In 1992, US intelligence agencies started to become concerned about China's designs for its next-generation nuclear weapons. A series of explosions monitored by the West suggested that the People's Republic of China was working on smaller, lighter thermonuclear warheads, with an increased yield-to-weight ratio. US officials did not think Chinese science was advanced enough to produce such sophisticated weapons on its own. They suspected something else-that the PRC had stolen US nuclear secrets.
Three years later the US received apparent confirmation of such thefts from the Chinese themselves. An unsolicited Chinese individual--a "walk-in," in the argot of espionage--turned a pile of PRC documents over to the CIA. Among them was a paper stamped "secret" which contained design information on perhaps the most advanced warhead in the US arsenal, the Trident II's W88.
Since then, the CIA has come to believe that the walk-in was a plant, someone who in fact worked for PRC intelligence. The US conclusion is that China, for some reason known only to its own top officials, had decided to flash a glimpse of its stolen knowledge in front of US eyes. If that is the case, it could turn out to have been a colossal misjudgment. The recent report of a special House panel, chaired by Rep. Christopher Cox (R-Calif.), on Chinese espionage has detailed years of systematic PRC spying, outraging many members of Congress. It might well have an impact on the relationship between Washington and Beijing for years to come.
That US companies, through their own laxity or greed, may have speeded the loss of secrets is faint comfort to Washington. China has obtained everything from US nuclear data to crucial help in missile upgrades and US computers and machine tools far more powerful than domestic Chinese models.
Many of the details alluded to in the House report remain classified and beyond public view. In general, however, purloined US technology and data could allow the Chinese to produce state-of-the-art nuclear weapons, upgrade their combat aircraft and submarines, conduct more extensive and effective anti-submarine warfare, equalize battlefields via information warfare, and improve their command-and-control capabilities, according to the Congressional study.
"The PRC seeks foreign military technology as part of its efforts to place the PRC at the forefront of nations," concludes the House Select Committee on US National Security and Military/Commercial Concerns with the People's Republic of China report-more simply called the Cox report. "The PRC's long-run geopolitical goals include incorporating Taiwan into the PRC and becoming the primary power in Asia."
Use of Western military technology obtained under questionable circumstances is not new for China, of course. Its current arsenal of CSS-4 nuclear-tipped ICBMs traces its design lineage to the US Titan ICBMs of the 1950s, thanks to CSS-4 lead designer Qian Xuesen, who worked on the Titan program.
A Chinese citizen educated in the US during the Japanese occupation of China, Qian became one of the world's top experts on jet propulsion during World War II. After earning a PhD at the California Institute of Technology and then working with a Cal Tech rocket research group, he was recruited to join the US military's long-range missile programs. He received a direct commission to colonel in the US Army Air Forces and began work on what became the Titan. However, spy allegations dogged Qian, and eventually he lost his security clearances. Negotiations between the US and the PRC resulted in his return to China in 1955. After serving as chief project manager in all PRC ballistic missile programs, he became head of the government arm responsible for all aeronautics and missile development research.
Theft and Diversion
Today, China uses what US intelligence calls a "mosaic" approach to the collection of technical data, which takes small bits of information collected by many individuals, then pieces them together in the PRC.
Classic spying remains a major part of this approach. Witness the case of Peter Lee, a Taiwanese-born, naturalized US citizen who worked at US national laboratories until evidence of espionage surfaced. In 1997, Lee passed China classified US developmental research on very sensitive detection techniques that could be used to threaten previously invulnerable US nuclear subs, alleges the Cox report. In 1985, Lee passed to China data about the use of lasers to create nuclear explosions on a miniature scale.
Mosaic intelligence also takes advantage of the relative openness of US society. PRC nationals attend US universities, host foreign scientific delegations, and pump visiting scientists for information that is on the edge of classified. The PRC also gets valuable bits from open forums such as arms exhibits and computer shows.
The report alleges that, at one recent international arms exhibit, PRC nationals were seen videotaping every static display and collecting all possible brochures. When a contractor left his booth unattended, Chinese spies stole a display videocassette that had been playing continual information on the US Theater High Altitude Area Defense system, a theater missile defense program.
"Converting the stolen cassette to a frame-by-frame sequence could yield valuable intelligence information to the PRC," says the select committee report.
Simple purchase of equipment plays a part in PRC intelligence gathering. Chinese front companies take advantage of US military downsizing to buy surplus high-tech US military goods, including some that are proscribed from export to all but close allies.
Two years ago, the US Customs Service seized more than $36 million in excess military property being shipped overseas illegally. Among the goods bound for the PRC and Hong Kong were 37 inertial navigation units for F-117 and FB-111 aircraft, Patriot missile parts, 500 electron tubes used in the F-14 fighter, and 26,000 encryption devices.
Military goods that find their way to the PRC can be reverse-engineered, or copied, for indigenous models. Thus the PRC's C-801 anti-ship cruise missile is thought to be a copy of the French Exocet anti-ship cruise missile. The Chinese Z-11 helicopter is a reverse-engineered French Aerospatiale AS-350 Ecureuil, according to the Cox report.
Because of the decentralized nature of the Chinese collection effort, Washington finds it very difficult to track, according to the report. It adds that, because of the FBI's historic focus on the Soviet Union during the decades of the Cold War, the US has never made monitoring the PRC's acquisition activities a priority.
"There is little or no coordination within the US government of counterintelligence that is conducted against the PRC-directed efforts to acquire sensitive US technology," concludes the Cox report.
It is in the area of nuclear weaponry that this lack of spy defenses may have hurt the US the most.
China has focused espionage activities on the relatively open environment of the US national labs for decades, according to the report of the select panel. The penetration "almost certainly continues today," claims the study.
Impetus for the PRC effort came following the end of the domestic chaos of the Cultural Revolution in 1976, when military planners sat back and assessed the state of their atomic weapons. PRC warheads of the late 1970s were large, multimegaton devices comparable to US technology of the 1950s. Officials may have decided that it was time to move to more advanced warheads and a new generation of ballistic missiles.
Over the years, the Chinese made major moves on American national laboratories located at Los Alamos and Sandia, N.M., Livermore, Calif., and Oak Ridge, Tenn. The effort evidently yielded the PRC a trove of stolen secrets. The Cox report says the Chinese obtained classified information on every currently deployed US Intercontinental Ballistic Missile and Submarine-Launched Ballistic Missile. Details remain classified, but the study says the warheads on which the PRC obtained information include the W56 warhead for the Minuteman II; the W62 for the Minuteman III; the W76 Trident C-4 SLBM, the W78 Minuteman III Mark 12A ICBM; the W87 Peacekeeper ICBM; the W88 Trident D-5 SLBM; and the W70 Lance short-range ballistic missile.
In 1996, US intelligence reported that China had stolen technology for the neutron bomb, which is intended to maximize radiation damage while reducing heat and blast. Such a weapon would be a useful tool if its possessor wished to wipe out human defenders but occupy the battlefield following conflict and avoid inflicting destruction on the area.
The PRC has also stolen data on weapons design concepts, on weaponization features, and on re-entry vehicles-the hardened shells which protect warheads during their plunge back into the atmosphere.
It may have obtained classified nuclear weapons computer codes. Theft of the so-called legacy computer codes, such as those used in development of the W88 Trident warhead, would fill in gaps in Chinese knowledge about how advanced thermonuclear devices perform when exploded. To successfully produce a W88-like weapon, the PRC may need dynamic, three-dimensional data on warhead packaging, primary and secondary coupling, and the chemical interactions of materials inside the warhead over time, according to the Cox report.
Specifics on the leaked codes remain largely classified. However, the House report confirms China acquired the MCNPT code, which is useful in determining a system's ability to survive electronic penetration; the DOT3.5 code, which performs similar calculations in a different manner; and the NJOYC code, which acts as a translator between the two other codes.
In the mid-1990s, US intelligence officials learned that China had acquired US technical information about insensitive high explosives. Conventional explosives are the first step in the chain reaction which leads to an atomic blast; insensitive high explosives are safer for use on mobile missiles. Such material can be dropped, struck, or even shot with a bullet but still not detonate.
The House Select Committee believes that the PRC theft of US secrets indicates that China will soon follow the US lead and move toward a nuclear force that is heavily reliant on lightweight, mobile, innovative nuclear weapons.
China is already known to be developing several new solid-propellant mobile ICBMs. The road-mobile DF-31, for instance, is likely to undergo first flight tests in 1999 and may be deployed as early as 2002, according to House data. The warhead for this smaller weapon would likely use elements of the US W70 or W88.
Chinese engineers may not be able to precisely match the sophistication of US warheads, but the difficulties they face in bending the US information to their own use are surmountable, according to the Cox report.
"Work-arounds exist, using processes similar to those developed or available in a modern aerospace or precision guided munitions industry," says the House study. "The PRC possesses these capabilities already."
The deployment of a new generation of thermonuclear warheads by China could prove strategically troublesome for the US.
For one thing, smaller, more efficient designs could allow the PRC to deploy missiles tipped with Multiple Independently Targetable Re-entry Vehicle warheads. The Chinese have frequently expressed opposition to US deployment of ballistic missile defenses, and MIRVs might allow the Chinese to put heavy stress on, or possibly break through, such a shield.
In addition, smaller, lighter warheads might allow China to extend the range of their SLBMs, enabling them to strike the US from distant Pacific waters.
Finally, China might not be the only nation that gets to take a peek at the secret US data. "The PRC is one of the world's leading proliferators of weapons technologies," says the Cox study. "Concerns about the impact of the PRC's thefts of US thermonuclear warhead design information, therefore, include the possible proliferation of the world's most sophisticated nuclear weapons technology to nations hostile to the United States."
High Performance Computers
US nuclear secrets are of little use to Chinese scientists unless they have access to modern computers. And the House Select Committee judges that High Performance Computer equipment recently acquired from the US represents a major leap forward in China's computing power.
In recent years, US export controls on HPCs have steadily relaxed. As a result, China now has more than 600 USorigin HPCs, estimates the Cox report. Three years ago, they had none.
Furthermore, "the Select Committee judges that the PRC has been using High Performance Computers for nuclear weapons applications," says the report.
High Performance Computers-defined as systems able to perform 1,500 to 40,000 MTOPS (Millions of Theoretical Operations Per Second)--have a wide array of legitimate civilian applications. They are useful in everything from financial market transactions and credit analysis to weather prediction and petrochemical research.
They are also essential building blocks of modern weapon design. Everything from nuclear weapons to anti-submarine warfare systems and command-and-control installations benefit from HPC power.
To keep HPCs from being used for military purposes, the Commerce Department controls their export. In general, the sale of HPCs with a performance level of greater than 2,000 MTOPS to nations other than reliable US allies requires some degree of Commerce scrutiny and/or licensing.
Under a law passed by Congress in the Fiscal 1998 defense authorization bill, Commerce is supposed to perform post-shipment verifications on all exports of HPCs with greater than 2,000 MTOPS to so-called Tier 3 nations, including China, Vietnam, and nations of the former Soviet Union.
China has long resisted any such inspections of purchased US technology, however. A June 1998 US-PRC agreement on end-use checks holds that China will consider requests for such inspections to be nonbinding. If inspections are carried out, they will be conducted by one of the PRC's own ministries.
"The Select Committee has reviewed the terms of the US-PRC agreement and found them wholly inadequate," says the Cox report.
At the time the House report was written, only one post-shipment verification had actually taken place. Yet Commerce and Defense Department data indicate that US HPCs have been obtained by Chinese organizations involved in the research and development of missiles, submarines, aircraft, communications, and microwave and laser sensors.
US companies have at times abetted such technology diversion. Compaq Computer paid a $55,000 civil penalty in 1997 to settle alleged charges that it had shipped equipment to the PRC without obtaining the proper export licenses. Digital Creations Corp. of New Jersey pleaded guilty to criminal charges that it had shipped a computer to China without the required license and was sentenced in 1997 to pay a criminal fine of $800,000.
The Select Committee believes that China is particularly interested in acquiring the kind of computer power needed for the simulation of nuclear blasts. As a signer of the Comprehensive Test Ban Treaty, the PRC can no longer legally conduct actual tests to judge the performance of weapons. Yet HPC performance in the millions of MTOPS is needed for adequate computer modeling of aging nukes.
"For this reason, the Select Committee judges that the PRC is almost certain to use US HPCs to perform nuclear weapons applications," says the report.
Inhofe Pierces Administration "Smoke Screen"
Ever since the Chinese espionage scandal erupted, harsh criticism has been falling on the Clinton Administration's team of national security advisors-and the President personally. Their foes in Congress and the media accuse them of incompetence, inattention, poor judgment, and playing low politics with the nation's defenses.
Few if any critics have been as fierce or well-informed as Sen. James Inhofe (ROkla.), member of the Senate Armed Services Committee. Inhofe's view: "This President and this Administration are singularly culpable for orchestrating a politically inspired cover-up [of Chinese spying] in order to advance policies they knew were causing harm to US national security."
Inhofe dismisses as a "smoke screen" the White House's suggestions that most of the cases occurred long ago and that all recent Presidents are equally culpable.
"Sixteen of the 17 most significant major technology breaches ... were discovered after 1994," charged Inhofe, citing data uncovered by a Congressional panel led by Rep. Christopher Cox (R-Calif.). "The notion that Presidents Carter, Reagan, and Bush knew the extent to which China's efforts to steal US nuclear and military technology were successful is fantasy."
In a recent statement posted on his Senate Internet site, Inhofe went on to say, "At least eight (and maybe more) of these breaches actually occurred after 1994. ... Among these breaches--occurring on the Clinton watch--are many of those that go the farthest in advancing China's potential as a direct nuclear threat to the United States."
According to Inhofe, the eight breaches are:
* Transfer of so-called legacy codes containing data on 50 years of US nuclear weapons development, entailing more than 1,000 nuclear tests.
* Sale and diversion to military use of some 600 High Performance Computers, enabling China to enhance its development of nuclear weapons, missiles, and advanced aviation equipment.
* Compromise of nuclear warhead simulation technology, thus enhancing China's ability to perfect miniature nuclear warheads without actual testing.
* Compromise of advanced electromagnetic weapons technology useful in the development of anti-satellite and anti-missile systems.
* Transfer of missile nosecone technology that enables China to substantially improve reliability of its ICBMs.
* Transfer of missile guidance technology enabling China to substantially improve the accuracy of its ballistic missiles.
* Compromise of supersecret space-based radar technology, which would give China the ability to detect our previously undetectable submerged submarines.
* Compromise of some other "classified thermonuclear weapons information" which "the Clinton Administration ... has determined ... cannot be made public."
Inhofe is especially incensed at the way that President Clinton's national security advisor, Sandy Berger, has cast his role in the infamous W88 nuclear warhead case. China's theft of the design of the W88 miniaturized warhead happened in the 1980s and was discovered in 1995. It was an "enormously significant" event, said Inhofe. However, Berger claims he didn't tell the President about the theft until perhaps as late as early 1998.
"The idea that Sandy Berger, ... who was fully briefed about the W88 technology breach in April 1996, did not immediately communicate this information to the President is preposterous," said Inhofe.
Inhofe went on, "The President had to have known about the W88 breach no later than April 1996, well before the 1996 election. The President deliberately withheld this vital national security information from key members of Congress for obvious political reasons. He withheld it for almost three years-a cover-up that is nothing less than a scandal of gigantic proportions."
Inhofe charges that the underlying source of Administration action was the desire to maintain close relations--especially trade relations--with China.
"Notra Trulock, the Energy Department's former director of intelligence who had first briefed Berger in April 1996, testified [that] he was prepared to brief members of the House and Senate Intelligence Committees as late as July 1998 but was denied permission to do so by acting Energy Secretary Elizabeth Moler, a political appointee. Moler reportedly ordered Trulock not to conduct the briefing because she said the information would be used to hurt Clinton's China policy."
Today's global market in space launch services is brutally competitive. China serves the same position in this market as it does for many consumer goods: It is the low-cost option. Its bids sometimes come in at half the price of Western launch firms, but it is not the high-quality option. PRC boosters have been known to veer off course and slam into nearby mountainsides, with disastrous results.
In general, the Cox report casts doubt on the wisdom of allowing US firms to put their satellites on Chinese rockets. The reason is that the boosters used are closely related to the PRC's military ballistic missiles. Launches financed by US firms and foreign agencies inevitably have given China the opportunity to refine booster reliability. In addition, US satellites are poorly guarded once they arrive in the PRC and present a tempting espionage target.
To bolster its point that the Chinese military benefits from civilian launches, the Cox report examines two cases in which US contractors may have skirted export restrictions to improve PRC boosters.
These companies--Hughes and Loral--were worried about the fate of their own satellites. Hughes Space and Communications, for instance, attempted in 1992 and 1995 to launch communications satellites on Chinese Long March rockets. Both satellites were lost when their launch vehicles exploded.
An internal Hughes investigation located the problem as being the Long March's hammerhead fairing-a sheath that protects the satellite as the rocket roars into orbit and then splits away as the payload is pushed into space. US engineers believed that the rivets that held the fairing together were not strong enough. They also thought the shape was slightly off and was vulnerable to strong winds during ascent.
The Chinese did not want to hear these points, at least not at first. They were very reluctant to admit fault in their boosters. However, commercial insurers were reluctant to back more Hughes launches in China unless changes were made. So Hughes conveyed their findings to the Chinese in a formal manner, and eventually the Long March 2E fairing was improved through such measures as an increase in the nose cap attachment screws.
Fairings are not necessary with single-warhead ICBMs. But multiple-warhead missiles use them to shroud re-entry vehicles, and the knowledge Hughes conveyed to China could help speed their development of MIRVs, believes the House Select Committee.
There is evidence that US government officials improperly approved at least some of the fairing discussions between the US firm and the PRC. Committee members allege, however, that Hughes knew that transferring the knowledge in question required additional review by the State Department.
"Hughes deliberately acted without the required State Department license," says the Cox report.
Loral was similarly worried about the reliability of the Long March rocket. On Feb. 15, 1996, a Loral Space Systems Intelsat 708 satellite was destroyed when a PRC booster tipped over even before it cleared the launch tower. The rocket crashed into a nearby hillside after 22 seconds of flight, devastating a village and killing upwards of 100 people, by some estimates.
A Chinese probe concluded that the spectacular accident was caused by a broken wire within the inner frame of the guidance system's Inertial Measurement Unit. Loral engineers thought that explanation did not come close to explaining the rocket's wild behavior. A Loral review pointed to two other possible causes: the IMU's follow-up frame or an open loop in the feedback path of the guidance system.
Loral faxed the report to the PRC in May 1996 without prior review by any US government authority, charges the Cox committee. China eventually concluded that Loral was right, and that the IMU follow-up frame had failed.
Improvements in the reliability of the Long March guidance system hurts US security because it is one of the candidates for use in the PRC's next-generation DF-31 ICBM, says the House Select Committee. Though not accurate enough to allow more than targeting of cities, the system is lightweight and compact.
One major danger in these technology transfers is simply that China has learned much about Western diagnostic processes, according to the Cox study.
"This exposure could improve the PRC's pre- and postflight failure analysis for their ballistic missile programs," says an interagency review team formed to answer questions about the Long March. "This, in turn, could increase the PRC's future ballistic missile reliability."
It is also possible that China has gleaned valuable technical information from the mere presence of USbuilt civilian satellites at Chinese launch sites. US firms are responsible for launch site security in the PRC, but buildings in which the satellites are prepared have numerous security weak points, from underground steam pipe tunnels to large unlocked window areas and paper door seals which can be peeled off, undetected, when cold.
Private US guards aren't exactly the epitome of professionalism, either. They routinely arrive for work drunk and then go to sleep, charges the Cox report. Trips to town to meet prostitutes are common.
The hunt for hookers became so intense at one point that a Defense Department monitor was approached by a PRC official who told him that one of the guards had been soliciting prostitutes in front of the local police department.
In another incident, a guard pulled a table out of line of sight of a video surveillance camera, to use it as a bed. Since the table blocked the room's door, the Defense Department monitor called the room to have it moved back. The guard reportedly responded that he was "not in the furniture moving business."
One guard even reported for work carrying a sleeping bag, charges the Cox study.
Espionage is not the only way the PRC obtains technological secrets. It also buys them, by relentlessly scouring the West for civilian items that may have military uses.
US law theoretically blocks sale of dual-use technology. However, determining what can and cannot be sold to the PRC is a difficult process, made harder by the PRC use of numerous front companies and long-term investments in Western firms.
Some efforts are rebuffed. In 1990, the PRC tried to advance its cruise missile program by buying the Williams FJ44 civil jet engine. This compact turbofan was derived from the power plant for the US Tomahawk, so the purchase was denied.
Others succeed. In 1993, a PRC company joint venturing with McDonnell Douglas to produce civilian airliners was allowed to buy 19 advanced US machine tools for its manufacturing plant. There were warning signs--the number of airliners to be built in China was cut by 50 percent, for example--but McDonnell Douglas insisted the tools were necessary for the PRC plant, so the Commerce Department approved the deal.
Two years later, McDonnell Douglas reported that six of the tools had been diverted to a factory that made military aircraft and cruise missiles, as well as commercial products.
Some attempted PRC purchases resulted in changes in US policy. In 1991, the Commerce Department decided to decontrol a popular series of civilian jet engines manufactured by AlliedSignal's Garrett Engine Division, the Garrett TFE-731. That meant that the engines could be exported without a license or US government review. The PRC quickly began negotiating with AlliedSignal over terms of a coproduction deal. Reportedly, the Chinese motivation was the need for a reliable engine for its developmental K-8 multirole military aircraft.
In July 1992, the Department of Defense learned of the negotiations. The reaction of military officials to the news sparked an interagency review of the decontrol decision. The co-production deal died after the review concluded that transfer of such jet engine production capabilities could threaten US national security.
"The PRC has mounted a widespread effort to obtain US military technologies by any means-legal or illegal," concludes the Cox report. "These pervasive efforts pose a particularly significant threat to US export control and counterintelligence efforts."