakistan against potential Soviet action.”
Deng Xiaoping: “That is a very good approach…We applaud this decision.”56
The Reagan administration thought much the same thing. Evidence of Pakistan’s covert nuclear programme was certainly an irritant, not least since it threatened to torpedo Congressional support for the upgrading of the US-Pakistan security relationship, but it was not the first-order concern. In 1981, an agreement was reached to sell Pakistan the F-16 fighter jets that it would later adapt to become part of its nuclear strike force. At the time, the only other recipients of the state-of-the-art aircraft were NATO allies and Japan.57 A Congressional amendment to the arms sales package specified that aid would be cut off if Pakistan tested a nuclear weapon. But the administration understood clearly that the nuclear programme would continue to move ahead, even if it stopped short of an actual detonation. The Chinese had not taken US proliferation concerns especially seriously even before the Soviet invasion. Kissinger had joked with Chinese leaders that the best way to contain India’s ambitions was to arm Pakistan and Bangladesh with nuclear weapons.58 The Chinese expected that any US objections could be weathered—and they were right.
When it came to Pakistani transfers of Western technology to China’s nuclear programme, the United States was ambivalent—and some in the US government were even tacitly supportive. In 1973, Kissinger had assigned a small group to assess Soviet threats to China and how the United States could help to address them.59 The conclusion: China’s nuclear arsenal was vulnerable, the PLA’s technology, logistics and industrial capacity were poor, the air force was mostly obsolete, and the navy was in an even worse state.
If the United States was willing to support a “hardening” of China’s defensive capacities, there would be several advantages. It could help tie down Soviet forces on its eastern frontier, reduce the temptation for Moscow to coerce China or launch surprise strikes, reinforce China’s anti-Soviet resolve, and minimize the prospects of a nuclear crisis between the two powers.60 Any transfer of US military technology or arms sales to China would be a matter of high sensitivity, of course, and some of it would have to be undertaken through friends and allies who faced fewer restrictions. While the bulk of the heavy-lifting would end up being undertaken by the Europeans and the Israelis,61 Pakistan—so recently the bridge between Washington and Beijing—was another obvious place to turn. The man spearheading early efforts to launch the programme of US-China military collaboration, Michael Pillsbury, told Pakistani officials that “logically, it would need Pakistan’s cooperation”.62
For a number of years, there was hesitation on the US and Chinese sides about proceeding. A further study in 1975 by James Lilley, the National Intelligence Officer for East Asia, which suggested that US military ties with Beijing could strengthen those Chinese leaders who favoured closer links with the West, was taken up by Kissinger on his next visit to China.63 Mao was cool to the American offer, stating, “As for military aspects we should not discuss that now. Such matters should wait until the war breaks out before we consider them,” to which Kissinger responded, “Yes, but you should know that we would be prepared then to consider them.”64 They were indeed. Six weeks later, during President Ford’s trip to Beijing, authorization was given by Washington for a sale by the British company, Rolls Royce, of 50 Spey jet engines that would be used to power PLA Air Force fighters, the first military-related technology sold to China by the West.65 Ford also agreed to the sale of two high-powered US computers that could be used by China for nuclear warhead and ballistic missile development.66 But it was only after the Soviet invasion of Afghanistan that the Sino-US security relationship began in earnest. China would end up receiving everything from arms sales and technology transfers to the US field manual for the “Air-Land Battle” doctrine that underpinned the US defence of Europe against Soviet invasion. But at the heart of US concerns was China’s nuclear arsenal. Its vulnerability to Soviet attack derived from its small size, its lack of sophistication, its w
eak command and control infrastructure, and the lengthy and complex preparations required before the weapons could be launched.67 The top US priority was to improve Chinese early-warning capacities,68 reducing the state of readiness in which China needed to keep its liquid-fuelled missiles and thereby the incentives for either China or the Soviet Union to launch first strikes.69
The United States chose to address the problem directly. In 1979, the secretive Sino-US “Chestnut” programme was put into motion.70 Between August and December, the CIA airlifted equipment to China for a pair of monitoring stations that were established in the Tian Shan mountains, at Korla and Qitai,71 close to Urumqi and the Sino-Soviet border, with operations beginning in late 1980.72 Chinese technicians from the PLA 2nd Department were trained at a SIGINT training centre near San Francisco.73 As well as monitoring military communications and radar signals from Soviet air defences, their antennae could detect any change in the alert status of Soviet nuclear forces. The listening posts meant that China was able to increase its warning time for nuclear attacks and Washington was able to replace the capacities it was losing with the fall of the Shah. The “Tacksman” listening stations in the mountains of northern Iran performed a similar function for the United States but had to be closed or destroyed following the 1979 revolution.74 The final details of the agreement were sealed during a secret trip by the head of the CIA, Stansfield Turner, who visited Beijing, in a disguise replete with moustache, in December 1980, his last as Director of Central Intelligence.75 “It was clear that the Chinese leadership, Deng especially, regarded this cooperation as a major strategic decision for them,” noted his aide, future Defense Secretary Robert Gates. “It was for us, too, as we sat down with people with whom we in intelligence had been at war since 1949.”76
But there was a more ambitious goal too. As one US army journal in 1979 argued: “The flow of Western technology made possible by the shift in U.S.-Chinese relations may strengthen [Chinese] military capabilities to the point where the Soviet Union is increasingly forced to pursue a conservative, defensive, and détente oriented strategy”—especially China’s strategic forces, which, although “relatively primitive”, could be “expected to improve strikingly as a result of China’s new emphasis on orderly technological development, and the flow of commercial and military technology.”77
Pakistan was certainly one of the early sources of this “flow”. The debate over whether Pakistan should receive an advanced radar system as part of the F-16 sales illustrated the balance of considerations. The CIA warned in 1982 that “the sale of the AN/ALR-69 Radar Warning Receiver to Pakistan entails a significant risk of the equipment being exploited by China…China has obtained French weapons—and possibly U.S. air-to-air missiles—from Pakistan and has negotiated agreements on joint weapons developments based on Western arms technologies acquired by Pakistan”.78 The radar would give China “the potential of a significant improvement in radar warning capability” and China “would benefit from access to Western avionics fabrication technology gaining several years in the development of a modern radar warning system”.79 The transfer to Pakistan went ahead regardless.
However, it is far from clear that the transfers of Western technology did a great deal to help China’s nuclear programme. A.Q. Khan’s P1 centrifuges—the stolen Dutch centrifuge technology—did not operate well and there is no reason to believe that China had more luck with them than Pakistan did. Chinese weapons-grade uranium throughout the period came from its gaseous diffusion plants—where it had achieved a genuine breakthrough in the enrichment performance in the early 1980s—and not from gas centrifuges.80 China’s centrifuge programme was still in development by the 1990s, and when it finally did establish large-scale centrifuge plants at Hanzhong, the very location where A.Q. Khan claimed to have helped to assemble a facility, China simply purchased them wholesale from a familiar source: Russia.81 China’s nuclear programme did benefit from another set of Western technologies, but these came directly from the source. A 1984 Defense Intelligence Agency estimate suggested:
There is evidence that the Chinese have been successful in assimilating into their nuclear weapons program United States technology in areas such as high explosive, radiochemistry, metallurgy, welding, super computers, numerical modeling, high speed photonics, and underground drilling…Increased access to this technology and continued Chinese efforts will in the 1980s and early 1990s show up as qualitative warhead improvements.82
The thanks for this, however, were owed to “overt contact with U.S. scientists and technology and covert acquisition of U.S. technology,”83 not to the Pakistanis. By this time US-China military exchanges and arms sales had become increasingly normalized, US defence and high-technology hardware sales reaching $5 billion in 1985.84 Whatever deal Bhutto struck with Mao, it is clear who got the better end of it.
This was even more obviously true of Pakistan’s missile programme. While there is a view that “if you subtract China’s help, there wouldn’t be a Pakistani [nuclear] program”,85 there is also a good case that Chinese aid was largely a “supplemental contribution”: Pakistan’s acquisition of the bomb certainly relied on its own scientific and technical prowess too.86 As one Pakistani nuclear scientist puts it: “It is quite likely that the development of nuclear weapons by Pakistan would have succeeded but without Chinese assistance this would have taken longer.”87 It is harder to make the same claims about delivery mechanisms for those weapons. Pakistan’s missile programmes had a far weaker technical base than its nuclear programme. Until the United States choked off its F-16 sales in 1990, it was still these planes rather than ballistic missiles that were seen as the primary delivery vehicle for its nuclear arsenal. But India’s tests of its first short-range ballistic missile in 1988 and an intermediate-range ballistic missile in 1989 prompted an attempt by Pakistan to demonstrate that it had its own matching capabilities. In February 1989, Pakistan proudly announced that it had tested its own Hatf missiles, named after the Prophet Muhammad’s sword. Foreign observers were unimpressed by the hurriedly developed series, which suffered from limited range and accuracy. The Hatf-1 was dismissed by US experts as an “inaccurate battlefield rocket that can travel 80km”; the Hatf-2 as “two Hatf-1s put together”.88 But even before these missile tests, Pakistan had again turned to its old friend for help. The Chinese surpassed themselves. They had already assisted in Pakistan’s crash effort to demonstrate an indigenous missile capability. But now, not only would they transfer to Pakistan some of their very latest models—the M-11 and the M-9 missiles had only recently been inducted by the PLA itself—but they ensured that Pakistan could develop its own rockets in the future.89 The M-11s, developed by the Sanjiang Space Group in Hubei Province, gave Pakistan the 300km range missile that Islamabad had pretended the Hatf-2 provided.90 The M-9s, the domestic version of which Beijing would use in an “exercise” during the 1995–96 Taiwan Strait crisis to intimidate Taipei, gave Pakistan the capacity to strike New Delhi.91
The outright handover of the M-11 missiles did not go unnoticed. In late 1992, US intelligence spotted a shipment of the missile parts passing through the port of Karachi.92 They were destined for the air force base at Sargodha, in western Punjab, which soon became the focus of international attention, just as the nuclear facilities at Kahuta had been before it.93 Thirty of the missiles were stored in crates there, and satellite photos revealed shelters for the crates, mobile launchers, and missile maintenance areas.94
After the unhelpful publicity the missiles attracted, China began supplying both M-11s and M-9s in unassembled form, which required the development of a dedicated missile assembly facility near Rawalpindi.95 Chinese experts showed up in Sargodha and other locations to train Pakistani technicians to become self-reliant for future production.96 Virtually every time a new missile was added to Pakistan’s arsenal, it could be traced to a Chinese prototype. The Shaheen-I, rolled out in 1999, bears a striking resemblance to the M-9. The 2000km-range Shaheen-II, displayed at the Pakis
tan Day parade in 2000, is believed to be based on China’s M-18 missile or an adapted M-9.97 Pakistan’s first cruise missile, the Babur, put into production in 2005, appeared to be based on China’s DH-10s (though both are ultimately reverse-engineered US cruise missiles). Its current missile defence system uses Chinese HQ-9 surface-to-air missiles. Even Pakistan’s rocket launcher for battlefield nuclear weapons was based on a Chinese design.98 The only notable exceptions were the long-range Ghauri missiles—the first missiles that gave Pakistan the capacity to strike any city in India—and they came from North Korea, in one of the most controversial A.Q. Khan deals. Pyongyang provided Pakistan with Nodong missiles, in return for which it not only received cash but may also have got its hands on crucial documents and components to support its clandestine uranium enrichment programme.99
Unlike in the early 1980s, during the crucial phase of Sino-Pakistani nuclear cooperation, which proceeded with little serious challenge, Chinese missile transfers generally took place in the teeth of international opposition. The transfers began when China was at the low ebb of its post-Tiananmen isolation, and when the collapse of the Soviet Union meant that the United States no longer had the same need to maintain its Cold War quasi-alliance with either China or Pakistan. The sale of the M-11 launchers resulted in US sanctions in 1991—a blacklisting of the companies involved—which were eventually followed by a two-year freeze of high-technology sales to China.100 On each occasion, China would make a new promise or sign a new agreement with the United States in order to have the sanctions suspended, only to continue its transfers exactly as they had agreed with the Pakistanis. At every point where they were challenged, the Chinese would counter with complaints about the US sales of F-16 warplanes to Taiwan, which had started shortly before the M-11 missile transfers.101 Beijing also remained on hand to support the Pakistani nuclear programme itself, exporting five thousand ring magnets in 1994, which, it was reported, enabled Pakistan to double its production of highly-enriched uranium.102 There are even suspicions that China tested a warhead on Pakistan’s behalf in 1990 at its facility in Lop Nur, Xinjiang, well in advance of the 1998 explosion in the Chagai hills in Balochistan that formally announced Pakistan’s membership of the nuclear club.103 Neither did that Pakistani test bring a halt to Chinese proliferation: China increased its shipments of specialty steel and guidance systems following the Indian and Pakistani tests.104 Over the course of the next decade, China would continue to provide technology and support for the expanded production of Pakistan’s ballistic missiles.