China and Weapons of Mass Destruction: Implications for the United States

Conference Report
5 November 1999

This conference was sponsored by the National Intelligence Council (NIC) with Armed Forces Journal International and the National Security Studies Program at the Edmund A. Walsh School of Foreign Service, Georgetown University.  The views expressed in this conference summary are those of individuals and do not represent official US intelligence or policy positions. The NIC routinely sponsors such unclassified conferences with outside experts to gain knowledge and insight to sharpen the level of debate on critical issues.

Introduction | Schedule | Papers Appendix I | Appendix II | Appendix III | Appendix IV

Introduction

This conference document includes papers produced by distinguished experts on China's weapons-of-mass-destruction (WMD) programs. The seven papers were complemented by commentaries and general discussions among the 40 specialists at the proceedings.

The main topics of discussion included:

The development of China's nuclear forces.

China's development of chemical and biological weapons.

China's involvement in the proliferation of WMD.

China's development of missile delivery systems.

The implications of these developments for the United States.

Interest in China's WMD stems in part from its international agreements and obligations. China is a party to the International Atomic Energy Agency (IAEA), the Treaty on the Non-Proliferation of Nuclear Weapons (NPT), the Zangger Committee, and the Chemical Weapons Convention (CWC) and has signed but not ratified the Comprehensive Nuclear Test Ban Treaty (CTBT). China is not a member of the Australia Group, the Wassenaar Arrangement, the Nuclear Suppliers Group, or the Missile Technology Control Regime (MTCR), although it has agreed to abide by the latter (which is not an international agreement and lacks legal authority).

The papers below reflect important trends in thinking outside the Intelligence Community on the issue of China and WMD. As noted on the title page, the views stated in the papers are those of the authors and are not necessarily those of the Intelligence Community or any particular US Government agency.

---

Schedule

Welcome	(9:00-9:05 AM): Robert L. Worden, Chief, Federal Research Division
Opening Comments	(9:05-9:15 AM): Robert G. Sutter, Moderator, National Intelligence Officer for East Asia
Panel One	(9:15-10:45 AM): WMD Capabilities
	Bates Gill and James Mulvenon - The Chinese Strategic Rocket Forces: Transition to Credible Deterrence
	Eric Croddy - Chinese Chemical Warfare Capabilities
	Commentators: Torrey Froscher and Catherine E. Johnston
Panel Two	(11:00-12:30 AM): Scope of WMD Proliferation
	Evan Medeiros - The Changing Character of China's WMD Proliferation Activities
	Shirley Kan - Chinese Proliferation of Missiles and WMD: Issues for US Policy
	Commentators: Harlan Jencks, Peter Brookes, Janice Hinton
Panel Three	(2:00-3:45 PM): China's Views on WMD
	Michael Swaine - The Chinese View of Weapons of Mass Destruction
	Mark Stokes - Weapons of Mass Destruction: PLA Space and Theater Missile Development
	Ken Allen - Key Indicators of Changes in Chinese Development and Proliferation of Weapons of Mass Destruction
	Commentators: Lonnie Henley and Vincent Bonner
Panel Four	(4:00-5:15 PM): Wrap-Up: Implications for US Interests and Policies
	Peter Almquist, Michael McDevitt, and Thomas Fingar
	Contributors
	Ken Allen is with the Stimson Center.
	Peter Almquist is with the Department of State.
	Peter Brookes is a member of the staff of the International Relations Committee, House of Representatives.
	Eric Croddy is a senior research associate at the Chemical and Biological Weapons Nonproliferation Project, Center for Nonproliferation Studies (CNS), Monterey Institute.
	Bates Gill is Senior Fellow in Foreign Policy Studies at the Brookings Institution, and Director of the Brookings Center for Northeast Asian Policy Studies.
	Thomas Fingar is with the Department of State.
	Torrey Froscher is with the Central Intelligence Agency.
	Janice Hinton is a specialist on Chinese affairs.
	Lonnie Henley is with the Defense Intelligence Agency.
	Harlan Jencks is with the Lawrence Livermore National Laboratory.
	Catherine E. Johnston is with the Defense Intelligence Agency.
	Shirley Kan is with the Library of Congress.
	Michael McDevitt is with the Center for Naval Analysis.
	Evan Medeiros is a senior research associate on the East Asia Nonproliferation Project at the Center for Nonproliferation Studies in Monterey, CA.
	James Mulvenon is Associate Political Scientist at the RAND Corporation, and Deputy Director of the RAND Center for Asia-Pacific Policy.
	Mark Stokes is with the Office of the Assistant Secretary of Defense for International Security Affairs.
	Robert G. Sutter is National Intelligence Officer for East Asia, National Intelligence Council.
	Michael Swaine is with the RAND Corporation.
	Robert L. Worden is Chief, Federal Research Division, Library of Congress.

---

Bates Gill and James Mulvenon¹

Introduction

The doctrine and force structure of China's Strategic Rocket Forces (also known as the Second Artillery from the Chinese di er pao) remain some of the most heavily shrouded and poorly understood aspects of the Chinese military. Yet, as China undergoes a continued modernization of its nuclear forces, to include improved mobility, reliability, accuracy, and firepower, concerned analysts are compelled to understand and analyze the Second Artillery more precisely, including its evolving doctrine, organization, and hardware, and their implications for international security.²

To date, the most prominent work on China's nuclear posture has either dwelled primarily on hardware and R&D,³ focused on doctrinal debates,⁴ or described the technological development of Chinese nuclear weapons in the form of political-military histories.⁵ Some past work, now more than 10 years old, attempts to weave several of these strands together in the context of a "cultural" explanation.⁶ More recent work by Johnston and Xue goes furthest in providing more unifying analyses that carefully draw together aspects of doctrine and force structure, yet this work requires some reexamination.⁷

In light of China's continuing nuclear weapons modernization program, an updated and more comprehensive framework is needed that fully pulls together theoretical analysis, China's declared nuclear principles, and an empirical assessment of its nuclear force structure. Taking such an approach, we reach four key findings on Chinese nuclear posture:

• First, from a theoretical perspective, traditional approaches such as neo-Realist and organization theory do not adequately predict and explain key aspects of Chinese nuclear doctrine and force structure. Rather, an understanding of such variables as domestic political, technological, historical, and cultural factors provide far greater insight and predictive capacity about the drivers that shape China's doctrinal and force structure decisions.

• Second, from a technical perspective, although we agree with analysts who highlight the role of technology in shaping Chinese doctrine, we go beyond the somewhat simplistic understanding that technology drives doctrine. Rather, we see patterns of rational strategic choice made for China's nuclear posture, though technology limited the realm of the possible for Chinese leaders. Perhaps it could be said that the Chinese made a virtue out of necessity in the construction of their nuclear deterrent, accepting the technological constraints of the system and making rational choices under those constraints.

• Third, we find that the evolution over time of China's doctrine and force structure is the story of trying to close the gap between real capability, on the one hand, and what one might call "aspirational doctrine" on the other. In the United States, the appropriate analog would be a comparison of current operational doctrine, as outlined in the Joint Doctrine publications series, with an aspirational doctrine, such as Joint Vision 2010. In the Chinese case, the discontinuity between reality and aspiration is of times referred to as the "capabilities-doctrine gap." At the present stage in the Second Artillery's modernization, China is nearing an historic convergence between doctrine and capability, allowing it to increasingly achieve a degree of credible minimal deterrence vis-à-vis the continental United States--a convergence of its doctrine and capability it has not confidently possessed since the weaponization of China's nuclear program in the mid-1960s.

• Finally, for the future, the doctrine and force structure of China's Second Artillery should be analyzed at three distinct levels, reflecting a multifaceted force with very different missions: a posture of credible minimal deterrence with regard to the continental United States and Russia; a more offensive-oriented posture of "limited deterrence" with regard to China's theater nuclear forces; and an offensively configured, preemptive, counterforce warfighting posture of "active defense" or "offensive defense" for the Second Artillery's conventional missile forces.

Theoretical Examination of China's Nuclear Posture

In reaching these findings, the work proceeds in five sections. First, we begin with a theoretical analysis of Chinese nuclear posture. Second, in the absence of an open and official declaration of Chinese nuclear doctrine, we examine China's declared nuclear principles to inferentially deduce certain aspects of China's nuclear doctrine. In a third and fourth section, we test these findings by closely examining empirical data on China's current and likely future nuclear force structure. A final section draws these findings together to reach conclusions about China's past, present, and likely future nuclear force posture.

One observer of China's nuclear program states that "for about 30 years after China exploded its first nuclear weapon there was no coherent, publicly articulated nuclear doctrine."⁸ In a similar vein, others have noted that China's nuclear weapons program "proceeded without such strategic guidance" and that "until the early 1980s, there were no scenarios, no detailed linkage of the weapons to foreign policy objectives, and no serious strategic research."⁹ In the absence of definitive official, authoritative open-source documentation to describe China's nuclear doctrine, how can analysts begin to understand Chinese nuclear posture? To start, one can briefly consider several theories, or "analytical lenses," to deduce likely Chinese doctrinal choices. The literature offers three principal "models," or explanatory frameworks.

The first framework to consider is neo-Realism. Neo-Realism stresses the state as the primary actor on the international scene, and focuses on the propensity of states to engage in "self-help" in order to preserve their interests in a hostile, anarchic world system. According to neo-Realist predictions about nuclear posture, China, as "revisionist power," would likely prefer offensive weapons and doctrines. Furthermore, neo-Realism would predict that as a country that faced a number of powerful adversaries in the formative years of its nuclear weapons program (first the United States and then the United States and the Soviet Union), China would wish to pursue offensive weapons and doctrines. Neo-Realism would also predict that, as a revisionist power with limited means to detect imminent attack, Chinese doctrine would favor offensive, preventive war strategies.¹⁰

Another theoretical approach, known as organization theory, looks to the presumed preferences of military organizations as a determinant of doctrinal outcomes.¹¹ An organization theory framework would suggest that under the highly militarized domestic conditions during the initial development of China's nuclear arsenal (from the mid-1950s to the early-1970s) China would have likely pursued an offensive nuclear posture. According to this framework, the strong presence of Chinese military interests in doctrinal and weapons development in the first decades of the People's Republic would likely result in the rejection of no-first-use posture, and would favor first-use options and counterforce targeting. According to the organization theory framework, this would be predicted by the fact that China's leadership during this period was made up of active and former military leaders, and the fact that the nuclear weapons program itself was conducted largely under the auspices of the military. In addition, because China went through a series of external security crises during the formative years of its nuclear arsenal, organization theory would warn of an even stronger likelihood that the military would actively pursue offensive deployments and doctrines.

A third predictive approach gives greater weight to domestic political, historical, and cultural factors as determinants for shaping doctrinal decisions. This approach, known as neo-culturalism in the academic literature, can be applied to the Chinese case by examining domestic political interests, civil-military relations, resource restraints, and historical experience. In the Chinese case, one can point more specifically to domestic political factors (especially the unusual dynamic of Party-Army relations), technical factors (particularly availability of resources), and other historical and cultural factors as critical variables compelling doctrinal decisions.¹² In examining these factors, neo-cultural explanations--unlike neo-Realist or organizational frameworks--would not necessarily predict a Chinese preference for offensive nuclear doctrines.

Certain aspects of the empirical record would lend support to the predictions of either the neo-Realist or the organizational theorist, or both. For example, the initial Chinese decision to go nuclear in January in 1955 is predicted by the neo-Realist approach that places great emphasis on threats and prestige as useful indicators. In another example, we see that midlevel Chinese military officers have been the most open in recent years to promote more offensively oriented deployments and doctrines, as shown in Iain Johnston's work.¹³

However, in taking the 45-year record of Chinese nuclear weapons development as a whole, neo-Realist and organizational frameworks would not predict the basic declared principles and empirical record of Chinese nuclear weapons posture overall. As explained in fuller detail in subsequent sections, China's nuclear posture overall has adopted such principles as no-first-use, has circumscribed use in the form of both positive and negative security assurances and the declared adherence to nuclear-weapon-free zones, provides no extended deterrence guarantees beyond its borders, and maintains qualitatively and quantitatively limited forces, resulting in likely "countervalue" (as opposed to "counterforce") targeting, and a delayed second-strike (as opposed to launch on warning or launch on attack) state of readiness.

Hence, in the Chinese case, considering the neo-cultural approach to help predict and understand Chinese doctrinal choices would be more helpful to us. What specific aspects of domestic politics, historical experience, and cultural tradition stand out in this regard?

From the perspective of domestic politics, we must recognize first and foremost that in the critical decades that Chinese nuclear weapons were first developed, Chinese nuclear weapons decisions were firmly dominated by the views and statements of Mao Zedong and a small number of other leaders under the powerful political sway of Maoist political ideology and rhetoric. Mao's own publicly expressed opinions about nuclear weapons served as the guiding principles for the development of the Chinese arsenal. Lewis and Xue have derived seven major principles from official Maoist statements in the 1960s and 1970s that helped define the future parameters of Chinese nuclear deployments and doctrine: (1) no first use; (2) no tactical nuclear weapons; (3) "small but better"; (4) "small but inclusive,"; (5) minimum retaliation; (6) quick recovery; (7) soft-target kill capability.¹⁴ A recent study by a Chinese missile scientist argues that many of these principles continue to carry great weight in determining the fundamental quantitative and qualitative parameters of China's nuclear weapons arsenal even today.¹⁵

A good part of this thinking with regard to nuclear weapons was derived from the wartime experience of the Chinese communist leadership, especially during the Chinese civil war (1927-49), and in the war or the communists against the Japanese (1937-45). According to Mao, Chinese communist military successes of "People's War" emphasized guerrilla tactics within a protracted war strategy, the importance of manpower over technology, the moral and physical attrition of the enemy over time, and the importance of controlling the strategic "hinterland" to surround the enemy's base in the developed urban centers. For nuclear doctrine, this translated into (1) opposition to quick or preemptive military actions from a position of weakness; (2) an appreciation for "strategic retreat" and the primacy of defense in the interest of eventual victory; (3) a subordination of a strictly military viewpoint to the political-military goals of the revolution; and (4) the ultimate superiority of man over weapons and technology.¹⁶

Mao's opinions also were influenced by his careful reading of Chinese history and its classic texts, especially the work of Sun Zi (Sun Tzu), who wrote the classic Art of War in the 6th century BC.¹⁷ Contemporary Chinese interpretations of this work emphasize the largely defensive and nonviolent nature of Chinese strategic thought, most often citing Sun Zi's well-known maxim: "To win one hundred victories in one hundred battles is not the acme of skill. To subdue the enemy without fighting is the acme of skill." Other aspects of Sun Zi's thought that favor "nonviolent" means to vanquish one's opponents--deception, wily strategy, and what is known today as "psychological warfare"--also are often cited as representative of traditional Chinese strategic thinking.¹⁸ Moreover, this interpretation of strategic thinking finds resonance in the larger context of Confucianism--the single-most-dominant philosophy of statecraft in Chinese history--and its overarching concern with abjuring violence and assuring order through moral--rather than strictly military--strength.

Interestingly, the term in China for "deterrence" itself may help explain Chinese nuclear posture. For example, a "Confucian" approach to nuclear doctrine may be reflected in China's frequently stated "opposition" to the policy of nuclear deterrence. This apparent contradiction only leads to suspicions about true Chinese intentions, especially from Western analysts who view deterrence as an essentially defensive and stabilizing condition. However, discussions with Chinese strategists suggest that this confusion may derive in part from Chinese perceptions of the word "deter," which in Chinese (weishe) connotes strongly the notion of "menacing" or "terrorizing with military force," and implies threatening rather than defensive intent. Alternative terms in Chinese for "deterrence" also imply threats: hezu,to frighten into inaction, and weixie, to awe and threaten. Not wishing to portray its nuclear weapons as threatening, China traditionally stated its opposition to deterrence.

Since late 1995, China's official position has adjusted slightly its stance to criticize the "obviously anachronistic . . . policy of nuclear deterrence based on the first use of nuclear weapons." Track-two discussions between US and Chinese officials were able to glean a further Chinese distinction to the effect that China exercises a "defensive deterrent," while the United States wields an "offensive deterrent."¹⁹

A second domestic political factor in the Chinese case that neo-Realist and organizational theory cannot fully capture is the unique dynamic of China's "Party-Army" relations. Both the neo-Realists and the organizational theorists assume a discernible distinction of preferences between "civil" and "military" leaders in a given state. The revolutionary history of the Chinese political-military leadership often belies that assumption, especially in the formative years of the People's Republic and the development of the Chinese nuclear arsenal. Chinese "civilian" or "Party" leaders--such as Mao Zedong, Liu Shaoqi, Deng Xiaoping and Zhou Enlai--had considerable experience as revolutionary military leaders, while members of the uniformed military carried significant political power as Party leaders and, by dint of their status, as revolutionary heroes. Powerful "military" interests and predispositions intertwined with "civilian" (or "Party") concerns to reach decisions of a broader "political-military" nature, which is reflected in the apparent doctrine of China's nuclear arsenal.

The notion of different "Party-Army" factions is a better approach to understanding how the Party and the Army interact for decisions in China. The differences between these factions are resolved at the highest levels of Chinese politics where both ostensibly "civil" and "military" leaders represent interests as individuals of the Chinese Party-Army state, rather than the corporate interests of bodies of which they are members. Three good examples of how this factionalism and resolution played out were the intervention of the military to quell the excesses of the Cultural Revolution, the overthrow of the Maoist "Gang of Four" in 1976, and the deployment of troops to crush the Tiananmen Square demonstrations of 1989. In these cases, different "Party-Army" factions formed across institutional boundaries to advocate different, often diametrically opposed, courses of action.

We should note how key decisions under the conditions of a symbiotic "Party-Army" relationship have traditionally been taken by China's topmost leaders, who by necessity must credibly bridge the gap between civil and military constructs. The result for strategy in the formative years of the Chinese nuclear arsenal was a more comprehensive and political-military doctrine, not a strictly "military" or "civilian" approach.²⁰

Third, an understanding as to how the Chinese define "doctrine" also helps explain what appear to be discrepancies between doctrine and capability. Briefly put, what Western observers might call "doctrine" is different from the Chinese definition. What the West often defines as doctrine in the Chinese context is better understood to be "basic doctrine, as distinct from operational doctrine." Doctrine for China is "less operational and practical, and is more of a systemic description of the theory or overall construct guiding the PLA's defense posture."²¹ In practice, we would differentiate between "aspirational doctrine" as opposed to "actual doctrine." In the United States, the appropriate analog would be a comparison of current operational doctrine, as outlined in the Joint Doctrine publications series, with an aspirational doctrine, such as Joint Vision 2010. Thus, just as "minimal deterrence" at the beginning of China's nuclear weapons program reflected hopeful thinking as much as on-the-ground reality, so too today discussions of a warfighting or "limited deterrent" are likely indicative of future goals rather than current capabilities. To state, for example, that "the PRC's announced strategic doctrine is based on the concept of 'limited deterrence'"²² not only misinterprets Johnston's research and wrongly implies that the Chinese have ever "announced" a formal doctrine, but also wrongly attributes a Western sense of "doctrine" to what amounts to a Chinese "aspirational" doctrine.

Finally--and again a point not well explained by either neo-Realist or organizational theory frameworks--the empirical record suggests that Chinese nuclear weapons options and doctrine were shaped by resource constraints, especially considerations of technological development.²³ As noted above, we find that Chinese doctrinal preferences were not the principal drivers behind technological deployments (as neo-Realists and organizational theorists would likely predict), but rather the other way around: doctrine was shaped by what was technologically desirable or feasible. As a developing world state, technical obstacles and resource deficiencies almost immediately limited Chinese deployments to a defensive, countervalue, minimal deterrence stance, the principal features of China's traditional nuclear weapons doctrine. For example, China's reliance on countervalue targeting derives from the questionable accuracy of its ballistic missile forces and large-yield warheads that made precise, limited counterforce attacks unfeasible.²⁴

Chinese technological restraints were further exacerbated by certain domestic political and arguably "cultural" or historical factors. In turn, these developments limited Chinese doctrinal options resulting in a reliance on largely defensive and minimalist approaches. First, China's historical perception of itself as a "victim" at the hands of aggressive, more powerful states limited political choices--especially in the early years of China's nuclear weapons development--which may have favored more offensive and threatening nuclear postures. Second, the period of China's early development and eventual deployment of its rudimentary nuclear arsenal coincided closely with a turbulent period of domestic political upheaval. As Lewis and Xue have written in reference to China's pursuit of a nuclear submarine armed with solid-fuel missiles, it is "a story of politics and technology in collision."²⁵

While China eventually--after a 30-year effort--deployed a nuclear-powered submarine armed with nuclear weapons, it did so only tortuously and at great technological cost; the single submarine currently serving as the third leg of China's strategic triad rarely leaves port and has constant operational difficulties.

Third, China's historical ambivalence and self-reliant stance toward political and technological dependency also had implications for its nuclear weapons development. This position, already well entrenched in Chinese thinking dating back to the Opium Wars of the mid-1800s, was considerably strengthened during China's "century of shame" and following China's "betrayal" at the hands of Krushchev in the late 1950s and early 1960s. These lessons of historical experience slowed the acceptance and integration of foreign assistance and technologies in the development of the Chinese nuclear force. This situation constrained doctrinal choice and contributed to the development of the Chinese minimal deterrent.²⁶

Taken together, the available evidence suggests that, in analyzing the underlying causes of Chinese strategic choices, we need to give far greater attention to an approach that carefully considers domestic political forces, resource restraints, and historical experience.

China's Nuclear Weapons Principles

Moving beyond an explanation of the causal factors behind Chinese nuclear posture, what specific nuclear principles have resulted, and what can we deductively infer from them as a way to describe Chinese doctrine? On the whole, these declared nuclear principles tell us more about when China claims it would not use nuclear weapons than when it would. Nevertheless, we can infer from these principles certain aspects of an otherwise undeclared nuclear doctrine. Overall, these declared principles support what the Chinese claim to be the generally defensive nature of its nuclear arsenal. As we will see, there is room to question this assertion, though we find that the principles generally conform to current force structures (see next section). We can consider these declared principles in three parts: China's no-first-use principle, its negative and positive security assurances, and its declared adherence to nuclear weapon free zone agreements.²⁷

No First Use
First, public Chinese statements consistently reiterate the "defensive" purpose of Chinese nuclear weapons to counterbalance foreign threats. China's long-held "no-first-use" (NFU) policy serves as the foundation of this aspect of China's declared defensive nuclear posture. Chinese leaders decided to pursue nuclear weapons in January 1955 due to US nuclear threats during the Korean war and Taiwan Straits crisis of the early 1950s.²⁸ In a statement issued on the day of its first nuclear explosion in October 1964, China cited this achievement in its "struggle to strengthen [its] national defense and oppose the US imperialist policy of nuclear blackmail and nuclear threats":

China cannot remain idle in the face of the ever-increasing nuclear threats from the United States. China is conducting nuclear tests and developing nuclear weapons under compulsion...China is developing nuclear weapons for defense and for protecting the Chinese people from US threats to launch a nuclear war.²⁹

This declaratory policy has changed little in the subsequent 35-plus years that China has been a nuclear weapon state. In a July 1997 speech to the US Army War College, Lt. Gen. Li Jijun, Vice President of the PLA's Academy of Military Science, reiterated China's public position regarding its nuclear posture:

China's nuclear strategy is purely defensive in nature. The decision to develop nuclear weapons was a choice China had to make in the face of real nuclear threats. A small arsenal is retained only for the purpose of self-defense. China has unilaterally committed itself to responsibilities not yet taken by other nuclear nations, including the declaration of a no-first-use policy, the commitment not to use or threaten to use nuclear weapons against non-nuclear states and in nuclear-free zones...In short, China's strategy is completely defensive, focused only on deterring the possibility of nuclear blackmail being used against China by other nuclear powers.³⁰

The cornerstone of this publicly declared defensive position is China's NFU policy. Since first detonating a nuclear device in October 1964, China has consistently declared an unconditional NFU policy,³¹ combined with a policy of no threat or use of nuclear weapons against non-nuclear-weapon states (negative security assurances) (see below).³² Since that time, China has persistently proposed that nuclear-weapon states conclude a no-first-use agreement. The achievement of such an agreement was one of China's initial bargaining points in its CTBT negotiations. Later, China sought to gain such an agreement with the United States in return for a Sino-US detargeting pledge. Neither of these efforts succeeded, though the CTBT was completed and a Sino-US detargeting deal was reached. China and Russia, however, signed a bilateral NFU accord in September 1994.

Several questions, nevertheless, attend China's no-first-use pledge. First, such a pledge is highly symbolic--it is not verifiable and any violation would not be detected until too late. Second, as a practical matter, the NFU pledge may be less an altruistic principle, and more a simple reflection of the operational constraints imposed on Chinese doctrine by the country's qualitatively and quantitatively limited nuclear arsenal: China maintains an NFU pledge because it fits with the realities of nuclear weapons inventory. Finally, over the years there have been some indications that China's pledge may not be relevant to the first use of nuclear weapons on Chinese soil. Faced with the threat of a conventional Soviet invasion in the 1980s, Beijing's military strategists argued that the first-use of nuclear weapons on Chinese territory would not have violated its NFU pledge. Similarly, Johnston unearths evidence in Chinese military writings that loosely interprets the NFU pledge to possibly advocate launch-on-warning or launch-under-early-attack policies.³³

Negative and Positive Security Assurances
Another set of nuclear-weapon-related principles issued by the Chinese involves both negative and positive security assurances (NSAs and PSAs). As for NSAs, China's declaratory stance is clear:

China undertakes not to use or threaten to use nuclear weapons against non-nuclear-weapon States or nuclear-weapon-free zones at any time or under any circumstances. This commitment naturally applies to non-nuclear-weapon States Parties to the Treaty on the Non-Proliferation of Nuclear Weapons [NPT] or non-nuclear-weapon States that have undertaken any comparable internationally binding commitments not to manufacture or acquire nuclear explosive devices.³⁴

DF-21 IRBM TELs at National Day Parade in Beijing, 1 October 1999

Of note here is China's pledge not to use nuclear weapons against non-nuclear-weapon states under any circumstances; the US NSA, for example, is conditional in that the country retains the possibility of nuclear weapons use against non-nuclear-weapon states that take part in an attack on US territory, armed forces, or allies.³⁵

As for PSAs, China has agreed with the other four major nuclear weapon states (France, Great Britain, Russia, and the United States) to work within the Security Council to take "appropriate measures to provide . . . necessary assistance to any non-nuclear-weapon State that comes under attack with nuclear weapons."³⁶ The precise nature of the assistance is not elaborated, and the Chinese statement makes clear that this position does not in any way compromise its desire for a universal NFU pledge and unconditional NSAs, nor does it endorse the use of nuclear weapons.

Of related note, Chinese declaratory policy is particularly critical of the policy of extended nuclear deterrence, or so-called "nuclear umbrellas," provided by other nuclear-weapon states to their allies. In operational terms, this means China officially opposes the deployment of nuclear weapons outside national territories, and states that it has never deployed nuclear weapons on the territory of another country, a point that is not contradicted by any open-source evidence. When Japan sanctioned China for continued nuclear testing in 1995 and 1996 during the course of the CTBT negotiations, Beijing derisively dismissed Japanese censure as hypocritical, citing the fact that Japan enjoyed the protection of extended deterrence. China also opposes the threat or use of nuclear weapons against non-nuclear-weapon states, and has repeatedly called on nuclear-weapon states to agree to a legally binding, unconditional NSA accord.

In practice, if China adheres to its NSAs and PSAs, its deployments and targeting would presumably be focused only on nuclear-weapon states and possibly other states not party to the NPT or similar arrangements. Several questions, however, arise about China's commitments, particularly with regard to NSAs. First, like the NFU pledge, China's NSAs are not verifiable or enforceable. Second, the pledge apparently would not apply to such states as India, Israel, and Pakistan, which are not members of the NPT. Even if they joined, we question whether China's NSA would still apply to a country such as India, which, although not formally recognized by China as a nuclear-weapon state, certainly has attained such de facto status.

Finally, some observers question the need for certain Chinese deployments--such as the DF-21 series--insofar as its range and basing mean its possible targets largely comprise non-nuclear-weapon states. For example, as discussed in the text accompanying table 2, the DF-21s' basing and ranges suggest targets in such places as Japan, South Korea, Okinawa, the Philippines, or Vietnam, in addition to targets in the Russian Far East and India. If true, as asserted by Lewis and Xue, that China's target sets for the DF-3 included US bases in the Philippines and Japan, this targeting also runs contrary to Chinese NSAs. That the DF-3 and -4 series missiles are already capable of reaching Russian and Indian targets raises further questions as to the purpose of the DF-21 series in the context of Chinese NSAs.

Nuclear-Weapon-Free Zones
China has become a signatory to several nuclear-weapon-free-zone (NWFZ) treaties: the Treaty of Pelindaba (Africa NWFZ), the Treaty of Raratonga (South Pacific NWFZ), and the Treaty of Tlatelolco (Latin American NWFZ). During the ASEAN Regional Forum minister's meeting in July 1999 China stated it also would sign the Southeast Asian NWFZ Treaty. In its 1995 white paper on arms control and disarmament, the Chinese government stated its support for "the establishment of nuclear-free zones in the Korean Peninsula, South Asia, Southeast Asia, and the Middle East."³⁷

At a conference focusing on a Central Asian NWFZ convened in Tashkent in September 1997, a Chinese Foreign Ministry official heading the Chinese delegation listed seven principles related to the establishment of NWFZs. Among them, China insisted that "any other security mechanism" should not interfere with the nonnuclear status of a nuclear-weapon-free zone, including military alliance relationships. In addition, perhaps with reference to the South China Sea, the Chinese official declared that NWFZs should not include "areas where there exist disputes over sovereignty of territory or maritime rights." He also called on nuclear-weapon states to commit to an unconditional pledge not to use, nor threaten to use, nuclear weapons against NWFZs.

In practice, China's adherence to NWFZ pledges does not greatly affect its nuclear weapon deployments, especially given that it deploys no nuclear weapons abroad. China's signing and ratifying the Southeast Asian NWFZ Treaty presumably would place an added political onus on its ability to threaten or use nuclear weapons against such targets as Vietnam or the Philippines. Depending on caveats, if any, at the time of its signing, the treaty also could affect use by China in the South China Sea. However, the pledges of nuclear-weapon states to adhere to NWFZs are not verifiable, and some include escape clauses. For example, in signing the Treaty of Raratonga (South Pacific NWFZ), China stated that it could reconsider obligations in the event that other nuclear-weapon states or treaty parties violated the treaty.

Taken together, several points can be gleaned from these principles on NFU, PSAs and NSAs, and NWFZs. First, these long-held principles are consistent with a "defensive" posture and a qualitatively and quantitatively limited nuclear arsenal. Given the reality of Chinese nuclear forces, therefore, these pledges come at little to no real "cost" in terms of reductions, disarmament, or dramatic alterations to Chinese nuclear posture overall. Second, with the possible exception of some deployments, such as the DF-21-series ballistic missile, the nuclear principles noted here are consistent with a posture largely concerned with the other major nuclear-weapon states (especially the United States and the Soviet Union/Russia), as well as India. Third, nothing in these principles necessarily precludes China's nuclear weapons modernization program, but might place political limits on targeting and use options. Finally, although these principles may give us an overall understanding about China's formally stated views about when it would not use nuclear weapons, they provide no details about when they would.

Second Artillery Force Structure

Inferences drawn from theory and from declared nuclear principles may be incorrect. Theoretical inferences have not been tested under actual warfighting conditions, and China may purposely misrepresent its principles for the purpose of deception. To unravel these potential analytic stumblingblocks, in the next two sections we take a careful look at China's nuclear force structure and hardware, draw inferences from this empirical data to clarify questions about China's doctrine and capabilities, and reach understandings about China's overall posture from the vantage point that means most for strategic policy: how does the posture of the Second Artillery actually affect the security balance in strategic, theater, and conventional terms?

History
According to Chinese sources, the Chinese Missile Research Academy (also known as the Fifth Research Academy) was established in October 1956 under the direction of Qian Xuesen.³⁸ Ten research institutions were set up under the Fifth Academy to focus on the development of China's ballistic missiles. China began "copy production" of its first ballistic missile--a Chinese copy of a Soviet R-2 missile--in October 1958, and the missile was first tested three times in November and December 1960. Since that time the exact number of missile tests is difficult to discern through open sources, but, by the end of the 1960s, China had conducted at least 30 MRBM (the DF-2 and -2A missiles) tests at ranges of up to 1,500 km. Major milestones in China's nuclear force modernization are noted over the following pages.

DF-2 and -2A. After a failed flight test on 21 March 1962--in which shortly after takeoff, the missile erratically flew with its engine on fire before crashing near the launch pad--the Chinese successfully tested the DF-2 numerous times in June and July 1964 following the first success on 29 June 1964. Following a February 1965 decision to increase the range of the DF-2, an increase of 20 percent in the range was achieved for the DF-2A, beginning with its first successful tests in November 1965. On 27 October 1966, the Chinese launched a DF-2 with an armed, live nuclear warhead from the Shuangchengzi to an impact area in the Lop Nur testing area.³⁹ The DF-2 series, with ranges of 1,000 and 1,250 km, respectively, and a yield of 20Kt, was "sited in Northeast China and targeted on cities and US military bases in Japan."⁴⁰ China was believed to have produced a total of 100 missiles between 1965 and 1971,⁴¹ deploying approximately 50 missiles at one time.⁴² Retirement of the system reportedly began in 1979 and was completed by 1990.⁴³

DF-3/3A. The DF-3 was China's first indigenously developed ballistic missile.⁴⁴ Official calls for an intermediate-range missile began in the summer of 1964, with formal approval to commence the R&D process granted in May 1965. After the difficulties with the DF-2's "volatile liquid oxygen fuel," the DF-3 was reportedly the first of a series of Chinese missiles designed to utilize storable liquid fuels.⁴⁵ The more stable fuels were also meant to improve readiness because the Cuban Missile Crisis had illustrated that missiles with nonstorable fuels (such as the SS-3s and SS-4s on Cuba) were ineffective in international crises, since they took long to prepare for launch and could not be maintained at high alert levels for extended periods of time.⁴⁶ The missile was first successfully flight-tested on 26 December 1966⁴⁷ although it was not until a third flight test in May 1967 that the Chinese were fully satisfied. Several years were required for the missile to be deployed, though the exact deployment date is in dispute. The IISS Military Balance lists a 1970 deployment, although the Nuclear Weapons Databook asserts a May 1971 deployment.⁴⁸ The DF-3 was designed to carry a 2,150-kg warhead to a distance of 2,650 km (intended, when first conceived in the early 1960s, to hit US military bases in the Philippines). Perhaps as many as 36 of these missiles were sold to Saudi Arabia in the late 1980s, as the slightly longer range (2,850 km) DF-3A was tested in December 1985 and January 1986, and commissioned in that year to replace the DF-3.

DF-4. The Chinese intermediate-range ballistic missile (IRBM) DF-4 was a more difficult undertaking. With a required range of up to 4,000 km ("to strike the B-52 base on the US island of Guam"⁴⁹), the Chinese formally authorized development of the missile in May 1965. This was to be China's first two-stage rocket (using the DF-3 as the first stage), and required technical breakthroughs in such areas as engine reliability in the near vacuum of the upper atmosphere, developing high-altitude test simulator beds, developing more heat-resistant materials, and improved guidance systems for the longer range missile. The first flight test of the missile failed in November 1969--the second stage was not ignited/separated and the missile self-destructed--but the missile was successfully tested in January 1970. According to Lewis and Hua, because of the Sino-Soviet Ussuri River clashes in late 1969, the range of the missile was subsequently raised to 4,500 km (and eventually attained a 4,750-km range) in order to reach Moscow.⁵⁰ According to Norris, et al., it "was initially planned to be deployed in silos but recognition of its vulnerability lead to reconsideration of rail-mobile basing."⁵¹ From 18 September to 2 October 1975, the Chinese conducted DF-4 rail-mobile tests over 8,000 km in 10 provinces.⁵² In 1977, the Chinese finally chose a deployment plan based on cave storage, whereby the missiles would be brought out of the cave for erecting, fueling, and firing.⁵³ A full-range test flight occurred on 2 August 1980.⁵⁴

DF-5 and DF-5A. China formally began development of the intercontinental ballistic missile (ICBM) DF-5 in March 1965; its progress also was delayed by the exigencies of the Cultural Revolution. A first flight test was conducted on 10 September 1971, although this test--entirely within Chinese territory--had to be conducted across a shorter range and different trajectory than the missile was designed for. Not until 18 May 1980--a full 15 years after the missile began development--could the Chinese conduct a full-range flight test from the mainland into the Western Pacific. This test was followed by a second full-range test on 21 May 1980.

Solid-fuel Missiles. According to Chinese sources, work on solid-fuel missiles in China date back as far as October 1956, when Qian Xuesen first began to set up the Fifth Research Academy.⁵⁵ First strides were made by the late 1950s and early 1960s in developing and testing prototype solid propellant. Static tests were made with 300-mm-diameter engines in 1965 and on 1,400-mm-diameter engines in December 1966.

Initially, work was conducted with the intention of using solid fuels for a single-stage rocket. But, deeming such missiles' ranges as too short, in March 1967 Chinese military-technical authorities decided to go forward in the development of two-stage, "medium-range" solid-fuel surface-to-surface strategic missiles, to be mated with the ongoing nuclear submarine under development (the submarine-based missile was later to evolve into the DF-21 land-based system). Again, owing to the exigencies of the Cultural Revolution, Chinese sources note that serious work on the solid-fuel missile program did not begin until August 1978.⁵⁶ After launch equipment tests in April and May 1984, followed by launch tests in May 1985 (DF-21) and May 1987 (DF-21A), these systems finally became fully operational in the early 1990s. This accomplishment culminated a nearly 30-year development effort.

Another version of the DF-21, the submarine-launched JL-1, was first tested from a submerged conventionally powered Golf-class submarine on 7 October 1982, but this launch failed as the missile lost control soon after ignition and self-destructed. On 12 October 1982 the missile was successfully launched from the submerged Golf submarine. As for launching from China's nuclear-powered submarine, the missile failed its first test on 28 September 1985, again turning over and self-destructing. Not until three years later, on 15 September 1988, did a fully successful JL-1 launch take place from the submerged Xia-class nuclear submarine; a second successful test was conducted on 27 September 1988, culminating a difficult 30-year development process for Chinese SLBMs dating back to the late 1950s. According to open sources, China, since 1988, has not test launched its JL-1 from the Xia-class nuclear submarine.

DF-15 SRBM Launch From TEL

By the early 1990s, China also had tested and begun deployment of two short-range, nuclear-capable ballistic missiles, the DF-15 (CSS-6/M-9) and 300-km-range DF-11 (CSS-X-7/M-11).⁵⁷ Both missiles were originally developed for export; only after China pledged not to export these missiles were they incorporated into the Second Artillery.⁵⁸ The DF-15 has been operational since 1994⁵⁹ and was tested approximately 10 times as part of the missile exercises China conducted around the Taiwan Strait in July-August 1995 and March 1996.⁶⁰ The CSS-X-7/M-11 probably was not deployed with Chinese forces by October 1998,⁶¹ though some foreign sources familiar with the PLA believe that the 300-km DF-11 already has been fielded by at least two PLA group armies.⁶² The 1999 DoD Report to Congress on the Security Situation in the Taiwan Strait reported thatan improved, longer range version of the DF-11 might be under development,⁶³ which later was verified by the 1 October 1999 military parade in Beijing.⁶⁴

Testing. China's 32-year testing program is the smallest of the five major nuclear powers, with 45 tests between 1964 and 1996. By comparison, the United States tested more than 20 times as much, with over a thousand blasts over a more than 50-year program. This static examination of the total number of tests gives us evidence of comparative scale, but changes in annual averages can also signal intent. The amount of Chinese testing increased marginally after 1979 from 1.3 to 1.7 tests per year, but American testing between 1979 and 1992 averaged 13.6 detonations per year.

By previous standards, Chinese testing accelerated significantly in the mid-1990s, though this intensified program was probably linked to China's stated intention from early 1994, at the outset of CTBT negotiations, to conclude a test ban by the end of 1996. This timeline suggests that a political decision to sign the treaty in principle had been made by 1993 or earlier and may have intensified in the face of increasing international condemnation of China's test program, which continued throughout the CTBT negotiation process.⁶⁵ The pace of Chinese testing certainly intensified over the period 1994-96. China's six tests over a 25-month period (June 1994-July 1996, which overlapped with the negotiations of the CTBT) more than doubled China's average testing pace. For the only time in Chinese history, nuclear weapons were tested twice in three successive years.⁶⁶ Also, this period marked the only time in Chinese testing history that blasts occurred in either July or August--outside the typical Chinese testing "season"--which also indicates a sense of urgency within the military and nuclear scientific communities.⁶⁷ Finally, the initial bargaining positions put forth by China--such as on verification and inspection procedures and leaving the door open to peaceful nuclear explosions--offered the military the possibility of further testing and may have succeeded in stalling the negotiation process, thereby granting China's testing program more time. Almost immediately after China announced in early June 1996 that it would have one more test, it stepped away from its objections to the treaty and allowed the negotiations to conclude.

The Cox Report strongly suggests that the combination of nuclear espionage and the intense series of underground tests described above has accelerated the PRC's attainment of advanced, MIRVable small warheads, but some important caveats must be offered. First and foremost, the warheads employed by US nuclear forces are highly complicated devices that are extremely difficult to build. They are the product of decades of dedicated research and development, using some of the most advanced techniques available. As such, there are limits on the amount of benefit that can be wrought from simply obtaining the designs for these weapons.⁶⁸ As one sober observer writes,

China's theft of the W-88 design used for the US Navy's Trident missile warhead, for example, does not allow its engineers to reconstruct the thousands of parts and electronic components that form the completed weapon. Even the computer codes China may have obtained are mathematical models of the physical characteristics of a nuclear explosion. They cannot be used to design and manufacture a warhead. Chinese engineers may well have obtained some useful information, but they lack the data and experience required to design and build replicas of sophisticated US warheads from the stolen information.⁶⁹

This line of reasoning is supported by the damage assessment by the intelligence community, which concluded that China had not deployed any operational system using the stolen designs, despite a lapse of more than 10 years since the alleged espionage.⁷⁰ Passage of the CTBT could have locked this situation in place for the foreseeable future, although its defeat in the Senate should prepare us for the likelihood of a resumption of Chinese testing, and, thus, the possible conquering of important developmental hurdles in the area of smaller warheads.

Current Force Structure
As a result of this historical progression, one of the most intriguing aspects of China's nuclear weapons program has been its quantitatively and qualitatively limited nature over time. These limitations are characterized in practice by a relatively small number of warheads; technically and numerically limited delivery vehicles; an overwhelming reliance on land-based systems; persistent concerns over the arsenal's survivability, reliability, and penetrability; and a limited program of research, development, and testing.

Table 1
Range of Estimates of Chinese Nuclear Weapon Delivery Vehicles

China's current nuclear weapons arsenal totals about 400 devices, 300 of which consist of warheads and gravity bombs for use on its strategic "triad" of land-based ballistic missiles, bomber and attack aircraft, and one nuclear-powered ballistic missile submarine (SSBN) (see table 1).⁷¹ According to the US Defense Department, over 100 warheads are deployed for use on China's ballistic missiles, with additional warheads in storage.⁷² The Chinese SSBN is thought to deploy 12 single-warhead missiles. The remaining warheads reportedly consist of about 100 tactical nuclear weapons, including bombs for tactical bombardment, artillery shells, atomic demolition munitions, and possibly short-range missiles.⁷³ China has the capability to increase the size of its nuclear arsenal using its existing stockpile of fissile material. One source indicates that China has an inventory of between 2 and 6 tons of plutonium and 15 to 25 tons of highly enriched uranium.⁷⁴ Iain Johnston estimates that China has enough fissile material to double or triple its arsenal.⁷⁵ According to the US Defense Department, however, "China is not currently believed to be producing fissile material for nuclear weapons, but it has a stockpile of fissile material sufficient to increase or improve its weapon inventory."⁷⁶

In addition to ballistic and cruise missiles, according to the US Defense Department, "China also has a variety of fighters, bombers, helicopters, artillery, rockets, mortars, and sprayers available as potential means of delivery for NBC [nuclear, biological, and chemical] weapons."⁷⁷ China is working to modernize its capabilities in terms of ballistic and cruise missiles, bombers, and multirole aircraft, but relies upon deterrent systems and technologies that are at least 20 years behind the capabilities of the four major declared nuclear powers. According to Chinese sources, the overall capabilities of the strategic rocket forces have advanced in recent years owing to better, more modern training, the development of strategic missile simulator training, improvements in technical reconnaissance, weather forecasting, geographical surveying, antichemical warfare and logistics support, and the introduction of some "1,000 technological research results."⁷⁸ Estimates of Chinese nuclear-capable ballistic missile forces are shown in table 1. Estimates vary as to the exact number of these missiles, but China benefits from a large, well-developed infrastructure for the development and production of ballistic missiles.

From table 1, the Chinese nuclear force structure clearly is primarily land-based, relying on a range of missile systems. On the short-range end of the land-based missile spectrum, China reportedly possesses several hundred DF-11s and DF-15s, which have ranges of 300 km and 600 km, respectively. The DF-15 can deliver a 500-kg payload to a maximum range of 600 km, with a CEP (circular error probable) of 600 meters.⁷⁹ The DF-11 reportedly has an 800-kg warhead and a 150-meter CEP.⁸⁰

In the medium- to intermediate-range inventory, the PRC fields three types of missiles (DF-3A, DF-4, and DF-21A). Deployed in caves and valleys to increase its survivability, China's liquid-fueled DF-3As have a range of 2,800 km and reportedly carry a single warhead with an estimated yield of 1-3 megatons.⁸¹ The liquid-fueled DF-4s, with a range of 4,850-5,500 km, are deployed in silos and tunnels and have a single warhead with an estimated yield of 1-3 megatons.⁸² The solid-fueled, mobile DF-21As have a range of 1,800 km and a 600-kg warhead with a yield of 200-300 Kt.⁸³

In the ICBM category, China's DF-5 ICBMs can reach targets in all of the United States.⁸⁴ Each silo-based missile carries a single warhead, with an estimated yield of 3-5 megatons.⁸⁵

In its weaker second leg of the triad, China has deployed 12 single-warhead JL-1s, a submarine-launched ballistic missile (SLBM) with a range of 1,700 km aboard its one Xia-class nuclear submarine.⁸⁶ These missiles have faced operational difficulties, and not until 1988 were they first test-launched successfully from the Xia-class submarine. According to Paul Godwin, "this troubled ship has spent most of its time docked or in local waters and is not considered operational."⁸⁷ The limited range of the missile, the problems it has had in deployment and operation, and the limited experience of the Chinese in long-range submarine operations limits the value of this system as a strategic weapon. Beijing also may have learned some valuable negative lessons from the experience of the Soviet Union, whose SSBN force was forced to retreat to bastions by a superior US attack submarine fleet.

China's bomber and ground-attack fleet is made up of two aircraft, both of which are based on 1950s Soviet designs: the Hong-6 (H-6) bomber (Soviet Tu-16 design) and the Qian-5 (Q-5) ground attack aircraft (a redesign of Soviet MiG-19). Given the nascent state of China's in-flight refueling capability, the maximum ranges of these aircraft are approximately 3,000 and 800 km, respectively. China reportedly halted production of the H-6 in 1982, and now deploys between 100 and 120 H-6s (some in a nuclear role). China deploys over 400 Q-5 aircraft (perhaps 30 currently in nuclear role).⁸⁸

Toward an Organic View of Chinese Nuclear Force Structure
Viewed as an organic whole, the Chinese nuclear force structure seems to defy simple categorization as either limited or minimal deterrence. Instead, the multifaceted force is made up of strategic, theater, and tactical systems of varying range, accuracy, and yield. The small ICBM force, anchored by the DF-5 family of missiles, appear to be second-strike minimal deterrence forces. The theater systems are unlikely to be used in a second-strike, minimal deterrent role following a preemptive strike. Instead, theater systems look like offensive systems meant to strike US forces and bases in Asia to degrade conventional capability. The short-range, ballistic missile forces, which are also nuclear capable, further confuse the situation by serving a variety of conventional warfighting and nuclear warfighting roles. Perhaps the best way to understand the nature of this multifunction force structure is to deductively infer the purpose of each element in the force by examining range and deployments, payloads and CEP, readiness, and C4I structure.

Table 2
Suspected Chinese Strategic Missile Bases
(Derived From Open Sources)

Ranges, Deployments, and Targets. The Chinese nuclear force inventory encompasses a wide variety of ranges, and the deployment of these forces offer a wide variety of potential targets. The range and basing of China's missiles are summarized in table 2.

From the locations of these bases and the ranges of their deployed missiles, several inferences can be drawn about the likely target for these missiles. The DF-3As and DF-21s of Base 80301 probably are targeted on Japan, South Korea, Okinawa, or the Russian Far East. The DF-15s of Base 80302 are almost certainly aimed at Taiwan. The DF-3As and DF-21s of Base 80303 probably are targeted against countries south and southwest of China, including the Philippines, Vietnam, and India. The DF-5s of Base 80304 are the major CONUS-oriented systems, while the DF-4s of both Base 80304 and Base 80305 might be aimed at Hawaii. Finally, the DF-3As and DF-4s of Base 80306 likely are targeted at sites in the former Soviet Union, including Moscow, or possibly India.

How Did the Structure Evolve to This Arrangement? Lewis and Hua maintain that China's nuclear weapons program "proceeded without such strategic guidance" and that "until the early 1980s, there were no scenarios, no detailed linkage of the weapons to foreign policy objectives, and no serious strategic research."⁸⁹ They even go so far as to say that neither the "Chinese leader nor his senior colleagues on the Central Military Commission considered, communicated, or authorized the investigation of the broader strategic purposes of the program."⁹⁰ As Lewis and Hua predicted, we have difficulty believing this to be true. From an examination of the sources of their collected works, no one can doubt the authors' access to critical personnel or documents from China's nuclear programs or missile programs, though the level of citation from central leadership documents is considerably lower. Although we doubt that the first generation of leaders, especially Mao, understood the scientific or technical aspects of nuclear combat, they at least were able to articulate the strategic targets for these weapons and task the weapons complex accordingly. Indeed, the authors seem to contradict themselves when they relate stories wherein researchers are told the specifications for specific missiles (i.e., range, payload, etc.) by central authorities, who then later change the range and payload requirements for individual missiles to reflect new strategic goals. For example, they assert that the military commission in 1970 commanded that the range of the DF-4 be increased from 4,000 km to 4,500 km, "bringing Moscow within range of bases in Da Qaidam, Qinghai Province."⁹¹ This story, along with others in the narrative about the sequential development of missiles capable of hitting the Philippines, Guam, Hawaii, and the United States, suggest that someone, somewhere, at a central level was making decisions about the strategic purpose and direction of various missile systems, which was then reflected in the seemingly logical pattern (defined as matching geographic location with range to target) of base and missile deployments.

One important dilemma that confronts any analyst trying to understand the overall nature of the Chinese nuclear force posture is reconciling the mixture of strategic and theater systems with claims of either minimal or limited deterrence. Comparative cases of nuclear force structure evolution, however, offer clues about China's intentions. In the Soviet case, we note that Moscow did not draw a sharp distinction between its strategic and theater nuclear weapons systems. The best example of this was the road-mobile SS-20, which was developed to decouple the United States from its allies in Europe and Asia by holding theater targets at risk and preventing Washington from defending allies. The Soviets referred to this combination of strategic and theater nuclear weapons as the "seamless web of deterrence." Is the same thing happening in China? Clearly, China and the former Soviet Union share some commonalties in their strategic environment and goals. Like Russia, China seeks to decouple the United States from its allies in the region, especially Japan and South Korea, by using the threat of theater nuclear weapons. In recent years, this threat has become particularly important in a Sino-US conflict over Taiwan, which could escalate to the point that it threatens to split the US-Japan defense alliance. The United States, however, withdrew its theater nuclear forces in 1991. How has this situation changed the rationale for the DF-21A and other Chinese theater nuclear forces, because they no longer have a second-strike role?⁹² To explicate this situation, a deconstruction of the Chinese force is required.

Payloads, CEP, and Targeting. Until the DF-31 comes online, the Chinese strategic nuclear force is dominated by missiles with high yield warheads and large CEPs. For example, the DF-4 ICBM has an estimated yield of 1-3 megatons and a CEP of almost a mile.⁹³ The mainstay of the Chinese ICBM force, the DF-5, is more accurate but still has a yield of 3-5 megatons and a CEP of more than a quarter of a mile. This combination of high yield with low accuracy suggests that the force is designed for countervalue, or "city-busting" attacks against "soft" targets such as concentrated population centers, and other locations of political and economic value.⁹⁴ Counterforce warfighting, by contrast, requires far more accuracy than offered by these systems.

Readiness and Survivability. In the past, the limited numbers, low level of readiness, and slow response times of China's land-based missiles and bombers left China vulnerable to an overwhelming and incapacitating first strike. China does not currently have space-based or land-based early warning assets. A senior US intelligence official has confirmed that Chinese missiles are usually unfueled and unmated to their warheads.⁹⁵ Furthermore, the process of loading the liquid fuel tanks and installing the warheads can take two to four hours.⁹⁶ Because of the lengthy prelaunch exposure times of more than 2 hours for the DF-3A, decisions were taken that led eventually to operating the DF-4 from caves and the DF-5 from silos.⁹⁷ Although cave and silo basing reduces prelaunch exposure, the basing mode could not significantly reduce the overall preparation time for launch, including fueling, arming, positioning (in case of non-silo-basing), targeting and range-setting, and other preparatory checks.⁹⁸ Given these time constraints, the Chinese DF-3A, DF-4, and DF-5A in today's arsenal may still require from 1 to 2 hours to launch. From this incomplete data, we tentatively infer that the Chinese nuclear force is incapable of launch-on-warning or launch-under-attack. This readiness and survivability level is consistent with a minimal deterrent posture.

DF-31 ICBM TELs. The DF-32 Is Still in the Test Launch Stage

China has also sought to improve survivability by establishing a credible triad. As early as the mid-1950s, China began developing a sea-based deterrent, though this small program continues to face a number of serious technological obstacles.⁹⁹ China has held only one known SLBM test from the Xia-class submarine, and the existence of only a single boat obviates the possibility of regular patrolling.¹⁰⁰ Efforts to further integrate Chinese bombers into the triad have been impeded by the vulnerability of PRC airfields and the high cost of modern aircraft capable of penetrating advanced air defenses.¹⁰¹ In addition, Chinese nuclear-capable bombers are limited in range and are highly vulnerable to sophisticated air defenses, making it unlikely that the bomber force would be effective in a nuclear delivery role against either Russia or US forces in the Western Pacific region.¹⁰² Despite strenuous efforts, therefore, the sea-based and bomber-based legs of China's triad are still relatively unreliable, especially in the context of intercontinental nuclear combat with the United States. As a result, China has been forced to focus on ensuring the survivability of its land forces by deploying road-mobile, solid-fuel systems.

C4I Structure. The Second Artillery (SAC) is tasked with implementing the reliable and secure command and control of China's nuclear and conventional missile forces.¹⁰³ The SAC was formally established in 1966, based upon a "special" artillery corps formed in 1958 following the Chinese decision to develop nuclear weapons. The SAC is a separate service arm, distinct from the army, navy, and air force. The central command and control center for all Chinese forces, including SAC, is located is Xishan, in the hills west of Beijing, where strategic operational orders originate. Direct communication with China's six launch bases would be passed through the SAC headquarters and its communications regiment. We must note that this system bypasses China's military region commands, and connects directly to base commands. Base commands, in turn, communicate with their respective launch brigades. The SAC reportedly operates about six launch bases, each led by a major general. Each base has two to three missile brigades each commanded by a colonel, with each brigade operating one type of missile. These brigades consist of up to four launch battalions (see table 2).

At a political level, ultimate authority to use nuclear weapons is "subject to the unified command of the Central Military Commission. Only the commission's chairman (currently Jiang Zemin, who is also head of the Chinese Communist Party and the Chinese President) has the power to issue an order to use such weapons after top leaders reach a consensus on the issue."¹⁰⁴ However, it is likely that such a decision would require a consensus decision within the Central Military Commission and other senior military elders.¹⁰⁵

As for the technical aspects of Chinese nuclear C4I, little open source information is available as to the precise systems employed to ensure safe and reliable communication between the central leadership and the launch bases. In recent years, however, reports increasingly have surfaced in the open literature describing various new technologies and systems that help strengthen China's command and control system. In some cases the "breakthroughs" reported suggest that the past level of command and control structures was not particularly advanced. For example, the official People's Liberation Army Daily in early 1998 noted that the SAC "after three years of arduous work" developed a new digital microwave communications system which now allows for a secure "all-weather" communications for missile launch. "With the new system," the article notes, "the Second Artillery will no longer be affected by natural conditions such as weather."¹⁰⁶

At the same time, however, the Pentagon reports that "China has made significant efforts to modernize and improve its command, control, communications, computers, and intelligence infrastructure."¹⁰⁷ Given the importance of nuclear weapons to Chinese security, we assume that similar advances in C4I modernization have occurred in the strategic rocket forces. Some evidence indicates, for instance, that the Second Artillery seeks to connect much of its infrastructure with secure, landline fiber-optic cable.¹⁰⁸ Moreover, open-source reports detail the deployment of an "automated command and control system."¹⁰⁹ From these changes, we can infer desire for greater survivability and positive control of nuclear weapons. They probably also reflect a greater desire for operational security, as well as enhanced denial and deception against increasingly advanced national technical means of other countries. By itself, however, the modernization of Chinese nuclear C4I does not automatically imply that the force is transitioning to a flexible response, counterforce footing. The changes might signal desire for eventual launch under attack (LUA) capability, but the current inventory of missiles and the next generation of replacements are not capable of the reaction times necessary for such a capability. More likely, the C4I modernization program is meant to improve the credibility of China's minimal deterrent posture in the short to medium term.

Future Nuclear Force Posture

Doctrine
Over the past decade, certain indicators suggest that these long-held aspects of Chinese nuclear weapons doctrine may be undergoing some reconsideration.¹¹⁰ As Paul Godwin argues,

Minimum deterrence, which uses a single countervalue punitive strike on cities to deter, is seen by many Chinese strategists as passive and incompatible with what they see as a future requirement for more flexible nuclear responses.¹¹¹

One observer argues that, consequently, some Chinese military planners are considering a shift to a "limited" deterrent posture, which could include the introduction of limited warfighting capabilities; improved command and control and early warning systems; smaller, survivable, mobile, more accurate, and diverse cruise and ballistic missile nuclear delivery systems; possible abandonment of the NFU policy; missile defenses; and the addition of counterforce targets.¹¹² This view has gained backing in other detailed research that notes that "China's strategic modernization R&D [research and development] supports this shift toward a limited warfighting approach to nuclear warfare."¹¹³ Such a capability would enable China to respond to "any level of nuclear attack, from tactical to strategic."¹¹⁴

As the previous pages suggest, however, from a strictly doctrinal perspective, such a shift probably will await shifts in the domestic political hierarchy and its view of the outside world, factors that have consistently driven Chinese doctrinal choices. Moreover, as noted in the previous section on force structure, technological constraints will remain one of the foremost drivers determining the direction of doctrine in the near term.

Rather than force a stark analytic choice between either a doctrine of "minimal deterrence" or one of "limited deterrence," drawing out two important nuances to better understand this debate is more logical. First is to recognize the differences between "operational doctrine" and what we might call "aspirational doctrine" in the Chinese context. Second is to recognize that the Second Artillery--which oversees strategic nuclear, theater nuclear, and conventional missiles--more likely operates on three levels of doctrine: credible minimal deterrence with regard to the continental United States and Russia; "limited deterrence" with regard to China's theater nuclear forces; and an offensively configured, preemptive, counterforce warfighting posture of "active defense" or "offensive defense" for the Second Artillery's conventional missile forces.

Force Structure
Various governmental reports suggest that Chinese nuclear force structure will increase in numbers and quality. In 1995, then Secretary of Defense William Perry stated that China "has the potential to increase the size and capability of its strategic nuclear arsenal significantly over the next decade."¹¹⁵ According to the US Department of Defense in 1997, "China probably will have the industrial capacity, although not necessarily the intent, to produce a large number, perhaps as many as a thousand, new missiles within the next decade."¹¹⁶ General Hughes, then Director of the DIA, testified in 1999 that "the number of Chinese strategic missiles capable of hitting the United States will increase significantly during the next two decades."¹¹⁷ Publicly released estimates of the number of ICBMs capable of reaching the United States range from "tens"¹¹⁸ to the Cox Committee's ambitious estimates of "up to 100" ICBMs with 1,000 MIRVed warheads by 2015.¹¹⁹ According to the Pentagon, "China plans to begin production and deployment of at least one solid-propellant ICBM that will provide China's strategic nuclear forces [with] improved mobility, survivability, accuracy, and reliability."¹²⁰

Two principal impetuses are behind the modernization of the Chinese nuclear force structure. The first is the predictable process of replacing aging weapons systems with more modern counterparts. Most of China's operational missile forces, especially the CONUS-capable ICBMs, are 1950s-vintage liquid-fueled systems. As General Hughes has testified, "China's strategic nuclear force is small and dated, and because of this, Beijing's top military priority is to strengthen and modernize its strategic nuclear deterrent."¹²¹ This effort has been assisted and accelerated in part by the ready access to technologies now available from Russia. The second driving factor behind Chinese modernization is a rising concern about the survivability of its nuclear deterrent, particularly given the prospect of the Strategic Defense Initiative in the 1980s and now the deployment of theater and national missiles defenses by the United States. Chinese perceptions about the survivability of its force were also undermined by Desert Storm, which highlighted the ability of US conventional forces to destroy fixed targets with precision-guided munitions and the concomitant inability of those same forces to destroy mobile targets. This realization no doubt reinforced the perceived desirability of modern, road-mobile nuclear forces.

The two principal programs in this modernization effort will be the DF-31 and the DF-41.¹²² The mobile, solid-fuel DF-31 will have a range of 8,000 km, and carry a payload of 700 km. The origins of this missile are controversial. Lewis and Xue argue that the First Academy drew up plans beginning in 1974 to develop not only the JL-1 SLBM, but three other solid-propellant missiles as well over the subsequent decade, namely the DF-21, DF-21A, and the JL-2 SLBM.¹²³ Another source claims that the DF-31 missile was an outgrowth of the DF-23 road-mobile, solid-fueled program, which began development in 1978 as a land-based missile, and was then modified to also serve as the basis for a submarine-launched SLBM, known as the JL-2. To confuse matters even further, a different Lewis article asserts that the R&D for the DF-23 began in August 1970, during "a particularly tense moment in Sino-Soviet confrontation."¹²⁴ Regardless of its development path, the DF-23 was renamed the DF-31 in January 1985, although the designation JL-2 was not changed. In August 1999, China publicly declared the first full flight test of the DF-31.¹²⁵ We expect that the DF-31 will be deployed perhaps by the early 2000s.

The planned follow-on to the DF-31, the DF-41, was officially initiated in July 1986.¹²⁶ The three-stage, solid-propellant ICBM will have a range of 12,000 km, thus making it capable of striking all targets in the CONUS. It is therefore the logical replacement to China's aging DF-5 force, which Beijing will begin replacing around 2010. According to Lewis and Hua, the final basing mode for the DF-41 is still unclear, although it will be stored in caves and is likely to be deployed on a road-mobile TEL.

Some reports indicate that China will launch a major effort to develop and construct a follow-on to the Xia-class nuclear ballistic missile submarines to be deployed after 2000. The next-generation submarine, the 09-4, probably would deploy 16 of the new JL-2 SLBMs, with a range of about 8,000 km.¹²⁷ However, political and technological constraints may delay or even suspend the deployment of this boat.¹²⁸

Implications

Mobility. Despite yeoman effort, the Chinese largely have failed to field a credible triad. Instead, the force remains highly unbalanced, with land-based missiles predominant over bombers and SLBMs, especially in the intercontinental category. As a result, Beijing has been forced to improve the survivability of its land-based missiles. Apart from the addition of solid fuels and improved C4I infrastructure, the Chinese began to move from silos and caves to a road-mobile force with missiles loaded on transporter-erector-launchers (TELs) as early as the 1970s.¹²⁹ With the planned deployments of the DF-31 and DF-41 ICBMs over the next 10 to 20 years, the Chinese nuclear inventory will thus become increasingly mobile over time. This move will have the effect of enhancing the credibility of China's minimal deterrent posture, as long as such a large force size asymmetry exists between China and the larger nuclear powers. Moreover, the deployment of the DF-31 and DF-41 theoretically increases deterrence stability with other nuclear powers by making China's force more survivable.

Solid Fuel. One impediment to greater flexibility and survivability in the Chinese force were the hazards associated with volatile liquid propellants.¹³⁰ The move to solid fuel increases the credibility of the Chinese force by improving reaction times, thus raising its overall readiness level. As Godwin points out, however, solid fuels also "contain less thrust than liquid fuel, requiring China to develop smaller, lighter warheads with much better yield-to-weight ratios than its older weapons."¹³¹

C4I Modernization. Speaking in 1999, DIA Director Hughes testified to Congress that China was actively engaged in "upgrade programs" for its nuclear C4I.¹³² Overall, the modernization of Chinese nuclear C4I increases the credibility of the Chinese force by strengthening command and control. Specifically, it enhances the leadership's positive control over the force, increasing the probability that the National Command Authority could survive an attack and respond. In the paradox of nuclear strategy, this development actually increases deterrence stability between China and other nuclear powers.

Accuracy. There is reason to believe that the Chinese SAC is attempting to improve the accuracy of its strategic rocket forces. Presurveyed launch sites increase the potential accuracy of the new mobile systems. Chinese research institutes are reportedly attempting to increase precision by developing better gyros and inertial measurement units.¹³³ According to the Pentagon, China is using the Global Positioning System (GPS) to make "significant improvements" in its missile capabilities. As an example, the DoD cites the use of GPS for midcourse guidance correction to improve missile accuracy, and also asserts that such satellite updates will "increase the operational flexibility of China's newer mobile missiles."¹³⁴ A RAND study on this subject concluded that GPS-aiding of ballistic missile guidance could improve accuracy by 20-25 percent.¹³⁵ Greater accuracy might signal a desire for eventual counterforce capabilities, although force size will be an important constraint on successful transition to a more offensive posture.

Greater Numbers. The Cox Report and other analyses predict that the Chinese nuclear force structure is likely to increase in size, and therefore pose a greater threat to the United States.¹³⁶ Why would the Chinese force increase in size? Increasing numbers of Chinese missiles would cause an opposing force to have greater difficulty in "decapitating" the Chinese force, which has been a prevailing fear since the beginnings of the program. The fear only has become more frantic in an age of growing American predominance in space-based reconnaissance. More Chinese missiles might signal a possible shift from a retaliatory countervalue posture to an offensive counterforce posture, particularly if accompanied by necessary improvements in accuracy. According to Godwin, a sufficient number of weapons could permit China for the first time to attempt intrawar escalation control because Beijing would retain enough forces to respond at a higher level if the aggressor chooses to escalate a nuclear exchange.¹³⁷

An increase in missiles is also the logical response to the deployment of theater (TMD) and national missile defenses (NMD) among the United States and its allies, which the Chinese view as an organic whole rather than separate programs (as one Chinese arms controller put it, "two sides of the same coin"). Proponents of TMD/NMD point out that the Chinese already are modernizing their missile forces, so defenses are not to blame for increases in the quality and quantity of the Chinese force. This claim probably is true but must also be accompanied by an honest recognition that TMD/NMD deployment is likely to accelerate this effort and push the Chinese to spend more money on such relatively cheap antimissile defense accessories as countermeasures and decoys. Perhaps the only good news is that limited increases in Chinese missiles would paradoxically increase deterrence stability between China and other nuclear powers and enable China to maintain a no-first-use principle by reducing the likelihood that the PRC's force could be destroyed in an all-out preemptive attack.

At the same time, we must also entertain the possibility that the new generation of missiles are meant only to replace the aging veterans of the fleet, particularly the DF-4 and DF-5. If the Chinese eventually exchange the road-mobile, solid-fueled DF-31s and DF-41s for these liquid-fueled, silo- and cave-based missiles on a one-to-one basis, or even two-to-one basis, then the net result is ceteris paribus an increase in the credibility of China's previously suspect minimal deterrent, not necessarily a fundamental shift to an offensive posture. Moreover, as the significant delays in the IOCs of past systems and the inaccurate estimates of DF-31/DF-41/DF-25 deployments in Lewis and Hua's 1992 article attest, we should not be overly optimistic about the production timelines or output estimates offered by the Chinese for the rollout of the next generation of missiles. Rather, we should maintain a sober view of the impressive but sometimes erratic production cycles in the Chinese missile system.

MIRVing? Since the late 1980s, China has conducted a series of smaller yield tests, apparently intended to develop smaller, lighter warheads with an improved yield-to-weight ratio,¹³⁸ although this trend could be traced as far back as 1970.¹³⁹ Most analysts agree that the purpose was to develop new warheads for single placement on China's next-generation solid-fuel ICBMs (DF-31 and DF-41) as well as ensure the safety and reliability of new warhead designs.¹⁴⁰ The antecedents of the DF-31 and DF-41 programs, which were initiated in the early 1970s, were the beginning of a move to develop mobile forces, which required the development of smaller missiles, which in turn required smaller warheads.

Other observers have added an additional, controversial motivation for the testing of smaller warheads--the development of a multiple warhead capability, possibly MRV or even MIRV.¹⁴¹ The Cox Committee, for example, concluded that "the PRC has demonstrated all of the techniques that are required for developing a MIRV bus, and that the PRC could develop a MIRV-dispensing platform within a short period of time after making a decision to proceed."¹⁴² Often, this desire is linked to a perceived future Chinese intent to develop flexible response, counterforce-oriented nuclear forces, though the smaller warheads could also be used as MIRVs on the existing DF-4s and DF-5As. Significant evidence suggests that the Chinese have been actively interested in developing multiple warhead technology for more than 20 years.¹⁴³ The current small size of the Chinese force and the mainstream projections of the size of the future force, however, make unlikely China's seeking multiple warheads for counterforce purposes. Instead, an examination of the timelines for MIRV research in China suggest that the focus of the multiple warhead effort is anti-ballistic-missile defense. Lewis and Hua assert that the Chinese began to study MRVs and MIRVs in 1970 as a response to US deployment of multiple warhead systems, but lowered the priority of the effort in March 1980 after more than a decade of problems.¹⁴⁴ Work on multiple warheads was resumed on 10 November 1983, however, when the First Academy included them in the DF-5A modification program.¹⁴⁵ Some reports suggest that missile tests undertaken between fall 1986 and late 1987 were for the development of multiple-warhead missiles, including at least one such test for the DF-5A ICBM.¹⁴⁶

Why the renewed interest after years of difficulty? Lewis and Hua give us no clues, but the US announcement of the Strategic Defense Initiative in March 1983 seems too great a coincidence to ignore. If we assume that US SDI and now NMD research is driving the current round of Chinese efforts to develop multiple warheads, then a number of potential implications can be offered. The first critical variable is the status of Chinese nuclear testing. Despite allegations of nuclear espionage, Chinese accession to the CTBT would significantly impair China's ability to make progress in this area, particularly given the conclusion of the Jeremiah Commission that China has not deployed a MIRV on its ICBMs.¹⁴⁷ Even if we assume that the Chinese have already achieved a level of miniaturization necessary for MIRVing or will do so in the near future, a second critical variable will be the size of the future Chinese nuclear force posture, particularly the CONUS-capable forces. If China maintains a relatively small ICBM force, eventually replacing its several dozen DF-4s and DF-5As with a comparable number of DF-31s and DF-41s, respectively, then Chinese MIRVing along with robust decoys and countermeasures is likely meant to try and overwhelm the proposed 100- or 200-interceptor NMD system, not necessarily perform offensive counterforce attacks. A larger force of ICBMs makes this distinction murkier, but the overwhelming, triadic force asymmetry of the United States vis-à-vis China for the foreseeable future severely reduces the possibility that China could hope to achieve its goals with a preemptive strike.

Conclusions

Based on theoretical analysis, a review of Chinese nuclear principles and doctrine, and a study of China's nuclear force structure, we reach a number of important findings. We conclude that the operational survivability of China's nuclear retaliatory capability vis-à-vis major nuclear powers was and probably still is open to question, particularly in the context of an all-out preemptive strike. At best, then, China's minimal deterrent was primarily psychological, although the potency of this aspect of the deterrent should not be underestimated. The PRC's missile modernization program, therefore, has been a quest to increase the credibility of this deterrence posture by improving the readiness and survivability of the force. Measures being implemented are a transition from volatile liquid fuels to more stable solid fuels, a transition from fixed basing to mobile basing, and the construction of a robust C4I infrastructure. The Chinese have not operationally deployed any of their planned solid-fueled, road-mobile ICBMs, though the shorter range DF-31 seems to be nearing IOC after more than 30 years of work. When these systems come online, the Chinese finally will have succeeded in fielding a much more credible minimal deterrent force, whose mobility and readiness theoretically increase the chances that some percentage of the force could survive a first strike and thus effectively deter potential attackers.

At the same time, however, the Chinese force has grown to encompass more than simply minimal deterrent forces, including theater and tactical systems. Viewed in its totality, the Chinese nuclear force structure seems to defy simple categorization as either minimal or "limited" deterrence. The multifaceted force is made up of strategic, theater, and tactical systems of varying range, accuracy, and yield, reflecting the very different missions it is required to perform. The small ICBM force, anchored by the DF-5 family of missiles, appear to be second-strike minimal deterrence forces. The theater systems, by contrast, are unlikely to be used in a second-strike, minimal deterrent role following a preemptive strike. Instead, theater systems look like offensive systems meant to strike US forces and bases in Asia to degrade conventional capability. The short-range, ballistic missile forces, which are also nuclear capable, further confuse the situation by serving a variety of conventional warfighting and nuclear warfighting roles. For the future, the doctrine and force structure of China's Second Artillery must be analyzed at three distinct levels: a posture of credible minimal deterrence with regard to the continental United States and Russia; a more offensive-oriented posture of "limited deterrence" with regard to China's theater nuclear forces; and an offensively configured, preemptive, counterforce warfighting posture of "active defense" or "offensive defense" for the Second Artillery's conventional missile forces.

How did the Chinese force evolve into this arrangement? First, our analysis tends to confirm the arguments of Lewis, et al., of the importance of technology as a determinant of Chinese doctrine. The progression of missile systems, with their gradually expanding ranges and capabilities, defined the limits of the possible for the Chinese leadership. We disagree, however, that technology alone determined the nature of the Chinese nuclear force posture. Central guidance on ranges and payloads, although admittedly vague, appears to conform with strategic-level perceptions of threats and goals in the external security environment, especially when matched with the corresponding logical deployment pattern outlined in section three. Perhaps we can say that the Chinese made a virtue out of necessity in the construction of their nuclear deterrent, accepting the technological constraints of the system and making rational choices under those constraints.

In the end, however, we question whether China ever actually achieved a fully credible minimal deterrent. Thus, our attention has focused on the discontinuity between reality and aspiration, which is oftimes referred to as the "capabilities-doctrine gap." At the present stage in the Second Artillery's modernization, China is nearing an historic convergence between doctrine and capability, allowing it to increasingly achieve a degree of credible minimal deterrence vis-à-vis the continental United States--a convergence of its doctrine and capability it has not confidently possessed since the weaponization of China's nuclear program in the mid-1960s.

But what about "limited deterrence"? Recent studies find that since at least the late 1980s, Chinese military writings have promoted the need for China to develop a "limited deterrence"--as opposed to a "minimal deterrence"--doctrine. Although these writings are not considered official declarations of doctrine, because they are written by military analysts and appear in officially sanctioned military publications they have a special salience that deserves further scrutiny. In analyzing these writings, Johnston observes the emergence of "more comprehensive and consistent doctrinal arguments in favor of developing a limited flexible response capability" and that "Chinese strategists have developed a concept of limited deterrence . . . to describe the kind of deterrent China ought to have."¹⁴⁸

In general and specific terms, these Chinese writings call for limited, counterforce, war-fighting capabilities "to deter conventional, theater, and strategic nuclear war, and to control and suppress escalation during a nuclear war."¹⁴⁹ According to Chinese analysts, such a posture requires:

a greater number of smaller, more accurate, survivable, and penetrable ICBMs; SLBMs as countervalue retaliatory forces; tactical and theater nuclear weapons to hit battlefield and theater military targets and to suppress escalation; ballistic missile defense to improve the survivability of the limited deterrent; space-based early warningand command and control systems; and anti-satellite weapons (ASATs) to hit enemy military satellites.¹⁵⁰

Because such a posture would require a significant increase in Chinese capabilities, Johnston correctly highlights the gap between this proposed doctrine on the one hand, and actual capabilities on the other. As Godwin points out, the lack of any space-based reconnaissance or early warning systems means that Beijing's command and control system does not have the ability in real time to determine the size and origin of the attack, making it difficult to determine what kind of response is required--an essential component of the more sophisticated versions of limited deterrence found in Chinese military journals.¹⁵¹ Johnston also notes that actually achieving such a deterrent posture is not an inevitable outcome, owing to a number of possible constraints.

We have little basis for questioning the findings of Johnston about internal military writings on nuclear deterrence, especially the striking lack of discussion of the term "minimal deterrence." There are a number of possible explanations. Paul Godwin suggests that Mao Zedong's death in 1976 and the implementation of Deng Xiaoping's military reforms in the late 1970s permitted China's military analysts to explore issues of doctrine and strategy "free from the stultifying requirement to verify everything they wrote with a literal interpretation of Mao's writings and statements."¹⁵² Second, Godwin points to the increased battlefield nuclear weapons threat on the Sino-Soviet border, which "raised the salience of strategic deterrence and nuclear warfighting to a level it had never before achieved," encouraging Chinese military analysts to read extensively in Western theories and journals.¹⁵³ Johnston himself offers some additional explanations in the last few pages of his International Security article.¹⁵⁴ Many of the PLA authors explicitly contrast limited and minimal deterrence, obviating the possibility that they have simply renamed the previous doctrine for bureaucratic purposes. The authors appear to be well placed to affect the operational doctrine of the Second Artillery, removing the possibility of a disjuncture between academic and military writings, as occurred between the writings of RAND strategists and the war-winning strategy of General LeMay at the Strategic Air Command. If limited deterrence is defined as flexible response, counterforce warfighting, then perhaps limited deterrence is the aspirational doctrine for a future Second Artillery, although the past production timelines of the missile industry should sober our expectations of its appearance anytime soon.

We would add three more caveats to interpret the emergence and meaning of an ostensible limited deterrence posture in China. First, assuming a continued adherence by China to its testing moratorium, and the possibility that it will ratify the CTBT in the future, we question the ability of China to confidently develop smaller, lighter, and more accurate nuclear warheads (including potential MRV and MIRV capability) consistent with the limited deterrent aspirations described by Chinese analysts in the late 1980s and early 1990s.

Second, the tripartite system we describe possibly is a confirmation of Johnston's conclusions about limited deterrence, and we have simply come to the same place from a different direction. Perhaps the Chinese, when they looked at the multifunctional force structure they created, felt that minimal deterrence no longer could encompass all of the various defensive and offensive, long-range and short-range systems in their arsenal. Borrowing from Confucius, they may have concluded that harmony could only be restored when the name of the thing matched the nature of thing, and the product of this zhengming was "limited deterrence."

Third, even if we accept limited deterrence as an overarching aspirational goal of this multifaceted system, however, we still reject the misinterpretation of Johnston's writings by some, such as the Cox Committee and others, to mean that the Chinese are unquestionably engaged in an aggressive modernization of their missile forces meant to enable counterforce warfighting. Indeed, as we have outlined in this paper, there are legitimate, alternative explanations for many of the hardware trends in China. Reforms in mobility, readiness, and C4I infrastructure are readily and more comprehensively explained as an attempt to increase survivability from foreign attack--simply the long-sought confidence of a credible deterrent, notwithstanding Chinese analytic differentiation between "limited" and "minimal" deterrence--and not necessarily to achieve a warfighting, war-winning strategy. Moreover, as long as the numbers of the force stay beneath a certain level, increases in accuracy and multiple warheads do not pose a threat to American and Russian overwhelming nuclear superiority. American strategic nuclear forces, we must remember, still number around 8,000 deployed on 575 ICBMs, 102 strategic bombers, and 17 SSBNs. Indeed, a single Trident SSBN, carries more missiles (24) than the entire Chinese ICBM inventory.

The troubling countertrend involves the introduction of theater and national missile defenses into the equation, dramatically complicating China's strategic environment. Whereas China previously faced a world marked by the threat of offense racing, the post-BMD world will be marked by the unpredictable interactions of offense racing, defense racing, and countermeasure/decoy racing. In this environment, China would be acting rationally if it accelerated the desultory pace of its missile modernization, spending more money on relatively cheap countermeasures and decoys. To develop smaller warheads for penetrating missile defenses, Beijing would be acting in its self-interest by opting out of CTBT and resuming testing. Finally, China might even seek to foil missile defenses by proliferating its countermeasures technology to other emerging nuclear states. All of these trends would reduce the security of the United States. We hope that a sober understanding of the nature and purpose of Chinese nuclear force modernization and doctrinal evolution could forestall such an outcome.

---

Eric Croddy

Summary

This paper divides the two disciplines of chemical and biological (CB) weaponry. First, it discusses the PRC view of chemical weapons from a historical perspective. Next, the immediate question of Chinese CB weapons is examined, presenting the likely capabilities of a former or existent offensive capability in either area, followed by a look at Chinese CW defensive preparations. The next section sketches the development of China's chemical industry, and how its uneven progress could have affected offensive CW capabilities. Looking at the state of chemical technology in the PRC can help to establish a framework to consider the production of CW agents.

The BW side of the ledger follows, noting its historical context, facilities in the PRC that are related to the science of biological weaponry, and whether recent allegations of specific BW activity on the part of China have merit.

The main points of this study are as follows.

History

• The same sources charge that the US military used chemical weapons against Sino-Korean forces, including mustard, cyanide, and chloropicrin.

• The PRC also alleges the extensive use of BZ (an incapacitating agent) by the United States in the Vietnam war.

Chemical Warfare: China's Offensive Capabilit

• In Chinese literature, three CW agents receive the most attention, and probably for good reason: blister, blood, and nerve agents.

• China possessed a significant quantity of chemical weapons at least until the late 1980s, although the amount of CW agent or number of munitions did not approach anywhere near that of the former Soviet Union or the United States.

• The only chemical delivery systems in China that could threaten Taiwan directly include ballistic missiles and possibly aerial munitions.

Chinese Views on Chemical Weapons and Arms Control

• The PRC is under the impression that coalition forces moved some 2,700 tons of weaponized CW agent near the Persian Gulf during the Gulf war (1991).

• With regard to the Chemical Weapons Convention (CWC), the PRC probably believes that for a country to clandestinely produce large amounts of chemical weapons and not be discovered is impossible.

The PLA's Chemical Defense Corps (Fanghuabing)

• The PLA's Chemical Defense Corps (CDC) to our knowledge first undertook offensive operations during the campaign in the Yijiangshan islands in January 1955, probably involving obscurant smoke and perhaps flame throwers.

• China was able to indigenously mass produce CW defense equipment only by the mid-1970s.

• A nuclear, biological, and chemical (NBC) defense reconnaissance vehicle recently was modified by the PLA using a chassis from the Beijing-Jeep line of SUVs.

• After 1979, a new series of CW defense materiel was designed, and, by 1987, a total of 50 different standardized models were used by the PLA.

Medical Defense Research and Organization

• During the 1960s and 1970s, China provided instruction in chemical defense medicine to students from Vietnam, North Korea, and Albania.

• The official history of military medicine in the PRC indicates China finally deduced the chemical formula and composition of VX only by the 1970s.

• The two carbamates mentioned in Chinese literature for nerve agent prophylaxis, Cuixingning and Cuixingan, offer the PLA effective nerve agent prophylaxis, possibly superior to that used in the West.

• One of the more important areas for medical defenses are efforts to protect PLA personnel from the toxic propellants and off-gases of rockets and other self-propelled weapons systems.

Development of China's Chemical Industry: 1978 to Present

• China's large oil reserves and petrochemical industry probably were adequate to manufacture blister (Lewisite, sulfur, and nitrogen mustards) in large quantities, perhaps by the mid-1950s.
• Since the founding of the PRC, production of elemental phosphorus for fine chemicals probably was a very difficult procedure for Chinese chemists to accomplish.
• If China has in fact given up an offensive CW capability, the PRC does so now when it is most able to produce a wide range of toxic nerve agents, and in large quantities.
• A pessimistic view is that, in the event of a major crisis, the PRC would have little trouble reconstituting a large chemical weapons arsenal within a relatively short period of time.

Chinese Perspectives on BW

• The PRC repeatedly makes assurances that China has no biological weapons, and categorically states that "China has never manufactured nor possessed biological weapons."¹⁵⁵

• Some specialized equipment has also been fielded in some unspecified numbers to counter the threat of BW to PLA troops, including mobile laboratory units and bioaerosol sampling.

• By 1984, M.S. degrees were being awarded in the related specialization of BW defense by the Military Medical Science University.

• Nonetheless, Chinese writings on BW reflect a rather outdated mode of thinking, with emphasis on destroying insects and vermin for defense against biological weapons.

Chemical weapons: The Chinese Historical View

In language remarkably similar to that of an East German source on the subject,¹⁵⁶ modern Chinese CW experts refer to the use of noxious chemicals by prehistoric man, who may have employed them either to scare off fierce beasts, or perhaps to smoke out prey ensconced in caves. Drawing upon the fecund, literary sources of their own history, the Chinese are also wont to proffer specific examples:

During China's ancient period, it is said that the rebel Chi You created a fog to confuse his southern enemies. This smoke caused such havoc that were it not for Emperor Huang Di's "directional chariot"--a legendary vehicle that could navigate in darkness--the Northern barbarians might very well have won that day. In 559 BC, the Qin kingdom is purported to have poisoned the Jing river, a source that supplied water for the Jin, Lu, and other warring states, with the result that many men and horses were poisoned, forcing their retreat.¹⁵⁷

In 1000 AD, a grenade invented in China is mentioned, consisting of arsenic and crotonaldehyde (badou), ¹⁵⁸ capable of poisoning the enemy by means of its issuing vapors.¹⁵⁹ Even the deified Gen. Guang Gong¹⁶⁰ who, while attacking the city of Fan, was hit by a poisoned arrow in the right shoulder, the toxin going straight to his marrow. Fortuitously, he was cured by a well-known physician who happened to be in the area.¹⁶¹

The Modern Period

Chinese writings on the subject of CW--admittedly a sparse selection--closely mirror western sources, but little time is actually spent on presenting other historical precedents in use of chemicals in battle, at least not until World War I.¹⁶² From the latter conflict, according to a PRC book on military history, major lessons can be drawn, particularly from the first major chemical attack at Ypres. One contributing author explains that the inattention of the British concerning intelligence that pointed to Germany's plans to attack with chlorine was a crucial misstep. After all, he points out, Germany had already tried a similar assault on Czarist troops earlier, and this should have been known to the British War Ministry.¹⁶³

Although mentioning that White Russian armies used British CW ordnance against Lenin's troops during the civil war in 1919,¹⁶⁴ Chinese sources do not discuss CW activity that existed during 'feudal' Republican China by the various warlords. Why not is difficult to ascertain.

During the 1920s, Zhao Hengti, Cao Kun, Feng Yuxiang, and Zhang Zuolin all expressed interest in purchasing or enlisting foreign firms to help manufacture chemical weapons. The latter warlord apparently contracted a facility to be built in Shenyang by Witte (Germany), contracting Russian and German chemical engineers to oversee the manufacture of mustard, phosgene, and chlorine, while Zhao took delivery of a relatively small shipment of "gas-producing shells" in August 1921. The warlord Wu Peifu considered such forms of warfare "inhumane,"¹⁶⁵ but by all accounts no widespread use of CW occurred during this period.

This (deliberate?) omission in China's semi-official history of CW might shed light on later, post-1949 activities in chemical agent manufacture. Reliable sources indicate that, among the former Japanese chemical weapons being unearthed in modern China are found some munitions that are not Japanese, but could have been a legacy of local CW activity two decades before the war.¹⁶⁶ Also, they could have been more modern munitions produced in the PRC, and dumped out of expediency.

Lessons From World War II

As one might expect, the Chinese are bitterly indignant over Japan's use of CW in World War II. One source notes that, despite Roosevelt's warning to Japan in 1942 over their use of such weapons against the Chinese, the United States never did take measures to retaliate in kind.¹⁶⁷ Although Japan certainly did employ a significant amount of CW agents during their invasion of China--including Lewisite, mustard, cyanide, phosgene, and probably a range of irritating gases--the same Chinese source probably exaggerates the overall importance of such warfare in Japan's success against KMT armies during this period. Quoting an "authoritative Soviet source," the self-same book claims:

During its war in China, the Japanese army had prepared 25% of their artillery shells to be chemical munitions, while 30% of its aerial ordnance were chemical bombs.¹⁶⁸

The authors, waxing in a nationalistic tone usually reserved for such historical judgments, also write, "The Chinese people finally gained victory on the battlefield, proving that the Chinese race are exceptional (youxiu), courageous, and cannot be broken down or subjugated." "Fascist" Japan used CW over 2,000 times, causing the death of 90,000, the authors continue, but "it is not a couple of new weapons here or there, but rather a just people (zhengyi de renmin) that will win a war, despite the great menace posed by chemical weapons."¹⁶⁹

The PLA's more objective view of the European theater in the Second World War may be somewhat revealing, although it is clearly derivative of at least two Western sources, the SIPRI volume and Brown's Chemical Warfare: A Study in Restraints.¹⁷⁰

Observing that CW did not figure into Heinz Guderian's doctrine of Blitzkrieg, Chinese authors recount that Hitler was persuaded not to use chemical weapons against the Allies in World War II, despite having a "monopoly on tabun (GA, nerve agent)." Hitler's advisors, using the open scientific literature as a means of intelligence gathering, were certain that the Allies, particularly Russia which had developed organophosphate chemistry for many years already, must have superior CW capabilities and no doubt maintained nerve agent stocks. (In fact, the G-series nerve agents were unknown to the Allies until at least 1945.) Figuring this into the potential costs of an Allied retaliatory attack, Germany's impressive array of offensive chemical weapons--including the exceptional power of tabun--became little more useful than "room decoration."¹⁷¹

The Korean War

In an otherwise objective source on chemical weapons, the Chinese charge that the US military used chemical weapons against Sino-Korean forces on more than 200 occasions, and lists the following CW agents by name: mustard, cyanide, chloropicrin, and chloroacetophenone (CN).

The authors, Wang Qiang, a captain and now research professor specializing in chemical defense, and Yang Qingzhen, a senior colonel and assistant professor at the National Defense College, write extensively concerning the United States and the Korean war, claiming that chemical weapons were used often by the US Army. Because one of the purported incidents is recounted in a nationalist film (available on VCD, incidentally), and is one of the more popular films at least among the patriotic mainland Chinese,¹⁷² it may be worth quoting at length:

In line with the summer and autumn offensives, the US Army made incessant use of poison gas against the Sino-Korean armies. From June to December 1951, poison was used seven times against our PVA 19 Corps alone. On the fourth of October, while defending a 331.8 meter elevation line and under unremitting attack from the US Army the 141 division of the PVA 47 Army was attacked over 20 times with yellow, purple, and brown-colored poison agents fired from rocket artillery shells. Following the explosions there issued forth heavy sulfur-smelling pall of smoke; those being poisoned had difficulty in breathing, tearing from the eyes, and went into an irreversible coma. According to this it can probably be determined that it was a mixture of two chemical agents, chloroacetophenone (ben lu yi tong, benzyl chloro ethyl ketone, CN) and chloropicrin (lu hua ku). On the 13th of October, during an attack by the 8th Army of South Korea there were also fired "chemical agent artillery shells" against the 199th division of 67 corps of the PVA.

The US Army's use of chemical weapons was an often used technique that was particularly effective against our army's tunnel defense system. Chemical munitions were usually combined with the use of explosives, brought in by artillery and military aircraft. First by destroying with explosives those fortifications and chemical warfare defenses, chemical munitions would then be fired, raising the effectiveness in causing casualties. Sometimes chemical bombs and smoke munitions would be used in tandem, disorienting our troops and widening further the killing zone.

At the point of a particular offensive in the war, after capturing one area and coordinated with a surrounding siege, the American Army hurled chemical hand grenades in a tunnel with our defending PVA army inside. At about the middle of June in 1952, the United States puppet army was in Kaesong [jin cheng] attacking the 100th regiment of the 12th PVA to the east, defending [Guan dai li xi shan] and the 39th PVA group at a 190.8 meter elevation southeast of [cheng shan]. As our defenders were retreating to the tunnels, the US Army hurled several times hand grenades that utilized sneezing powder [pentixing duji]. In October, the US Army during its so-dubbed "Operation Showdown" attack on Kumhwa against the 45th division of the 15th PVA [ganling] at 597.9 elevation, tunnel no. 2, and 537.7 elevation at the Paeksan summit tunnel, there also were thrown chemical weapon hand grenades many times. The motion picture "On [ganling] Ridge" in one vivid scene accurately recreates the use of chemical weapons by the enemy against the PVA, reminding us that the victory in Korea was not going to be easy.

(Curiously, a mainland book series on the history of major battles since the PRC's founding makes little mention of this, at least not in the section devoted to the Korean conflict.¹⁷³)

The PRC generally has taught its citizens, among other things, that the United States and its "puppet" ally in the South instigated the Korean war by invading the north. (This idea is given serious thought among some Western revisionist historians as well.) Allegations of CW use by the United States also could be accepted matter-of-factly in mainland China, despite no foundation for such charges.

The Vietnam War

The PRC also alleges the extensive use of BZ (an incapacitating agent) by the United States in the Vietnam war, to have been delivered using M44 and M43 CW agent munitions.¹⁷⁴ In one of these supposed attacks, a whole platoon of NVA apparently became anesthetized and were subsequently wiped out by bayonets. One NVA soldier, however, was undiscovered and after waking up after three days reported back to his barracks.¹⁷⁵ Of interest here is the fact that the PRC clearly takes credit for having, along with other unnamed countries, given the North Vietnamese training in CW defense, as well as supplying protective gear and equipment during the conflict against US forces.¹⁷⁶

CW Offense

China possessed a significant quantity of CW agents, and this would include chemical weapon delivery systems, at least until the late 1980s. The amount of CW agent or number of munitions, however, probably did not approach anywhere near that of the former Soviet Union (40,000 tons, according to the general consensus).

Former Soviet chemical munitions could have constituted an early Chinese inventory, perhaps before 1960.¹⁷⁷ If so, these were probably first- and second-generation CW agents only, such as phosgene, mustard, and Lewisite. Although certainly potent by themselves, this chemical ordnance probably was not augmented by the modern nerve agents, at least not for some time. The weaponization en masse of the G-series nerve compounds did not proceed quickly in the Soviet Union, despite the Soviets having discovered German tabun and sarin manufacturing facilities in 1945. Krause and Mallory write that in the former Soviet Union,

It is safe to assume that during the 1950s there was small-scale production of nerve agents such as soman and sarin and that testing and development activities took place in order to familiarize Soviet military officers with the effects of these new agents. . . . Once again, the Soviet military's greatest problem was its technological backwardness in the field of military chemistry. There is evidence pointing to some "development aid" rendered to the Soviet Armed Forces by East German military chemists. However, it seems that these activities did not start before 1965 or 1966.¹⁷⁸

With Chinese-Soviet relations ever worsening in the 1960s, the same could probably be said of chemical weapons work in the PRC.

Current Status of Chemical Weapons in the PRC

The PRC in submitting its data declaration to the CWC reported that it destroyed three production facilities, capable of producing militarily significant amounts of CW agent (from low hundreds to thousands of tons). This claim is consistent with PRC statements that deny any previous production of biological weapons, but make no categorical claim regarding past work in CW agents or weapons. The aforementioned declarations, according to some who were in close proximity to the offices that handled such documents, recorded Chinese CW agent-related activities in voluminous detail. Possible Chinese chemical munitions could have followed in the Soviet model:

Literature on Chinese CW in its offensive context is practically nonexistent. One of the only credible hints surfaced in 1989, when a marketing manager of Duphar medical devices was told by mainland Chinese that, in addition to having nerve agent antidotes, the PRC possessed much more than they were letting on: "But we don't know what, and we can't verify the claim at all," the manager was reported as saying.¹⁷⁹ Another source indicates that a Chinese-manufactured mustard shell of unknown caliber was recently found among Albanian munitions, and although containing live agent it appeared to be intended primarily for training purposes.¹⁸⁰ Finally, a report in March 1997 alleged that Ukraine sold China 500 tons of sarin left from former Soviet stocks, in addition to chemical protection equipment.¹⁸¹ (The original report apparently began with a comment from a Taiwanese "intelligence officer."¹⁸²) This story was vigorously denied by the Ukranian Ministry of Defense.

The training and research in handling the effects of chemical weapons is routine in the PRC, but to date no defector or other report in the open literature has elucidated any detail on actual Chinese chemical munitions or offensive doctrine in CW. Later in this report, the role of Chinese medical sciences in CW defense will be treated in some detail, but for now it is sufficient to point out that the PRC is cognizant of all known CW agents, except perhaps novel agents such as the Russian novichok,¹⁸³ has developed a nominal defensive infrastructure to deal with these threats, and is quite knowledgeable from both indigenous research and second-hand (foreign) information.

Chinese literature regarding chemical and biological warfare, often draws directly from Western sources, and one can even pinpoint certain passages that were translated practically word for word, such as the SIPRI volumes on CBW by Robinson, Leitenberg, et al. Therefore, when an officer of the PLA suggests that multiple launched rocket systems (MLRS) offer the most efficient means of delivering CW agent, he is not necessarily speaking from experience or drawing from any doctrinal axiom. He is just as likely quoting directly from the aforementioned SIPRI volume on chemical weapons.¹⁸⁴ However, he does point out that "at present, the United States and Russia both have this type of weapon system to fire CW agent rockets."¹⁸⁵ From the Chinese point of view, and considering their intimate knowledge of Soviet MLRS capabilities, the CW threat from the ersatz Soviet Union must have been an especially unsettling one.

And once the Soviet Union shrank back to pre-Revolution borders, and even now is cooperating militarily with the PRC, the Chinese apparently have little incentive to maintain an offensive CW capability. The remaining land-based opponent, India, could pose a threat to China, but would this justify holding on to a form of warfare that does not coincide with the new revolution in Chinese military affairs? We cannot say for certain, but it does not seem likely.

As for Taiwan, the only delivery systems remaining would be ballistic missiles and possibly aerial munitions--a Chinese concept for a binary nerve bomb that could be dropped over Taiwan will be addressed shortly. With China already armed with nuclear warheads, however, offensive chemical weaponry utilized against Taiwan seems redundant, possibly anathema, particularly when considering their shared past and kinship ties. At least two PLA officers regarded the use of chemical weapons to be equivalent to employing nuclear war:

Chemical weapons could be the fuse to ignite a nuclear war, for as soon as mass casualty weapons such as CW are used, there is no reason why nuclear weapons won't be as well. Once CW begins, it will be just like releasing the evil spirits from Pandora's box, eventually slipping towards the abyss of nuclear war.¹⁸⁶

When it comes to the actual chemical weapons themselves, we can identify some of the impressions of the PLA, however. In some respects these are surprising to an American observer

• Three main agents receive the most attention, and probably for good reason: blister, blood, and nerve agents. Blister agents, or vesicants, include mustard and Lewisite and are standard for CW arsenals. Sulfur mustard in particular requires a low level of technology investment compared to nerve agents, is well suited for a country well endowed with petroleum, and has a proven track record of effectiveness in battle. Nerve agents, as explained in the chemical industry section below, would have presented a challenge to early PRC technological capabilities, but this situation has changed dramatically in the past two decades.

• Chinese CW experts on CW defense center upon the Soviet-style method of contaminating the ground with chemical agents of various kinds.

A 1985 CBW defense encyclopedia reported that "the Russian military has been equipped with thickened mustard gas for many years now, and recently it has also come to possess thickened soman."¹⁸⁷ As a means to counter such threats, viscous preparations of nerve agents--the Chinese cite methacrylate polymers and tributyl phosphate¹⁸⁸ as possible thickeners for CW agents such as soman as well as mustard--would have given China the full range of persistent application of chemicals needed to slow down a Soviet armored advance. Although it may very well be that China was not able to mass produce VX until the 1970s (see medical research below), the utilizing tributyl phosphate--a compound that is easily produced¹⁸⁹--as a thickener would have afforded sufficient viscous character to other Chinese nerve agent preparations.

• Binary chemical weapons are given special attention, indicating their compatibility with low-cost production of weapons of mass destruction (WMD).

China and the Chemical Weapons Convention

Since ratifying the Chemical Weapons Convention (the treaty coming into force in 1997), and having submitted its declarations to the executive body responsible for verification, the PRC ostensibly has no chemical weapons. Also an official from the Chinese Ministry of Foreign Affairs repeated this claim to me.²⁰⁰

China is displeased because it perceives lack of benefit from joining the CWC. An Iranian official stated that "China sees a lot of liability with little benefit in being a State Party." Although this was said in the context of inspections, the PRC clearly is disappointed that it is not obtaining the technology, assistance, and bonuses it anticipated gaining from joining the Convention. Perhaps China expected more after having destroyed its remaining chemical arsenal, although Beijing's stockpile probably was quite small.

Chinese Views on Chemical Weapons and Arms Control

The handbook on chemical weaponry written by two PLA officers is skeptical that arms inspections can stop the proliferation of chemical weapons technology. These authors state that a fundamental concern is that the basic components involved in manufacturing binary chemical munitions are not far removed from technology used in industry. No matter how many intrusive inspections are carried out, they cannot stop the basic research conducted by civilians, thus making the spread of such CW technology easy.²⁰² As chemical weapons proliferate, the possibility of their being used increases when a nation, equipped with a CW apparatus, is pitted against another country that has none.

China's View of the Gulf War (1990-91)
The PRC seems to be under the impression that, in addition to 1,000 tactical nuclear warheads deployed by the United States,²⁰³ coalition forces also moved some 2,700 tons of weaponized CW agent near the Persian Gulf, posing a "name-brand recognition" type of threat to Saddam Hussein. The latter claim, of course, cannot be supported by the available evidence,²⁰⁴ but the PRC believes this deployment of chemical weapons played an important role in the course of the war, demonstrating that CW is "by no means inferior" to high-tech weaponry.²⁰⁵

Referring to the export control efforts of CW precursors and equipment by Australia Group, the authors above suggest that:

Although many countries have adopted these measures, it is doubtful that they will be effective. Because companies want to earn high profit margins, they are not going to concern themselves with governmental prohibitions, and will secretly export these kinds of materials. The reality is that nations are helped by foreign business, supplying them with the materials, equipment, and technology to acquire a chemical weapons capability.

A solution, they suggest, is to establish chemical-weapons-free zones.²⁰⁶

Reiterating the futility of stopping the proliferation of CW-related technical know-how, the authors nonetheless concede that, if international agreements like the CWC use intrusive inspections (yange de hecha cuoshi), a country will have great difficulty--perhaps near impossibility--clandestinely producing large amounts of chemical weapons without being discovered.²⁰⁷

The PLA's Chemical Defense Corps (Fanghuabing)

History and Defensive Materiel
The 8^th Route Army in 1939 established a Chemical Group (huaxuedui) at the Chinese People's anti-Japanese Military College. The group received rudimentary instruction, probably from Soviet instructors, on measures to defend against Japanese chemical warfare.²⁰⁸

According to Maj. Gen. Jiang Zhizeng--chief of the Chemical Defense Department of the PLA in 1989 and a significant contributor²⁰⁹ to an encyclopedic treatment of NBC defense--during the period following the war for "liberation" separate chemical groups were established in the 7th, 9th and 13th columns of the PLA's East China field army. "According to the recollections of comrade Liu Baicheng," writes General Jiang, "the 2nd Field Army established a large Chemical Group." On 11 December 1950, following the personal approval of Chairman Mao and Premier Zhou Enlai,²¹⁰ the first Chemical Defense Corps school was founded, leading to the formation of the Chemical Defense Corps (CDC). Its earliest instructors at this point were former Nationalist officers who had prior training in CW defense, and who had apparently "revolted from the KMT."²¹¹ In 1951 an Oxford-educated chemist, Dr. Huang Xinmin, left England to direct the chemistry department in the PRC, along with several Soviet advisers, in "protection" against CW agents.²¹²

With regard to the allegations that chemical weapons were used during the Korean war (see also below), the aforementioned preparation in CW defense prior to China's involvement could have made the PLA overly amenable to suggestion. When aggressive use of artillery, napalm, and aerial bombing hit the Chinese People's Volunteer Army during the war, the resultant off gases and suffocating smokes no doubt had the semblance of real chemical weaponry. These factors, together with posturing and outright fabrication for propaganda purposes (which are similar to charges that the United States used BW), are the best explanation of why the PLA continues to assert that the United States used CW (for more on the Korean war, see "History" below).

The wholesale import of Soviet-made CW defensive gear, including detectors, clothing, decontamination equipment, smoke generators, and flame throwers began in 1953.²¹³ The latter two types of equipment would be deployed as early as 1955, according to General Jiang.

In December 1954, Zhang Aiping was ordered by Mao Zedong to prepare an assault on Yijiangshan island.²¹⁴ Full-scale military operations were conducted during 18-20 January 1955, to seize control of the Yijiangshan island off of the coast of Zhejiang Province. General Zhang noted that "this was the first organized operation in which sea, air, and land forces worked in concert," and quickly finished off remnants of KMT soldiers (Jiang's Bandits) in this rather lopsided affair.²¹⁵ These are the first known operations for the nascent PLA's Chemical Defense Corps.

A grainy photograph published in PLA Pictorial shows troops boarding landing craft, wearing protective suits of some kind, and many carrying portable tanks on their backs, consistent with portable smoke generators or flame throwers.²¹⁶ Upon the quick victory by the PLA, a congratulatory telegram to the front lines was sent by the commander in chief of the East China army (presumably from Zhang Aiping).²¹⁷

Bingqi Zhishi has it that on 19 April 1955, the Central Military Commission named Zhang Xigeng as the first minister of the Ministry of Chemical Defense. General Jiang, however, uses the date January 1956 for its founding.²¹⁸ In any event, General Jiang states the mission of the CDC in this way:

[The Chemical Defense Corps] is to guarantee the protection of our army while under battle conditions that include nuclear, chemical and biological weapons. It is composed of troops in the CD (surveying, reconnaissance, decontamination), those responsible for flame throwers, smoke generation, etc. Among the major responsibilities are: Directing the use of collective protection against chemical weapons, carrying out of survey and reconnaissance for nuclear radiation and chemical analysis, testing for agents and infection, the neutralization of poisons and infection, providing an organized and assured obscurant smoke, as well as directing the coordinated use of flame throwers with advancing troops in combat.²¹⁹

A Chinese NBC defense manual dating from 1957 grouped CW agents in four categories, the systemic poisons, asphyxiating gases, blister agents, and irritants. Probably in the interest of simplicity for its intended--chiefly juvenile--audience, nerve agents were not mentioned by name. They were listed along with cyanide under the heading of systemic or blood agents, referring to them as being "odorless and colorless liquids, very poisonous, not very easy to detect, demanding that special caution be taken." It also gave simple instructions on how to build shelters, don protective CW suits, and decontaminate one's skin.²²⁰

By 1959, of the 20-odd different types of CW defense materiel formerly imported from the Soviet Union, the PRC became 90 percent self-sufficient in their development and manufacture. For example, in the mid-1960s, the Model 64 respirator mask was an indigenous product.²²¹

In 1971 the Chinese developed a detector (type 65) alarm for organophosphorus (OP) compounds (i.e., nerve agents). Although the type 65 OP detector is praised for its easy use and sensitivity to detect nerve agents at a considerable distance, like Western counterparts it is prone to interference²²² (and probably susceptible to false alarms). Nonetheless, the 65 and type 75 testing kit (analogous to the M256A1) are the current CW agent detection accouterments used by the PLA.²²³

But the ability to indigenously mass produce CW defense equipment, at least enough to outfit significant numbers of personnel, was achieved only by the mid-1970s, when the imported and copied Soviet-style equipment finally began fading out of service. In 1975, mechanized chemical and radiological surveying became more specialized, and CW defensive gear became standardized for the battlefield. A CW defense reconnaissance vehicle was modified using a chassis from the Beijing-Jeep line of SUVs, the same outfitter for the Gonganbu (Public Security Bureau), among others. This vehicle represented the first generation of laboratory testing on wheels. Although personnel must get in and out of the vehicle to perform field recon, the PLA could see many improvements in automation.²²⁴

Combining appropriate gas masks, individual chemical testing kits, and CW agent alarms, the Chinese Navy, Air force, and Second Artillery were already equipped at this time with CW defense equipment. After 1979, a new series of CW defense materiel was designed, and by 1987 a total of some 50 different standardized models were used by the PLA.²²⁵

Food and Water Testing
Various testing methods were supplied in kit form to examine provisions for CW agent contamination, beginning with the types 59 and 62 testing kits that were put together in the 1950s. Later, when the toxins VX and BZ came to light, improvements were made in the new type 67 kit supplied to the PLA. Later in the 1980s, a more comprehensive list of CW agents could be tested for in food and water by the type 85, and a kit especially designed for water quality testing was developed in the Shenyang type 81, which is well-suited for use by mobile armed forces.²²⁶

Gas Masks: Measure Twice, Cut Once
The Chinese military depended upon used and foreign-made gas masks going into the Korean war. The PLA notes that in the beginning it encountered an immediate problem, namely, the gas masks did not appear to fit the Chinese face very well. The PLA worked hard to find a solution:

It was necessary to make it suitable for the shape of the head that typifies our nation's race. In 1958, data was culled from the measuring of some 40,000 PLA soldiers heads, resulting in a lightweight and very protective model 64 mask.²²⁷

Later, types 65 and 69 masks were made for more flexible use on the battlefield, the latter model having activated charcoal in its filter. An additional model 87 was introduced, along with one specifically crafted for rocket propellants, the model 75.²²⁸ No. 75 is considered a "special-purpose mask," having a filter/canister construction best suited for personnel who are stationed near rocket propellants and fumes. It is also designed for tank crews and use in aircraft by connecting directly with the oxygen system.²²⁹

Chemical Suits
The first-generation protective garments in the PLA were and still are the venerable, 1966-vintage butylene polymer rubber suit. Having strong resistance to acids, mustard, VX, etc., and weighing some 2.5 km,²³⁰ this suit must be terribly uncomfortable, especially in the many hot days of the year in southern China. The CDC seems to be using this suit for most of its specialized training and operations. For battle front troops, mercifully, a gas-permeable suit layered with activated charcoal has been made available since its introduction in 1982.

Decontamination Equipment/Vehicles
Having noticed the former Soviet TMC-65 turbine engine platform--basically a jet engine that uses the force of water to decontaminate vehicles--the Chinese seem to have adapted their own. Whereas the Soviet system did not necessarily require special decontamination fluid²³¹ or hypochlorite solutions--heat and kinetics of the spray are probably violent enough for the purpose of sustaining combat operations--the PLA includes a tank of decon fluid in its diagram. This "Jet Exhaust Decontamination Vehicle" vents with a flow rate of 400 meters/second, with vapor immediately out of the nozzle reaching temperatures of over 500¼C, and reducing to 200¼C upon reaching the intended surface. Onboard computerized control can adjust the rate of fuel (diesel) burn, heating, etc.

One of the more interesting aspects of this arrangement, which includes a wireless automated control and a secondary driver's booth, is that the same engine can be used for laying down smoke screens. Apparently, the secondary operator changes the intake to allow supply of smoke-generating fuel into the turbine.²³²

We do not know if this endeavor represents a serious effort to deploy such vehicles in large numbers in the PLA forces, or even within the smaller organization of the CDC. Nonetheless, Chinese authors on this topic are apologetic concerning this particular need:

Although today the world is gradually moving towards a peaceful trend, the Chemical Weapons Convention (CWC) has been signed, and despite the reduction in risk from chemical attack, there is still an unending research and improvement in decontamination equipment.²³³

Decontamination fluids used in the CDC are made up of the usual types found in other armies, the PLA often utilizes a 3:2 ratio of calcium hypochlorite and calcium hydroxide, in addition to bleach, and chlorides of ammonia are especially recommended for dealing with V-agent contamination.²³⁴

Individual Decontamination and First Aid Kit
In 1958 the PRC copied the Soviet-type "IPP-51" decontamination kit, and later developed types 58, 63, 71, and model 1-0 for cleaning skin exposed to CW agents. Other models, No. 14 and 25 in particular, were specifically formulated to handle V-agent threats.²³⁵

The modern kit contains pharmaceutical preparations for countering the effects of nerve agents, and a moistened, chemically impregnated cloth for decontaminating skin. A cylinder holds tablets, presumably for a carbamate, and beneath it is a spring-loaded atropine injector. Although the main purpose of this kit is, again, for nerve agent first aid, it is also recommended for decontaminating other agents, such as mustard and Lewisite.²³⁶

Chemical Warfare Defense and the Chinese Antichemical Corps

Medical Defense Research and Organization
The following are highlights from the official history of Chinese military medicine.

China's development of a professional cadre to treat CW casualties took form in 1951, when the first Military Medical Sciences Learning Hospital was founded. Expertise in medical defense against CW was initially brought to China from the Soviet Union, and the first semester of high-level training in this area began with 45 students in 1954. Early emphasis on mustard agent (the king of CW agents) soon gave way to even more serious attention on the nerve toxins, and Chinese staff of the General Hygiene Department (Zonghou Weshengbu) visited the Soviet Union for advanced studies. In 1958, the disciplines of military toxicology, pharmacology, and biochemistry were combined into a Pharmacological and Toxicology Research Institute headed by Yang Tenghan, also referred to as the Chemical Defense Medical Science Research Institute, (and later changed in 1987 to the Toxicology and Pharmacology research Institute). Nationwide conferences that dealt with military chemical defense were held in 1961, 1974, and 1979.

In the beginning of the 1960s, owing to "strategic demands," a Chemical Defense Testing Unit (Fanghua Jianyan Fendui) was formed, shortened to "Fangyandui," and later called the Chemical Defense Medical Science Specialized Unit (Fanghua Yixue Zhuanye Fendui). Its duties were to assist in evaluating conditions on the borders or within the country, to quickly ascertain threats and provide medical, testing, and other support for chemical defense medicine. Depending upon their anticipated requirements, each military region formed its own version of this type of organization.²³⁷

With gradual improvements in technology, and the accumulated scientific knowledge, chemical defense medicine became even more important in the 1960s and 1970s. Qualitative improvements were made in the general treatment of mustard casualties, ways to counteract incapacitating agents, treatment for poisoning from cyanide compounds, as well as prophylactic defense and antidotes for nerve agents. In addition to packets made for skin decontamination, a testing kit was also designed for alerting one to the presence of contaminated food and water provisions. By 1963, the realization that the irreversibility of enzyme by nerve agents due to aging also led to renewed efforts at finding better acetylcholinesterase reactivators. During the 1960s and 1970s, instruction in chemical defense medicine was provided to students from Vietnam, North Korea, and Albania, with specialists sent to help these and other countries establish their own testing laboratories.²³⁸

Along with investigation of therapeutic herbs to counteract the toxic effects of CW agents, in the 1980s enzyme immobilization indicator technology was developed, along with more advanced spectrographic, immunoassay, and ionization-based (fangshe) detection systems. During this time the PLA also sent abroad specialists to study the problem of treating the toxic effects of nerve agents, communicating with the experts in the field in the United States, Great Britain, France, Japan, Switzerland, Australia, among others. From 1971 to 1989, the General Logistics and Sanitation Departments cultivated 81 specialists in the area of chemical defense medicine. In 1989, 19 Masters degrees were awarded in hygiene and chemical defense medicine.

Chinese Research in Defense Against Nerve Agents
For protection against nerve agent exposure, three main types of compounds are commonly used, both before and after the fact.

Carbamates are reversible inhibitors of acetylcholinesterase (AChE), and protect the enzyme from irreversible (aging) or long-term impairment by the G-series and VX. Because soman, for example, can age enzymes in a matter of minutes, carbamate prophylaxis is especially important. Pyridostigmine bromide (PB) is used in the United States and most other countries, but physostigmine, a naturally occurring carbamate found in the calabar bean, also is effective, although it probably is of higher toxicity. (In my view, current speculation that PB has a role in Gulf War Syndrome is completely unfounded.)

Oximes are administered after nerve agent intoxication to remove the nerve agent from the enzyme, hopefully revitalizing enough AChE to put the victim on the road to recovery. Pralidoxime HCL (Protopam Chloride, 2-PAM-Cl) is used in the autoinjector supplied to the United States and NATO, but may be replaced in the future with more effective oximes.

Anticholinergic compounds are those that block acetylcholine, restoring some normalcy following nerve agent poisoning. Atropine is the drug of choice for military chemical defense, although other similar compounds could be used, depending on the level of perceived risk.

Chinese development of antidotes for nerve agents may be broken down into three stages: 1) the initial treatment regimen typed No. 11, consisting of atropine and an oxime, etc., 2) efforts to find treatment for soman poisoning, and 3) general efforts to raise the capabilities of treatment, made more pressing by reports of the V-series of agent revealed from foreign reports in the 1970s.

With regard to the V-series of nerve agents, open literature referring to the basic structure of VX goes back to at least 1958.²³⁹ Between the actual date of discovery (1952), and its open publication revealed in 1975,²⁴⁰ Soviet military intelligence (GRU) probably had filched the basic formula in 1955.²⁴¹ But regardless of how it actually obtained the information on V-agents, the Soviet Union then proceeded to replace much of its G agents with its own structural isomer, VR-55 by 1960.²⁴² Are we to believe that, at least according to the official history of the Chinese military medical sciences, the PLA pharmacology and toxicology department only became aware of the details of VX by the 1970s?

The chemical formula and composition of VX was finally deduced in China after considerable laboratory investigation. Finally, China subsequently introduced the type 85 emergency antidote for treating nerve agent intoxication, "bringing the PRC to international standards." This effort, along with work in carbamates for protecting acetylcholinesterase from the effects of nerve agents, led to the exhaustive research of more than 15,000 compounds, 2,000 of which were novel formulations. Of these, more than 10 were shown to be effective, some of them typed as No. 11, No. 60, No. 68, No. 51, No. 73, No. 85, and others for emergency treatment of nerve agent poisoning, No. 85 simultaneously winning a second-level national prize as well as an award for advancing military technology.

In terms of enzyme protection with carbamates, traditional medicine yielded Cuixingning and Cuixingan. As work on natural sources of carbamates began in 1968, in the realm of enzyme reactivation, Song Hungqiang, et al., synthesized two new types of oxime compounds in cooperation with the Beijing Medical Pharmacological Industry Research Institute. These two compounds showed effectiveness against soman poisoning, surpassing international standards. For blocking acetylcholine, Zhang Qijie and others finally was able to synthesize new compounds, among them one that is found in traditional medicine.²⁴³

The two carbamates referenced above, Cuixingning and Cuixingan, are worth mentioning for they offer the PLA effective nerve agent prophylaxis, possibly superior to that of pyridostigmine bromide used in Israel, the United States, and many other Western countries. Lieske, et al., credit Ahmed and Robinson with having first prepared the following compound, referred to in the Chinese literature most commonly as Cuixingning, but also "youselin," and "Jiebiling" in the 1997 Junshi Yixue Cidian.²⁴⁴

The Chinese claim, however, that the latter compound was first synthesized in China during the early 1960s.²⁴⁵ A study performed at the US Army Research Institute of Chemical Defense (USAMRICD) found that Cuixingning showed promise as an effective prophylactic for nerve agents, while also having an acceptable index of toxicity.

Another compound, Cuixingan, is also mentioned in Chinese writings on CW defense.

Jiebling
Cuixingan

Regarding this compound, the Chinese claim that:

The pharmaco-toxicological action of cui-xing-an is the same as that of physostigmine and cui-xing-ning, but its toxicity is only one-tenth that of cui-xing-ning. The nicotinic action of cui-xing-an and its effects on the cardiovascular system are milder that those of cui-xing-ning.²⁴⁶

Little information exists in Western literature, however, to support such claims.

In the field of anticholinergics, Chinese investigators undertook substantial research on herbs and other traditional forms of medicine. As early as 1959, the PRC copied from abroad an atropine autoinjector, developing later a partially automated, ampoule-style injector mechanism, built with an extruded plastic injector. A fully automated injector has been supplied since the early 1980s, and the PLA Veterinary College successfully developed an OP antidote syringe (jielinzhen) for animals. Additional work through the 1980s included the elucidation of the mechanism of nerve agents, including soman and VX, as well as work in the area of using hydrolyzing enzymes to protect against lethal doses of soman. Many clinical trials involving this knowledge were performed on the personnel, reflecting a "a spirit of selflessness" on the part of the researchers. Such tests in chemical defense medicine alone were performed in individuals on some 3,000 occasions.²⁴⁷

Blister Agents (Vesicants)
Recognizing the "Great Old Difficulty" in treating mustard agent casualties, Chinese chemical defense medicine is resigned--much like everyone else--to the fact that supportive care is at present the only realistic course of action. A complete report elucidating mustard's effects on the human body, including its comprehensive toxicity, was produced in the 1980s. The official history does, however, refer to a typed No. 14 ointment designed to absorb and decontaminate exposed skin to vesicants, and the Kunming Military District's 60th Hospital and Medical Research Institute treated five serious mustard casualties, four of them "successful" (presumably, the fifth did not survive.) These casualties may very well have been caused by Chinese mustard, as the history refers to "other regional hospital departments having treated victims of mustard from leftover Japanese munitions, gaining much experience in the process."

The PRC seems to have adopted its own mercaptan type of treatment for Lewisite exposure, a certain dithiosodium butyrate, different from British Anti Lewisite (BAL) and the Soviet variety. As a CW agent, however, Lewisite does not receive as much attention as does mustard. ²⁴⁸

Incapacitating Agents
When it was learned that the United States had developed BZ weapons in the early 1960s, the PRC undertook efforts to characterize the compound and to develop treatments for BZ intoxication. BZ (3-quinuclidynil benzilate) was prepared in China by 1965, the official history noting that the compound was finally revealed in the open literature in 1972. In the 1970s, the Military Medical Science University, the Jinan Military Region's 88th Hospital, as well as the PLA General Medical Institute successfully used "Jiebiling" [the aforementioned carbamate, Cuixingning] to counter the effects of BZ poisoning. (Treatment in the United States for anticholinergic poisoning, such as Jimson's weed, calls for the judicious administration of a carbamate (physostigmine), if only for diagnostic purposes.)

Reports also surfaced that work overseas was being done on what were termed "body incapacitants" (Qutixing shinengji) in the mid-1960s, substances that would cause personnel to become numb and paralyzed (mabi tanhuan). These might be references to fentanyl and its derivatives. The Chemical Defense Medicine Research Institute then speedily researched effective antidotes in the event such agents were to be encountered.

Blood Agents (Systemic Poisons)
In 1961 the Seventh Military Medical University Medical College Protection Teaching Research Institute, together with the Chemical Defense Medicine Research Institute of the Medical University College of Sciences, carried out investigations into treating cyanide casualties, both HCN and cyanogen chloride (CK). Novel treatments were researched between 1970 and 1980, including the use of 4-dimethyl amino-aniline [4-erjiajianjifen] and p-amino benzylacetone. In the 1980s an antidote kit was developed, typed the No. 85 anticyanide injector. The No. 85 is also used in industrial and shipboard settings where accidents involving cyanide might occur.²⁴⁹

Asphyxiants
Phosgene and diphosgene present similar problems to mustard in that very little exists in therapy other than supportive care. In the 1960s, the Fourth Military Medical University's Protection Teaching Institute laboratory emphasized research in the choking gases, discovering, among other things, that vitamin C lessened the severity of pulmonary edema. Some 70 clinical trials were carried out in investigative drug therapies, many showing promise. Treatment for phosgene and diphosgene gas, including rapid diagnosis, also was investigated thoroughly by the Fourth Military Medical University and the Lanzhou Military Region Medical University Research Institute.²⁵⁰

Protection Against Rocket Propellants and Off-Gases
One of the more important areas for medical defenses was undertaken simultaneously with China's push for strategic aerospace weapons, primarily to protect personnel from the toxic propellants and off-gases, and especially to provide expertise in protection from "harmful gases produced from underground nuclear explosions."²⁵¹

In the latter part of the 1950s, the upper echelons directed scientists to fully elucidate the dangers of unsymmetrical dimethylhydrazine (UDMH), and in 1960 the Military Medical Science Institute established a Rocket Propellant Toxicological Research Laboratory, with a staff of about 20. The institute focused attention on protecting eyes and skin from possible exposures, decontaminating tissue safely, and treating casualties.²⁵² Determining that UDMH could be used safely in aerospace vehicles was considered a major achievement, winning the institute a 1965 National Discovery award. Additional toxicological research institutes conducted studies on decaborane, another propellant, and in 1966 Zhu Kun, et al., found that Vitamin B-6 was helpful in treating UDMH exposures.

As the strategic missile program grew even larger in the 1970s, solid fuel propellants also demanded toxicological study, including those compounds used in missiles and torpedoes. Both the National Defense Science Council and the SAC each established medical defense groups (fangjiandui) and laboratories (fangjiansuo) dedicated to addressing the problems of toxic propellants.²⁵³

Development of China's Chemical Industry: 1978 to Present

The PRC's Chemical Industry Base
For a nation to manufacture chemical weapons, a sound chemical industry base is obviously advantageous, particularly for the unfettered supply of important precursors and intermediates. It becomes even more important when the production of militarily significant quantities (in the hundreds of tons) of CW agent fill are required. During World War I, first-generation CW agents (chlorine and phosgene) were originally seconded from German dye industry stocks: The gas attacks at Ypres in April 1915, for example, required roughly 500 tons of chemical agent,²⁵⁴ and represented half of Germany's supply of chlorine for that year.

At the same time, some countries can do much more with a lot less. Iraq, for example, is not among the most highly developed nations in terms of a comprehensive chemical industry. However, its large phosphate reserves and imported technology (e.g., a French phosphorus trichloride manufacturing plant) enabled it to produce, with the possible exception of Soman (GD),²⁵⁵ every known CW nerve agent and in large quantities (including an obscure chemical, cyclosarin or GF). In terms of its own chemical industry base, a similar picture could be drawn for China, with one major difference, namely, the lack of foreign technology and the withdrawal of Soviet assistance, especially during the years 1959-78.

Although today the PRC could be ranked seventh in the world in terms of total GDP, its level of technological competitiveness is still rated much lower; by one account China ranks 28th.²⁵⁶ No Chinese conglomerate appears in the top 50 chemical producers in 1998, for example. Taiwan's Formosa Plastics comes in at 46 (with Chevron following at 47).²⁵⁷ More recently, the nurturing of China's chemical industry has brought some rather spectacular results.

To understand better the environment in which the PRC stands in terms of CW weaponry, assessing its past and present levels of chemical technology is useful, particularly since the Sino-Soviet schism in 1959. This section sizes up the development of China's chemical industry, with emphasis on its course since the founding of the PRC.

Background on Chinese Chemistry: History
In 1928, an Institute of Chemistry was originally founded in Shanghai as part of Academica Sinica, was subsequently transferred to Kunming during the war, and eventually returned to Shanghai following hostilities. The Peking Academy also contained under its auspices an Institute of Chemistry that was formed a year after its Shanghai counterpart. In the late 1930s the Chinese Chemical Society had a membership of about 2,000, and by 1950 there were 218 Chinese research institutes devoted to chemistry.²⁵⁸

Not surprisingly, the Soviet Union played a very important role in the formation of scientific societies in Communist China, and in early 1956 a 12-year plan was instituted that prioritized technological research in the following order by the CCP leadership:

Peaceful utilization of nuclear energy.

Radio and associated electronics.

Jet/turbine propulsion technology.

Remote control and automation.

Exploitation and exploration of minerals/petroleum.

Metallurgic applications.

Fuels and fuel technology.

Heavy machinery and power equipment.

Control of the Yellow and Yangtze rivers.

Chemical fertilizers and agricultural mechanization.

Disease prevention and eradication.

Basic theory in natural science.²⁵⁹

On 8 October 1956, the CCP announced the beginnings of the Chinese space program, establishing the first Rocket Research Institute under the Fifth Research Academy, led by Marshall Nie Rongzhen.²⁶⁰ With the concurrent programs in both atomic bomb assembly and missile development under way in the PRC, the demands for the production of chemicals must have been especially acute, and even more so when the Sino-Soviet agreements fell apart in 1959 and the ironically titled "Great Leap Forward" (GLF) began in earnest.

To address the needs of advancing technology and the building of a chemical industry in particular, in 1957 the Chinese Society for Chemical Engineering was established. Although this move certainly was a step in the right direction, the shortage of technical expertise in chemistry was such that in 1959 China was unable to find the necessary materials for building a launching pad for a sounding rocket, nor could it acquire liquid oxygen.²⁶¹ (This is despite the fact that the eminent scientist Qian Xuesen, who himself had carried out post-war intelligence work on Werner Von Braun's V-rockets for the United States, had been the primary project leader for the PRC's Fifth Research Academy since 1956.)

The 1950s were not devoid of any progress in the field of applied chemistry. In 1958 the PRC developed special methods to produce both superphosphate and calcium-magnesium phosphate for the production of fertilizer.²⁶² Despite such efforts, however, chemical fertilizers were more cost effective to import than would be procuring grain from abroad,²⁶³ and even in 1993 fertilizers were still being imported for reasons of cost.²⁶⁴

Unfortunately--not just for the 30 million victims but with regard to China's scientific progress as well--the late 1950s also heralded a period during which Mao's mass movements were just gaining speed. Making a virtue of necessity, these exhortations to pull the wisdom from the grassroots can only be described as antiintellectual pogroms. Reminiscent of Lysenkoism (Lysenkovshina) in the Soviet Union--where the experimental musings of peasants drove Russian biological sciences back to the stone age--elitism in scientific research was decried by all media outlets in China, while workers and farmers were praised for their practical applications of "science," however loosely defined.²⁶⁵ Work in basic research (i.e., that which may not have immediate or obvious practical use) was listed last on the CCP's to do list for the above-mentioned 12-year plan. (This aversion to theoretical work in the sciences and preference for applied chemistry persists even to this day in the PRC.²⁶⁶) Although mass movements may have assisted in prospecting for uranium,²⁶⁷ and the near elimination of schistosomiasis (recall the "People's War against the Snail" in 1950),²⁶⁸ scientific and other higher institutions were being led by party hacks and slogan-mouthed rubes, and as a result "red" nearly always trumped "expert," severely retarding technological advances. Until Marshall Nie Rongzhen intervened, more than two-thirds of Chinese rocket scientists suffered from edema stemming from malnutrition, much like everyone else (save for the party leadership.)²⁶⁹

A condition approaching normalcy returned in 1962, but the GLF clearly had taken its toll on the sciences, especially in the field of chemical engineering. By 1965, a substantial effort made progress bringing back science and technology in China. Between 1950 and the mid-1960s the number of institutes in the Chinese Academy of Sciences (CAS) had grown from 20 to more than 120.

But such progress was stymied by the Cultural Revolution, which brought back the antielitist themes of the GLF, displacing real academics with soldiers, workers, and other political cadres who then were put in charge of the universities. While the inmates were running the asylum, both the Chinese Chemical Society and the Chinese Society for Chemical Engineering were shut down,²⁷⁰ and, according to a CAS academician Tang Youqi, organic, inorganic, and physical chemistry were no longer taught at colleges. Fei Changpei told Chemical & Engineering News that "during the 10 years of the Cultural Revolution, no new scientists and teachers were trained."²⁷¹ Unless Chinese chemists were working directly for the strategic rocket and hydrogen bomb projects, which enjoyed special status, they were not protected from the political onslaught by the Red Guards; many scientists left the PRC during this chaotic period.²⁷²

By at least one account, Mao Zedong and Zhou Enlai tried to reintroduce the need for advanced and professional training in the sciences in the early 1970s. In January 1975, Zhou urged that agriculture, industry, national defense, and science in general be modernized, looking toward the year 2000 as a goal. An "Outline Report" delineating areas where attention was most needed was produced thanks to Zhou's urging; with its emphasis on cultivating professional technicians and professors, however, the document offended the retrograde tendencies of the "Gang of Four." Only after Mao's death in 1977 was a serious discussion of scientific progress possible, and, finally, a National Science Conference was held in March 1978.²⁷³

As Deng Xiaoping's reforms were starting to be implemented in the late 1970s, scientists involved in research or teaching in China could be grouped into the following two cohorts: 1) those 55 and older who had obtained Ph.Ds in the United States, 2) scientists aged between 45 and 55 who had obtained graduate-level training in the USSR, and others at Chinese universities before the 1949 revolution. Chemical engineering departments in the PRC during the late 1970s, for example, primarily consisted of those who had studied in the United States, returned to China during and after the revolution, and were still fondly nostalgic of their previous time spent abroad. Those potential scientists between 30 and 45 years old, however, who would have ushered in the next 20 years of scientific development in the PRC, were effectively lost because of the tumult of the Cultural Revolution.²⁷⁴ The effects still are felt today: official PRC sources indicate that by the year 2000, the generation of leading scientists and academics--and quite a few of these holding influential government positions--will begin retiring, producing an "academic vacuum."²⁷⁵

The late 1970s witnessed a move toward the importation of foreign technology in chemical engineering, as well as investment from overseas. At this stage of development, as far as it pertains to CW agent production, the technical and infrastructure base for chemicals in China probably would have been self-sufficient to produce the first generation of weapon fills, namely, mustard (sulfur and nitrogen), Lewisite, chlorine, phosgene, and hydrocyanic acid. Because changes or rapid improvements in the Chinese chemical industry were unlikely to have occurred during the Cultural Revolution, data from 1977 could be extrapolated modestly downward to the early 1970s, and still have an approximate picture of what China could produce in the way of intermediates and final products:

If China did maintain a stockpile of, at the very least, blister (mustard) and nerve agents (e.g., sarin), stocks of such vital precursors as thiodiglycol (mustard) and phosphorus trichloride (nerve agents) would be needed. Chinese chemical technology developments suggest the rudimentary necessities of three separate CW categories. Example agents demonstrated how production and weaponization could have been affected:

The demands for these starting materials and their CW agent products would include the following.

Mustard (Sulfur)
Mustard had, at least until the end of World War II, been considered the "king of CW agents," and this was nowhere more true than in the Soviet Union, where both Lewinstein and thiodiglycol processes were successively used in its production. Initially, the Lewinstein process used in the Soviet Union was probably the combination of ethylene and sulfur chloride:

2C₂H₄ + S₂Cl₂ ®(CH₂-CH₂Cl)₂S + S

or alternatively with sulfur dichloride,

2 C₂H₄ + SCl₂          (CH₂CH₂Cl)₂S²⁷⁶

This method had many drawbacks, not the least of which was the rapid degradation of mustard as well as explosive (hydrogen) off-gasses that required constant maintenance.²⁷⁷ Even more distressingly, within a span of five years the Soviet Union gauged that 25 percent or more of its mustard decomposed while in storage.²⁷⁸ China almost certainly would have been aware of, and much more in favor of, the thiodiglycol method (Victor Meyer-Clarke process) invented by Germany in World War I. If the experience in the later Soviet Union and Iraq²⁷⁹ is any guide, China would have followed in analogous fashion:

Oxidation of Ethylene

C₂H₄ + HOCl          CH₂-OH-CH₂Cl

A sulfonification step, most likely using hydrogen sulfide

2CH₂-OH-CH₂Cl + Na₂S        [or H₂S]             (CH₂CH₂OH)₂S + NaCl

Utilizing hydrogen disulfide to obtain thiodiglycol, according to Hirsch, led to a 70-75-percent yield. The remaining step is a straightforward chlorination reaction:

(CH₂CH₂OH)₂S      + 2HCl     (CH₂CH₂Cl)₂S+     H₂O²⁸⁰

The precursors and intermediary steps of either process would have presented no difficulty for the PRC, and likewise for the nitrogen mustards. Hirsch reports that nitrogen mustard was probably made in the Soviet Union in World War II "in the usual way by chlorinating the chlorhydrate of triethylamine with thionyl chloride or phosphorus trichloride. A method of chlorination by HCl is also known, but this has many technical difficulties."²⁸¹ If the PRC did follow in similar fashion, again, the materials and processes--with the possible exception of phosphorus trichloride--should not have been problematic, since at least the mid-1960s.

Lewisite
Lewisite production was carried out in the former Soviet Union by reacting arsenic and a chlorinated ethane-mercuric chloride compound. Although as a blister agent it does not receive as much attention as does mustard in PRC publications, it could also have been produced in large amounts with little (relative) difficulty.

Nerve Agents
The major challenge for the production of G-series (V-agents would not appear until at least the mid-1960s²⁸²), would have been finding precursors for nerve agents, primarily phosphorus trichloride (PCl₃), perhaps phosphorus oxychloride (POCl₃), and perhaps later phosphorus pentasulfide (P₂S₅/P₄S₁₀) for VX.

During the 1950s and 1960s, supplies of technical-grade phosphorus in China would have been sparingly small. Although dedicated facilities to thermal phosphoric acid production could have been built to feed the military use of chemicals, similar to the Muscle Shoals plant in the United States during the 1950s, no open-source data exists on China's approach. Moreover, competing needs for basic chemicals among different industries and even strategic weapons programs may have limited developments in this area.

With regard to civilian use, phosphates are extremely important in providing fertilizer and feed supplements to farm animals, food preservation, metal finishing, oil additives, flame retardants, and pesticides, among other uses. Fluorine and phosphorus (in the form of tributyl phosphate), for example, are used both in nuclear fuel processing as well as the manufacture of sarin and soman nerve agents. But, whereas the latter chemicals are utilized within closed systems and some recycling could take place, CW agent production in the hundreds of tons would have presented rather daunting challenges, at least in the years before 1978:

In 1972, technology for the production of elemental phosphorus via thermal methods was held by the United States, United Kingdom, and West Germany, the latter (Uhde of Farbwerke Hoechst) having built the Chimkent plant in the Soviet Union.²⁸³ Relations between the PRC and these countries were at their nadir during the early 1970s. Without similar turnkey plants China would have experienced great difficulty becoming sufficiently self-sufficient to build large stockpiles of nerve agents, although modest amounts could be produced by diverting elemental phosphorus from other uses, or via the time-consuming purification of phosphorus via the wet (acid) process.

Further, China claims to have developed an organophosphorus agent detector in 1971,²⁸⁴ apparently to fulfill a need to detect the use of contact insecticides, and perhaps in part stemming from concerns regarding Soviet CW capabilities on the border. Insecticides of concern probably would have included ethyl parathion and methyl parathion, both in use by the late 1960s, and resulting in many accidental poisonings in China.²⁸⁵ By 1971, the price of either insecticide dropped rather dramatically compared to 1966 levels, from US $0.75 per pound to US $0.40,²⁸⁶ leading to their wide-scale use. Perhaps an example of "off the shelf" technology used for military purposes, this detector may not have changed substantively since its inception.

Brain Power
Economic reforms in the late 1970s resulted in a concomitant increase in institutions devoted to applied chemistry and research. By 1991, 240 R&D agencies had been established for chemical industry, consisting of 20,000 scientists and engineers, plus an additional 12,000 technical staff.²⁸⁷ This progress is remarkable, particularly when considering that in 1984 the PRC only produced 15 graduates in science at the Ph.D. level, and these were primarily in theoretical research.²⁸⁸ Using data from a year earlier, these numbers reflect a large proportion of effort toward the chemical sciences, comprising 40 percent of all scientific research institutes, and another 40 percent of those personnel classified as "scientists or engineers."²⁸⁹ Today, China produces thousands of Ph.D.s in the sciences, although the significant problem of brain drain to more lucrative jobs in foreign countries continues.²⁹⁰

Infrastructure
In a barter arrangement that included mostly petroleum, Japan and China stuck a deal in 1978 for $20 billion in which Japan was to sell manufacturing facilities to produce ethylene, fertilizer, and synthetic leather.²⁹¹ From 1982 onward, foreigners invested in some 940 ventures over the next 10 years.²⁹² By 1994, 5,540 chemical enterprises with overseas funding were started in the PRC,²⁹³ and in 1997 there were 6,800.²⁹⁴ Foreign chemical conglomerates have since participated in the building of large chemical facilities, particularly ethylene plants to satisfy large demands for plastics and other polymers:

Production of Phosphorus and Organophosphorus (OP) Compounds in the PRC
In 1999, chemical outputs in China surpass levels originally targeted in Beijing's respective five-year plans, particularly in the area of pesticides. For example, the production plan in 1995 for pesticides (i.e., herbicides and insecticides) was 230,000 tons, with an actual output of 349,000 tons.²⁹⁵ This year the capacity is estimated at 750,000 tons for all pesticides, and outputs have averaged nearly 400,000 tons per year. In chemical fertilizers, the PRC has nearly always been self-sufficient in terms of nitrogen but continues to import potash and phosphate fertilizers.²⁹⁶

As demonstrated previously, production of elemental phosphorus for food-grade phosphates and chemical intermediates in OP chemistry has typically been via thermal processes, although recently the world market has been changing toward wet-process phosphates. The significant point about phosphate production in the PRC is the ability to produce large amounts of pure phosphorus trichloride (PCl3) and phosphorus pentasulfide (P2S5), precursors for G-series nerve agents and V-agents, respectively. With large phosphate reserves, even if the phosphate content in rock is of low quality,²⁹⁷ China has increased dramatically its production of several phosphate-related chemicals. Yunnan Province, for example, utilizes hydroelectric power to provide a source for thermal phosphoric acid. In 1990 plans were made to develop a 60,000 metric ton/year phosphoric acid plant in Yunnan.²⁹⁸ Bottlenecks and other production problems remain. In 1997, a manager based at a foreign-invested chemical company in China that is heavily involved in pesticide production, remarked:

China lacks the essential intermediates to carry out contract synthesis. They often have to do more steps due to this deficiency. We often have to ship raw materials from the West to ensure production schedules.²⁹⁹

In 1998, however, apparently there was enough phosphorus pentasulfide to allow the clandestine shipment of 500 tons to Iran via a Norinco front company in Hong Kong.³⁰⁰ Although it violates the Australia Group chemical precursor restrictions, which brought about sanctions from Great Britain and the United States, P2S5 technically is not a controlled substance under the CWC. Though not a member, and a vociferous opponent, of the Australia Group, China has voluntarily added P2S5 to its list of controlled chemical exports.³⁰¹

Phosphorus oxychloride, which is produced in China by more than 20 facilities, may also be considered within the context of nerve agent precursors, but primarily for tabun (GA) production. Interestingly in this regard, Lianshiu Chemical Works has introduced a process that combines sulfur, chlorine, and phosphorus trichloride over a catalyst to produce both phosphorus oxychloride (POCl3) and thionyl chloride, 1 ton of POCl3 producing about 0.8 ton of thionyl chloride during the process.³⁰² (Thionyl chloride is considered a chemical that has proliferative use in nerve agent synthesis.)

The PRC has made a concerted effort to move away from chlorinated hydrocarbons to organophosphorus (OP) pesticides.³⁰³ Two major pesticide research institutes were recently established in China, one in Shenyang (National Pesticide Engineering Research Center), and the other in Shanghai (National Southern Pesticide Formulation Center). In 1995, for the first time since the Communist revolution, representatives from the China Pesticide Industry Association visited their counterparts on Taiwan. Then, more than 10 joint-venture projects were financed by Taiwan firms in the PRC and were devoted to the export of pesticides, amounting to "tens of million[s] US dollars."³⁰⁴ In 1998, China produced 382,000 tons³⁰⁵ of pesticides and of this total exported 100,000 tons, earning a reported US $320 million.³⁰⁶

One of the more significant developments has been this ability on the part of the PRC to produce large amounts of OP pesticides, including an indigenous supply of precursors that could be used in nerve agent synthesis, particularly phosphorus trichloride and phosphorus pentasulfide. Although the production of OP pesticides is quite different from that of the extremely toxic nerve agents, the basic expertise and the basic starting materials are not that far removed. Significantly, the PRC produces the following OP pesticides, on the order of 5,000 tons/year (or more, 1994-1995):³⁰⁷

Pesticide Table

If local production and consumption of pesticides in the PRC (estimated by Monsanto to be approximately US $ 600,000³⁰⁸) remains at current levels, starting materials will also be in strong demand, especially for phosphorus trichloride. Of the pesticides listed above, two in particular often use the phosphorus pentasulfide route in synthesis. Many others probably are produced in smaller quantities in China:

Dimethoate (Dongguo)

example, 4CH₃OH + P₂S₅ Þ 2(CH₃O)₂PSSH + 2H₂S (+ additional organochlorine steps)³⁰⁹

Parathion (Duiliulin)

example, 4C₂H₃OH + P₂S₅ (or PCl₃) Þ(Chlorinating step in the case of synthesis with P₂S₅)³¹⁰

These two formulations alone would require enormous quantities of phosphorus if 5,000 tons of agent were to be produced. Thus, the phosphorus industry in China is capable of producing large quantities.

Conclusion

In terms of chemical technology and knowledge base, some crossover from erstwhile Soviet assistance would have occurred in the late 1950s, particularly in the areas of fluorine chemistry and organophosphorus compounds. (The latter two are critical for the enrichment of uranium, and form the basis for the German series of toxic nerve agents, respectively). Nonetheless, at the time of the Sino-Soviet feud, as far as military chemistry is concerned, China was even more backward than the Soviet Union. (Substantial East German assistance to the Soviet Union probably did not occur until 1965--long after the Sino-Soviet split.³¹¹)

We do not know what degree of technological competence and production levels in chemical manufacture existed in China before 1978, especially during the times when economic data was considered "secret." By the 1990s, however, China clearly has mastered many commercial methods of producing fine chemicals, including key precursors and intermediates that could be diverted to CW-agent manufacture.

If China has in fact destroyed its chemical weapons--and by its reported documentation to the Organisation for the Prohibition of Chemical Weapons (OPCW) in The Hague, it has--the PRC did so at a time when it could produce nearly any of the known CW agents in mass quantities. From an economic point of view, joining the CWC was for China a strategic decision to ensure that it's "pillar industry," namely chemical, would not be impeded by international export controls. An optimistic assessment would be that Deng Xiaoping's policy to subordinate the military to a strong economy applies to the Chinese chemical industry as well. The pessimist would note that, in the event of a major crisis, the PRC would have little trouble reconstituting a large chemical weapons arsenal within a relatively short time.

Chinese Perspectives on BW

Official PRC histories of BW, justifiably, recount at length the experience of Japan's invasion of China, and the gruesome experiments conducted by Gen. Ishii Shiro and his Unit 731. Also mentioned is a report about "Operation Golden Triangle," allegedly from a Russian defector who fled to Germany, claiming that near the end of the Second World War the Soviet Union conducted experiments with plague, anthrax, and cholera in Soviet-occupied Mongolia.³¹²

Allegations that the United States routinely conducted BW during the Korean war, however mendacious and insupportable, also seem to be accepted as fact by the PLA. The book on BW printed by the PRC's National Defense Press, for example, extensively covers the issue. Defense Minister Chi Haotian, who served during the Korean conflict and wrote the preface to the series on weapons and war, including CBW,may have influenced the book to publish a lengthy laundry list of "biological crimes" committed by US forces. Nonetheless, despite no reliable evidence of US complicity (and even recent proof that the Chinese themselves have colluded with North Korea to fabricate biological weapons), the charges are ingrained among senior Chinese leaders. The recent publication by Endicott and Hagerman,³¹³ as well as the unreconstructed claims of Maoist fellow traveler Joseph Needham, may have sealed the idea even further, for now there is "Western" concurrence to the allegations.

At the very least, this legend provides a historical starting point for the PLA's development of anti-BW defense measures and training. But with regard to future arms control agreements and intelligence assessments, the belief of the PRC that the United States employed biological weapons during the Korean war is significant. The Chinese, who see even the Opium War of the 1840s as having happened only yesterday, will be influenced by their interpretation of such historical events, no matter whether true or false.

China alleges the following US BW attacks in North Korea:

• While the United States was retreating south under attack by the united Sino-Korean Army, in December of 1950 the United States military disseminated smallpox against the Korean capital of Pyongyang, Hwanghae do, and other areas.

• On the 28th of January, US forces used aircraft on areas such as [Lung Zhao dong], southeast of Inchon, [Long shui dong], etc., to disseminate large quantities of three insect types never before seen in Korea: The first type was a kind of black fly, the second was in the form of something similar to fleas, and the third was a kind of tick.

• Laboratory test evidence showed that the insects disseminated by the United States carried plague, choloera, and other infections disease-causing pathogens.

• Accounts have revealed that the US Chemical Corps operations department produced 16 different types of deadly BW agents in large quantities. In March 1951, [Brigadier General Crawford] Sams³¹⁴ who was in charge of the Public Health and Welfare department of the "United Nations Army" command, led the No. 1091 microbiology lab on a landing boat to Wonsan harbour, and onward to Koje island. They used POWs as targets for biological weapons experiments. As the US military progressed in their manufacture of biological weapons, they utilized the work of the Japanese war criminal Ishii Shiro, Wakamatsu Yujiro, Kitano Masaji, etc., and even sent them to South Korea.

• [E]xamples of various technologies used ranged from fountain pens filled with infectious disease-causing black ink to feathers contaminated with anthrax bacilli, as well as fleas, lice, and mosquitoes infected with plague and yellow fever. Various kinds of flies, fleas, spiders, beetles, bedbugs, crickets and other insects were found, many of which had never been seen before in Korea.

• The types of bacteria found were Vibrio cholerae, Salmonella typhi (typhoid), Yersinia pestis, paratyphoid (A and B types), the causative agent of typhus, and Shigella dysenteria. Laboratory results showed that the insects tossed down carried plague, cholera, and other infectious diseases. . . . Not long after discovering these containers, many people came down with plague or cholera. Of 53 total plague victims, 39 died.

• According to relevant information, from the 28th of January, 1952, to the 31st of March, the US military disseminated bacteria as many as 804 times in North Korea.

• Several years later, the American government acknowledged that they had used biological weapons during the Korean War.³¹⁵

Similar conspiracy type of allegations seem to continue into the 1990s. For example, the PLA may actually believe that unusual outbreaks of hemorrhagic fever that occurred in Kenya in 1995, were in fact the results of US BW experiments,³¹⁶ and makes similar insinuations concerning the Ebola virus outbreaks in Zaire.³¹⁷

BW Offense
Writings are scanty on Chinese CW capabilities and even more so on BW. A PRC official from the Chinese Ministry of Foreign Affairs assured me that China has no biological weapons.³¹⁸ A book on the subject, with the imprimatur of Chi Haotian, states categorically that "China has never manufactured nor possessed biological weapons."³¹⁹

According to its submitted Biological and Toxin Weapons Convention (BWC) declarations, the PRC has declared the following facilities as having a "national defensive biological warfare R&D program," and listed the following facilities:³²⁰

Dual Use/BW Defense Research Facilities (1993)

• Institute of Microbiology and Epidemiology.

Vaccine Production Facilities

• National Vaccine and Serum Institute.

• Shanghai Institute of Biological Products.

• Lanzhou Institute of Biological Products.

• Changchun Institute of Biological Products.

• Wuhan Institute of Biological Products.

• Chengdu Institute of Biological Products.

• Institute of Medical Biology, Chinese Academy of Medical Sciences.

The PRC claims that no BL-4 (highest containment for extremely contagious and virulent organisms) laboratories exist, at least as far as BW-related research is concerned. Most biological weapons, however, can be produced and studied in BL-1-3 conditions, and a BL-4 facility is less relevant from a weaponization capability standpoint.³²¹

Little of the scientific literature that the PRC reports in its BWC declarations is worth noting except for public-health-related research on bioaerosols and reviews on staphylococcal toxins. The remaining citations consist of the typical infectious disease reporting and epidemiological studies on hepatitis (of just about every type), hemorrhagic fever with renal syndrome (HFRS), and insect abatement programs.

Allegations of BW Activity in Xinjiang Province
Ken Alibek, formerly with the Soviet/Russian Biopreparat BW complex, suggests that an outbreak of hemorrhagic fever in Xinjiang Province near Lop Nor was the result of Chinese activity in BW research:

Intelligence sources found evidence of two epidemics of hemorrhagic fever in this area in the late 1980s, where these diseases were previously unknown. Our analysts concluded that they were caused by an accident in a lab where Chinese scientists were weaponizing viral diseases.³²²

Another source in Taiwan told me that he felt certain a BW facility of some sort did exist in Xinjiang Province, not far from the nuclear testing facilities.³²³

As for the allegations of the source of outbreaks in Xinjiang, we should be cautious because of the natural occurrence of Xinjiang hemorrhagic fever (HF) endemic to the area, a variant of Crimean-Congo HF of the bunyaviridae-type virus that occasionally strikes in northeastern China, and where a significant outbreak occurred in 1968.³²⁴ But even if we discount the 1980 outbreaks as having military-related origin, we cannot rule out the actual existence of the BW-related facility. The list of declared research and production sites above shows nothing further northeast than Gansu Province. The Soviet Union, in open violation of the BWC, built the largest BW capability thus far known. Given the poor track record of the BWC as it is currently implemented (or more accurately, is not being implemented), China probably is withholding much information about its BW research, although such research primarily may be defensive in nature.

Agricultural BW
A newspaper in the United States intimated that the foot and mouth disease (FMD) outbreak in Taiwan could have been due to mainland Chinese sabotage.³²⁵ The largest known FMD outbreak, it has caused more than $5 billion damage to the Taiwanese pig farming industry.³²⁶ After hearing a presentation by Dr. Terrance Wilson on the subject,³²⁷ and following discussions with some knowledgeable Taiwanese, I am fairly certain that the FMD outbreak was purely accidental. A similar conclusion was also reached in the Taiwan agricultural community. For example, Stock-Farming of Tendays [sic] (Nongmu Xunkan), 25 September 1999, writes:

The outbreak of FMD in Taiwan was caused by the introduction of virus through either the smuggling of goods or related agricultural products. As a consequence, the defense against such smuggling is of great importance. . . It was finally determined by means of analysis in foreign research institute(s) that the FMD outbreak was absolutely the same as that in the mainland, thus proving that infection was brought into Taiwan from the PRC. It was completely because of smuggling meat products across the boundary by smuggling that caused great economic losses to Taiwan amounting to one percent of (1997)'s [GNP].³²⁸

The few Chinese writings on the subject of BW preponderantly discuss the allegations of US use of BW during the Korean war. Thus, even today, there is emphasis on training and equipment to rid the immediate environs of insects and vermin, as if modern armies would deploy such crude methods of delivery. For example, to foil the enemy's germ-laden, flying insects or plague-infested rats, the PLA handbook on BW even suggests how to use simple brooms and nets, and procedures for burying the offensive detritus.³²⁹

BW Defense in the PLA

In keeping with the definition of BW as "public health in reverse," PRC writings on the subject treat the matter more in terms of infectious disease control, an approach that is standard everywhere. As one would expect, considerable amount of research has been conducted in China on potential BW agents including tularemia, Q fever, plague, anthrax, West and Eastern Equine Encephalitis, psittacosis, among others.³³⁰ Some specialized equipment has also been fielded in some unspecified numbers to counter the threat of BW to PLA troops.

Type 76 Microbe Sampling Kit³³¹
First introduced in 1975, and includes the 76-1 variant,³³² this portable laboratory can test surface, waterborne, and airborne particles to determine the presence of BW agent threats, and also has five different types of insect and small animal reference specimens. Resembling a low-tech, gravitation/settle plate,³³³ a small, rotating mechanism is placed windward, and aerosol particles will adhere to the sampling or petri dish. Disinfectant is supplied along with culturing supplies.

Large-Volume Electrostatic Air Sampler³³⁴
This equipment has no classification number, and little information is provided concerning its attributes. It probably is similar to the corona discharge-based large volume air sampler (LVAS) used in the West. This technology in general offers excellent results, and is capable of isolating viral particles from the air, including rabies and human respiratory disease viruses.³³⁵

JWL-I Model Bioaerosol Sampler³³⁶
Like the LVAS mentioned above, the reference to this equipment offers little in the way of details. This automated air sampler resembles most closely a single stage impactor, drawing in air and depositing aerosolized particles onto agar for further testing. An example of this type of instrumentation is the Casella slit-to-agar, single-stage impactor used in civilian environmental monitoring.³³⁷

In 1974 an improved version of the WJ-85 microbiological laboratory vehicles was introduced,³³⁸ and could have resulted in this motorized laboratory platform, described as somewhere between "a railway car and a sedan," is separated into three sections, with airtight sealed gaskets on the doorways. The forward section houses the driver and carriage for occupants, the midsection contains the laboratory room (See Mobile BW Assessment Laboratory), and the rear section contains decontamination apparatus plus extra clothing. Laboratory equipment includes a glass glove box for handling infectious material, a bacteriostatic device, a refrigerator, an incubator (hengwenxiang), a fluorescent microscope, an inverted microscope, culture media, diagnostic reagents, cell culture instruments, etc. A separate station allows testing for bacteria and viruses, accommodating up to four people. Some 200 bacteria and 50 virus samples for reference and identification are supplied with the laboratory vehicle.

PLA Military Medicine and BW Defense
The earliest semblance of routinized BW defense in the PLA were the 1952 sanitation/anti-plague units, formed during the involvement of the Chinese People's Volunteer Army in Korea. At the same time, educational campaigns to rid disease-carrying pests were conducted, and, when combined with experience of the supposed BW casualties treated during the Korean war, "a great victory was achieved in anti-bacterial warfare."³³⁹

Building a more formal curriculum in BW defense, the PLA continued work in anti-plague research, and in 1954 delegations and students visited the Soviet Union for expertise in microbiology and infectious disease.³⁴⁰ Perhaps in tandem with the fanatical anti-pest campaigns carried out during the Great Leap Forward, a full-fledged, national investigative research project was carried out during 1958-61, led by the Military Medical Science University and sanitation units, from every military region, on down to individual cadres. By 1984, M.S. degrees were being awarded in the related specialization of BW defense by the Military Medical Science University.³⁴¹

---

Evan S. Medeiros

Beginning in the early 1980s, China's weapons proliferation activities emerged as an issue of growing concern for US policymakers. This trend has persisted for close to 20 years. Chinese companies in the last two decades have exported to several countries a variety of goods useful in building nuclear weapons, chemical weapons, and ballistic and cruise missiles. In some cases, China has provided critical materials, equipment, and technical assistance to nations who could not otherwise acquire these items for their weapons programs. Most notably, China provided Pakistan with a basic nuclear weapon design and substantial assistance in fabricating weapons-grade nuclear material. Moreover, China has provided some countries with production technologies, allowing these nations to indigenously build certain missile systems with little external assistance. Although China's proliferation behavior over the last two decades has been highly egregious, it has also improved dramatically in recent years, especially since the mid-1990s. The Chinese government has gradually signed onto a number of key nonproliferation treaties, such as the Nuclear Nonproliferation Treaty (NPT) and the Chemical Weapons Convention (CWC) and has developed internal bureaucratic and regulatory structures to carry out these commitments. This is not to say that China no longer engages in exports of proliferation concern to the United States. Rather, the nature of the China-proliferation problem is fundamentally different, and US nonproliferation policies on China should be changed accordingly.

This paper addresses one central question: What is the current scope of China's proliferation activities related to weapons of mass destruction (WMD),³⁴² and how has it changed over the last 20 years? Answering this question will help to establish the factual and conceptual basis for understanding the nature of the problem and determining viable options for US policy-makers in an effort to change Chinese behavior. To evaluate the scope of Chinese proliferation activities, this paper considers three indicators: the geographic scope of China's WMD exports, the types of exports (e.g., weapon-specific or dual-use technologies) and their contribution to WMD programs, and the frequency of such transfers. These three indicators are applied to three case studies around which this paper is structured; the three case studies cover China nuclear exports, missile (ballistic and cruise) exports, and chemical exports. The paper examines each of these case studies over a 20-year period to provide a historical perspective on the shifts and changes in the scope of China's WMD proliferation activities.

Drawing on this analytical framework, this paper argues that in the last two decades the overall scope of Chinese proliferation activities has declined across the board. The geographic distribution of Chinese proliferation-relevant exports has narrowed from almost a dozen countries to three: Iran, Pakistan, and to a lesser extent North Korea. The character of China's exports similarly narrowed from a broad range of nuclear materials and equipment (much of it unsafeguarded) and complete missile systems to exports of dual-use nuclear, missile, and chemical technologies today. In addition, during much of the 1980s and 1990s, China's nuclear and missile assistance directly contributed to the nuclear and missile programs in other countries; today such assistance is indirect, at best. The frequency of such exports also appears to have declined to a dribble of dual-use items, albeit declining less than the scope or technical character of China's exports. Despite this overall narrowing of China's WMD-related exports, further progress will be slow. Significant policy differences between Washington and Beijing exist about controlling dual-use nuclear, chemical, and missile goods to Iran and Pakistan. These contrasting policies are based on profound differences between the respective foreign policy approaches of the United States and China to Iran and Pakistan, the utility of supply-side technology control regimes, China's ability to implement and enforce its export control laws, and linkages to such bilateral issues as US arms sales to Taiwan.

This analysis of the scope of China's WMD exports requires a major caveat, however. Tracking China's nuclear, chemical and missile exports based on nonclassified, open-source information is an inherently difficult task. Reliable and comprehensive information is scarce. Much of the information--especially detailed technical data--is based on press accounts of leaked intelligence information. This classified data is often leaked for specific political purposes, is often incomplete, and thus is of questionable reliability. To offset these informational weaknesses, this paper relies on multiple sourcing combined with extensive conversations with US and Chinese officials from a variety of government agencies in both Washington and Beijing.

China and Nuclear Proliferation³⁴³

Chinese nuclear exports have changed dramatically over the course of the last twenty years. The geographic distribution of Chinese nuclear exports has narrowed, the character of nuclear items sold and their relative contribution to nuclear proliferation has positively changed and the frequency of nuclear exports (including technical assistance) has decreased significantly. As of 1999, US concerns about Chinese actions that contribute to nuclear proliferation are fundamentally different as compared to 20 years ago. To detail these trends, this section compares China's nuclear exports in the 1980s and 1990s.

Chinese Nuclear Exports in the 1980s³⁴⁴

Beginning in the early 1980s (only a few years after Sino-US normalization), Chinese state owned companies began providing a variety of nuclear assistance to an eclectic mix of countries all over the world. Chinese nuclear companies used nuclear exports as a means to generate hard currency as China opened up to the outside world and sought to better integrate its economy with Western ones. Central authorities in Beijing encouraged nuclear exports as a way for China's large military-nuclear complex to diversify into producing civilian goods. The profits from these activities were then funneled into improving China's dilapidated nuclear infrastructure, both military and civilian.³⁴⁵ Given these pressures, Chinese companies began providing nuclear equipment, materials, and technical assistance to such countries as Argentina, Algeria, Brazil, Chile, India, Iran, possibly Iraq, Pakistan, and South Africa. Initially, Chinese companies sought to create long-term relationships with many of them. Throughout the 1980s, China signed nuclear cooperation agreements (NCAs) with Argentina, Algeria, Brazil, Iran, Pakistan, and a variety of other countries of lesser proliferation concern in an effort to create sustained export relationships.³⁴⁶ Many of these NCAs are currently active, although trade between China and many of these countries has been scaled down in recent years.

China's nuclear sales covered a variety of nuclear items and technical assistance that directly contributed to the military nuclear activities in several countries. In the early 1980s, Chinese nuclear exports were not placed under International Atomic Energy Agency (IAEA) safeguards, which facilitated their use in military nuclear activities. Chinese firms exported different types of reactor fuel, complete reactors, reactor technologies, technical assistance for indigenous nuclear projects, and nuclear facility training. In the case of Pakistan, China also provided substantial direct assistance in designing and building nuclear weapons. Beginning around 1983, China sold Argentina a wide variety of nuclear materials such as uranium concentrate (yellow cake), uranium hexafloride, 20-percent low-enriched uranium (LEU), and heavy water. None of these exports was under IAEA safeguards, and all probably were used in Argentina's dual-use nuclear program. China's exports to Brazil were less extensive but also probably were diverted to Brazil's military nuclear activities. China sold some 200 kg of LEU (3-20-percent enriched) to Brazil in the early 1980s, none of which was subject to international safeguards. Of greater proliferation significance were China's nuclear exports to South Africa, which operated a dedicated nuclear weapon program--as opposed to the "military options" programs in Brazil and Argentina. South Africa purchased unsafeguarded LEU and uranium hexafloride that probably were used to fuel its pilot enrichment plant at Pelindaba East. In addition, China sold South Africa 60 metric tons (MT) of unsafeguarded heavy water for other nuclear projects. China's strong financial motives for exporting nuclear items were especially evident in its willingness to provide nuclear fuel to its strategic competitors. Between 1982 and 1987, China provided India with 130-250 MT of unsafeguarded heavy water; this item was probably used in India's CANDU reactors that for many years served as the main plutonium producers for India's nuclear weapons program.

Chinese nuclear exports in the 1980s went beyond nuclear fuel. In 1983, China and Algeria signed an agreement for the construction of a small 15 MW heavy-water research reactor.³⁴⁷ The reactor initially was not subject to any international inspection, and several indicators suggested the reactor could have been part of a nascent nuclear weapons program in Algeria.³⁴⁸ Chinese officials originally argued that the reactor deal was exempt from inspection because the contract was signed in 1983, a year before China joined the IAEA. Only after significant US and international pressure was applied beginning in 1988 (when US satellites noticed the reactor's construction) did China and Algeria agreed to open the reactor to IAEA inspection when it was completed.³⁴⁹

Ironically, China's nuclear assistance to Iran was of lesser concern in the 1980s. China's nuclear relationship with Iran was just taking shape in the 1980s and did not flourish until the 1990s. This assistance involved limited amounts of training and nuclear equipment exports; none of it was directly applicable to nuclear weapon development, and all of China's assistance was placed under safeguards. Reports say that China and Iran signed a secret nuclear cooperation as early as 1985. The first manifestation of this accord was the training of Iranian technicians in China; by 1991 some 15 nuclear engineers from Iran's Isfahan facility had been trained in nuclear reactor design and research in China.³⁵⁰ In 1989, China's initial nuclear exports to Iran were minimal and involved transfers of two or three electromagnetic isotope separators (EMIS or calutrons) and a 27-kilowatt (kW) subcritical reactor. Although EMIS is used to enrich uranium, it is highly inefficient, and hundreds are needed to produce significant quantities of enriched uranium.³⁵¹ The Chinese-supplied calutrons were placed under IAEA safeguards and stationed at two facilities in Iran. The Chinese used the subcritical reactor to began training the Iranians in basic nuclear physics, isotope production, and reactor operation. Such training--both in Iran and in China--provided the Iranians with a technical baseline from which greater expertise and presumably nuclear weapon knowledge could eventually be developed. Yet all of China's aid was consistent with and, in fact, encouraged under the Nuclear Nonproliferation Treaty (NPT).

During the 1980s, China's nuclear relationship with Pakistan was Beijing's most extensive in terms of the technologies/assistance provided, the contribution to proliferation, and the frequency of transfers. China directly assisted Pakistan's nuclear weapon program. In the early part of the decade, China reportedly provided Pakistan with a nuclear weapon design of a crude but highly reliable Hiroshima-sized weapon; reports say China also transferred enough HEU for one or two cores for this weapon; and in 1989 China may have allowed Pakistani scientists to observe nuclear tests at Lop Nor.³⁵² In addition, Chinese technicians provided equipment and assistance to several of Pakistan's unsafeguarded fuel-cycle facilities that supported the nuclear weapons program. In 1986, China concluded a comprehensive nuclear cooperation agreement with Pakistan. Under this accord, Chinese companies supplied Pakistan with a variety of nuclear products and services, ranging from uranium enrichment technology to research and power reactors. Specifically, Chinese scientists may have assisted Pakistan with construction of the PARR-2 research reactor and operating uranium enrichment centrifuges at the Kahuta facility. China also reportedly transferred enough tritium gas to Pakistan for a few nuclear weapons.³⁵³

China's extensive nuclear exports to Pakistan, Argentina, Brazil, and South Africa during the 1980s largely are explained by the weakness of China's formal nonproliferation commitments combined with the relative lack of bureaucratic infrastructure in China to support nuclear nonproliferation. For years, Chinese officials had rejected the NPT as a biased and inherently discriminatory treaty and viewed nonproliferation as a means for the superpowers to entrench their nuclear superiority by denying other nations equivalent capabilities.³⁵⁴ This view began to change slowly in the 1980s as China re-engaged with the international community.

Beginning in 1984, China made two initial nonproliferation commitments, neither of which was verifiable or enforceable. First, China joined the IAEA and pledged to require safeguards on all of its nuclear exports to non-nuclear-weapon states; this promise also included third-party retransfer prohibitions. Second, China's then Premier Zhao Ziyang provided a verbal commitment in a White House toast that China does not "advocate or encourage nuclear proliferation" and that China "does not engage in nuclear proliferation ourselves, nor do we help other countries develop nuclear weapons." Both of these commitments probably were motivated by the Chinese desire to conclude negotiations on a bilateral nuclear cooperation agreement so that China could gain access to US reactor technologies. The ability and/or willingness of the government to implement them was limited. Neither of these commitments was part of China's export law and probably were not communicated to the Chinese nuclear companies involved in exporting goods. At that time, China had no functioning export control system, set of export control laws, or technology lists that governed China's nonproliferation commitments. China's arms control and nonproliferation community similarly was underdeveloped. China's nascent community of arms control and nonproliferation experts were based mainly at the UN in New York or at the Conference on Disarmament in Geneva and focused on broad arms control issues like nuclear disarmament and nuclear testing. Nonproliferation was not an independent discipline in China. Also, there was little bureaucratic support in the Foreign or Trade ministries to understand or implement the 1984 pledges. China's continued nuclear relationship with Pakistan throughout the 1980s provides the best evidence of the limited scope and weaknesses of China's initial nonproliferation pledges.³⁵⁵

Chinese Nuclear Exports in the 1990s

By the early 1990s, the character of China's nuclear exports had begun to change. Chinese companies stopped providing nuclear-specific materials, equipment, and technologies to unsafeguarded facilities in countries with suspected nuclear weapons programs like Argentina, Brazil, India, and South Africa. The geographic scope of China's nuclear exports declined to cover mainly Iran and Pakistan; the character of China's remaining nuclear exports gradually shifted to dual-use nuclear goods; and the relative contribution of these exports to nuclear proliferation accordingly declined. These developments were further enhanced by the gradual expansion throughout the 1990s of China's formal nuclear nonproliferation commitments (China signed the NPT in 1992), its nuclear export control laws, and bureaucratic support within China for nuclear nonproliferation. These trends are detailed below.

China's nuclear cooperation with Iran expanded in the early part of the 1990s, but by the end of the decade it had almost entirely stopped. This contraction was a direct result of US pressure on China to cease all nuclear cooperation with Iran. Beginning in the early 1990s, China signed several reactor deals and contracts for other fuel-cycle-related facilities with Iran. China sold Iran a small zero-power research reactor and a zirconium tube production facility. Both of these were placed under IAEA safeguards and have been visited several times by inspectors. During this same period, China and Iran concluded a deal for a small 20 MW research reactor. Yet, by 1992 China canceled the deal under US pressure. Chinese officials were concerned that the deal would have complicated China's bid to secure renewal of Most-Favored-Nation (MFN) trading status with the United States.³⁵⁶

Around 1992, China and Iran signed another, larger contract for the export of two 300 megawatts electric (MWe) Qinshan-type reactors and a uranium hexafloride (UF₆) production facility. As before, the United States opposed these transactions, fearing they would contribute to Iran's nascent nuclear weapons program. US officials argued that two reactors and the UF₆ facility--although legal under the NPT--would move Iran further up the "nuclear-weapon ladder." Chinese officials countered that Iran was a member of the NPT, previous inspections had found no evidence of noncompliance with the treaty, and all these facilities were subject to IAEA safeguards.³⁵⁷ Sino-US debates about these facilities came to a head in 1997 as Beijing and Washington began to discuss implementation of the dormant 1985 US-China nuclear cooperation agreement. Both sides finally reached an agreement during the Clinton-Jiang summit in October 1997. In exchange for China's cancellation of these two projects and its agreement to halt all future nuclear cooperation with Iran, the United States would allow the NCA to enter into force. As part of this deal, China was allowed to continue two nuclear projects: the zero-yield reactor and the zirconium production facility.³⁵⁸ The CIA has verified in several reports to Congress that since 1997 China continues to adhere to this pledge to end all nuclear cooperation with Iran.³⁵⁹ Thus, as of 1999, almost all of China's nuclear exports to Iran have stopped. This virtual halt to Sino-Iranian nuclear cooperation stands in stark contrast to the ambitious plans for bilateral nuclear cooperation that Tehran and Beijing reached at the beginning of the decade.

During the 1990s, direct assistance to Pakistan's nuclear weapons program appears to have ended, while the scope of China's other assistance has narrowed significantly. Chinese firms provided Pakistan with a variety of nuclear goods and technical assistance that indirectly contributed to Pakistan's nuclear weapons program. Much of the assistance over the last 10 years involved exports of dual-use nuclear goods and nonnuclear technologies to unsafeguarded facilities involved in fabricating nuclear materials for weapons. China's assistance to Pakistan on three projects will help to elucidate the scope of the relationship.

First, China reportedly provided Pakistan with construction assistance for a 50-70-MW plutonium production reactor at Khushab; this facility is not under IAEA safeguards and, if operational, would provide Pakistan with an unsafeguarded source of plutonium-laden spent fuel. In 1995, for example, a Chinese company exported a special industrial furnace and high-tech diagnostic equipment to the Khushab facility.³⁶⁰ Although these technologies have clear civilian functions, their destination suggested a more pernicious end use. China has since promised to halt all assistance to this and other unsafeguarded facilities.

Second, Chinese firms reportedly were assisting Pakistan with the construction of a partially completed, unsafeguarded reprocessing center located at Chasma; if Pakistan completes this facility, then operating it in conjunction with the Khushab facility would provide Pakistan with an unsafeguarded source of plutonium. Also at the Chasma site, China is building a 300 MWe power reactor for electrical generation purposes. The reactor has little proliferation relevance and will be under IAEA safeguards.³⁶¹ Yet, Chinese work on the reactor could function as a "cover" for assistance to the Chasma reprocessing facility or other projects in Pakistan. Some sources indicate that Chinese and Pakistani experts already have considered this possibility.³⁶²

Third, in 1995 a Chinese firm supplied Pakistan's Kahuta Research Laboratory with 5,000 custom-made ring magnets for use in high-speed gas centrifuges. This plant, which is not under international safeguards, serves as Pakistan's main source of HEU for the nuclear weapons program. The proliferation relevance of these specialized magnets is not readily evident, however. They are a dual-use item that are not listed on any international nuclear trigger list but rather are part of a key technology, magnetic suspension bearing, which is a controlled as a dual-use item. Yet, the sale of these magnets raised concern on the part of the US due to their custom-made design for enrichment centrifuges and, more important, their destination at the Kahuta facility. The ring magnet incident was particularly significant because it raised questions about the ability of Chinese officials to control the actions of Chinese firms. Chinese officials claimed not to know about the magnet deal, and thus argued they should not be held accountable for it.³⁶³

The ring-magnet incident was especially important because it both highlighted the emerging problem in the 1990s of the government's difficulty in controlling exports, and catalyzed China to institutionalize many of its nonproliferation commitments. Following the episode, Chinese officials began to clarify its nuclear nonproliferation commitments and to codify them in domestic law. In 1996, following the incident, China publicly pledged not to "provide assistance to unsafeguarded nuclear facilities." This promise built on China's 1992 NPT obligations by expanding them to cover dual-use nuclear items or any nonnuclear goods to unsafeguarded facilities in Pakistan or other countries. These pledges were followed by the promulgation of nuclear export control laws that incorporate the Nuclear Supplier Group (NSG) trigger lists.³⁶⁴ China's first nuclear export control law was issued and published in 1997 and a second one, specifically covering dual-use nuclear goods, was released in June 1998. The latter law importantly includes a "catchall" clause to stop any and all dual-use nuclear exports not specifically mentioned in the regulations; this step even goes beyond the NSG restrictions on dual-use exports. (See Appendix III.) Since the early 1990s, China has also developed the bureaucratic infrastructure to help implement these commitments. The China Atomic Energy Agency or CAEA (Zhongguo Guojia Yuanzineng Jigou) in conjunction with MOFTEC and the Foreign Ministry have assumed responsibility for overseeing the nuclear export control process. Recent organizational changes in China have further bolstered this process. First, the CAEA was separated from the China National Nuclear Corporation which is China's main exporter of nuclear materials, equipment, and technologies; thus the CAEA is no longer subject to the direct pressure of the CNNC when making export control decisions. Second, the Foreign Ministry within the last two years established a Department of Arms Control and Disarmament Affairs (junkong si) under the directorship of one of China's most experienced arms control experts, Sha Zukang. This department has an entire division of some 10 experts devoted to Chinese nuclear affairs including nuclear exports and export control issues.

Although these bureaucratic changes represent a step in the right direction, concerns about Sino-Pakistani nuclear cooperation persist. First, the Chinese Government continues to have difficulty implementing and enforcing its nonproliferation commitments and nuclear export control laws. There are Chinese companies, usually small ones, that either do not know the government's laws or that disregard them in an effort to earn hard currency. China's commercial nuclear ties to Pakistan are deep, which may facilitate continued nuclear-relevant exports. The ring-magnet incident in 1996 represented the first public instance of the continuing problem of how to promote respect in China for the government's international commitments and domestic laws. Until the central government is able to control the activities of these small, "rogue" firms, Chinese nuclear exports will remain an issue of concern for US policymakers.

Second, aside from illicit exports of dual-use equipment and materials, Chinese scientists and technicians may still be providing secret technical assistance to their Pakistani counterparts. Although China has adopted controls on exports of nuclear materials, equipment, and technologies, tracking and controlling technical exchanges by personnel is inherently difficult. Mutual visits by key scientists to weapons-related facilities in both China and Pakistan probably continue. In one instance, China's existing nuclear cooperation with Pakistan on the Chasma power reactor may provide a cover for exchanges related to Pakistan's construction, operation, and maintenance of unsafeguarded facilities.³⁶⁵

China and Missile Proliferation³⁶⁶

In the last 10 years, China's exports of ballistic and cruise missiles and related technologies have undergone an evolution similar to, but not as dramatic as, the reduction in China's nuclear exports. The geographic scope of China's missile exports has narrowed to include Iran, Pakistan, and, to a lesser extent, North Korea. The character of China's missile exports has shifted from sales of complete systems to exports of dual-use missile technologies. China also has assumed a growing number of missile nonproliferation commitments. In contrast to the nuclear area, however, many of them are vague, lacking legal basis, and poorly implemented. Significant concerns also persist about China's interpretations of its pledges. As of 1999, the principal US concern about China's missile proliferation revolves around the continued export by Chinese firms of dual-use missile technologies and production technologies to organizations in Pakistan, Iran, and North Korea that are involved in missile development.

Beginning in the late 1980s and ending in the early 1990s, China actively marketed and sold a variety of complete ballistic and cruise missiles to several countries. As early as 1986, China sold hundreds of HY-2 Silkworm and C-801/YJ-8 cruise missiles to Iran and Iraq; in Iran some of these systems were fitted on land-based batteries for coastal defense, and others were mounted on fast-attack crafts and used to threaten Persian Gulf shipping.³⁶⁷ China also provided Iran with production technologies to facilitate indigenous construction of these systems. As China's missile cooperation with Iran began to expand rapidly, China exported 30-35 DF-3 (CSS-2) intermediate range ballistic missiles to Saudi Arabia in 1988. These missiles, drawn from China's stock of aging missiles, possess a range of approximately 2,800 km that allowed Saudi Arabia--for the first time--to target most Middle East capitals.³⁶⁸

Chinese exports of complete missiles continued in the late 1980s when Chinese firms began to market and sell the newly developed M-9 and M-11 missiles. The M-9 and M-11 were developed specifically for export and were welcome additions to the international missile market in the late 1980s. These missiles, which are Chinese designed and solid fueled, were far more reliable and accurate than the majority of the Scud-derivatives available at that time. China negotiated with Pakistan, Iran, and Syria for the sale of both M-9s and M-11s. By late 1989, China and Syria reportedly signed a $285 million contract for approximately 30 M-9 missiles and launchers; the Syrians even provided advance funds for the missiles that the Chinese promptly spent before deliveries began.³⁶⁹ China and Iran had also engaged in extensive discussions about exports of M-9 missiles. One report indicated that by January 1990 China and Iran agreed on the export of M-9 missiles and production tooling, suggesting the possible sale of production technologies along with the full missiles.³⁷⁰ Other reports indicated Iran financially supported the M-9's development as Tehran is known to have done for North Korea's Nodong missiles.³⁷¹ There is little evidence to suggest that Iran was interested in the M-11 missile, however. China also began selling Iran a short-range, battlefield missile with a 150-km range; it was known as the 8610 or CSS-8. By 1989 China had sold some 150-200 of these systems to Iran and also had begun providing technologies for the creation of a production line to facilitate Iran's indigenous development of the 8610 system.³⁷² China's discussions with Pakistan focused on the possible supply of the M-11 missile. Sino-Pakistani negotiations proceeded quickly, and by 1990 China had transferred a training M-11 missile and launcher. A final shipment of 34 M-11s reportedly arrived in November 1992.³⁷³

In response to China's missile marketing, the United States actively sought, and in many cases succeeded, in curbing China's behavior. Reeling from the shock of China's DF-3 sale to Saudi Arabia and its perceived impact on Middle East stability, the Bush Administration immediately launched a vigorous effort to halt China's exports of M-9 and M-11 missiles. This campaign involved several rounds of bilateral discussions combined with the imposition in 1991 of limited economic sanctions for violations of the 1990 Missile Control Act. Finally in late 1991 and again in 1992, Chinese officials pledged verbally (and later in writing) that China would adhere to the guidelines and parameters of the MTCR. By assuming this commitment, China was forced to cancel the proposed sale of M-9 missiles to both Iran and Syria; this was especially difficult in the case of Syria because a contract had been signed and advance funds had been provided to Chinese firms. Neither the Iranian nor the Syrian deal went forward. The 1991/1992 MTCR commitment is particularly important in evaluating the changes in the scope, content, and frequency of China's missile export activities. Beijing's 1991 MTCR commitment provided a tangible, upper-bound limit on China's missile export activities. Since