There has been some controversy as to the status of riot control agents under the Chemical Weapons Convention. In fact, however, the status of these agents is clear from the Convention's text. Every "riot control agent", as defined in Art. II(7), is also a "toxic chemical", as defined in Art. II(2). Thus, riot control agents are subject to the General Purpose Criterion of Article II(1)(a), which applies by its terms to "toxic chemicals." Under that provision, riot control agents are "chemical weapons" "except where intended for purposes not prohibited under [the] Convention, as long as the types and quantities are consistent with such purposes." The definition of "purposes not prohibited under [the] Convention" appears in Art. II(9); it includes both II(9)(c), "military purposes not connected with the use of chemical weapons and not dependent on the use of the toxic properties of chemicals as a method of warfare", and Art. II(9)(d), "law enforcement purposes including domestic riot control." If toxic chemicals, including riot control agents, are produced, held or transferred for purposes other than these permitted purposes, or in types or quantities inconsistent with those purposes, then they are "chemical weapons" and are subject to the Convention's declaration and destruction requirements. For instance, a stockpile of 105mm howitzer shells loaded with a riot control agent would be a "chemical weapon" and be subject to the Convention's restrictions. But, although riot control agents can be chemical weapons under certain circumstances, virtually all presently intended United States uses of riot control agents are permitted by the Convention.
A number of questions of interpretation may arise with respect to the Convention's provisions on riot control agents and other toxic chemicals. The following statements are intended as guidelines to help resolve such problems.
1. The term "law enforcement" in Art. II(9)(d) means actions taken within the scope of a nation's "jurisdiction to enforce" its national law, as that term is understood in international law. When such actions are taken in the context of law enforcement or riot control functions under the authority of the United Nations, they must be specifically authorized by that organization. No act is one of "law enforcement" if it otherwise would be prohibited as a "method of warfare" under Art. II(9)(c).
2. The uses of toxic chemicals prohibited as "methods of warfare" include any use of toxic chemicals by virtue of their toxic properties against enemy combatants (whether regulars or irregulars), and any use of toxic chemicals by virtue of their toxic properties against noncombatants if designed to advance a specific military objective in war.
3.A toxic chemical used by virtue of its toxic properties is only of a type consistent with the purpose of law enforcement, in the sense of Article II(1)(a), if it meets the Convention's definition of a "riot control agent" in Article II(7).
Thus, such chemicals must be "not listed in a Schedule" and must "produce rapidly in humans sensory irritation or disabling physical effects which disappear within a short time following termination of exposure." However, any chemical not on Schedule I may be used in carrying out the sentence of a duly constituted tribunal against a natural person.
I. Applicability of the General Purpose Criterion to Riot Control Agents.
Some officials have suggested that the Convention's only restriction on "riot control agents" is that they not be used as a "method of warfare", Art. 1(5). If this claim were correct, a state could develop, produce, retain, and transfer riot control agents in any form and in any quantity, so long as it did not actually use them as a method of warfare. A large class of toxic chemicals would be exempted from the Convention's controls by this interpretation: in the words of Art. II(7), it would include all chemicals "not listed in a Schedule, which can produce rapidly in humans sensory irritation or disabling physical effects which disappear within a short time following exposure." With respect to this group of chemicals, the situation would be little changed from that which existed under the majority interpretation of the 1925 Geneva Protocol, which similarly barred the use of chemical weapons but did not prohibit their development, production, and transfer. For example, this interpretation would permit a State to produce, maintain and test a large stockpile of howitzer shells or bombs loaded with a riot control agent, even though riot control agents of such a type (howitzer shells or bombs) and in such a quantity (a stockpile) would have no application other than as a method of warfare.
The claim underlying this interpretation derives in part from Article I.5 of the Convention, which obliges states "not to use riot control agents as a method of warfare." The only explanation for this provision, it is claimed, is that use of riot control agents as a method of warfare is not barred by any other provision of the Convention.
This explanation is incorrect both as a matter of history and as a matter of treaty interpretation. As a matter of history, there is a better explanation of Article I(5). The provision's authors almost certainly intended it to resolve the question of the international legal status of the use of riot control agents in warfare, a question raised by the position of the United States that the use of such agents in warfare was entirely outside the 1925 Geneva Protocol. In order to ensure that a similar question did not arise with respect to the Chemical Weapons Convention, the Convention's authors reiterated a rule that was already implicit in the Convention's text.
As a matter of treaty interpretation, this claim would exempt riot control agents from the Convention's central provision, the "general purpose criterion" of Article II(1)(a). That provision defines as "chemical weapons" subject to the Convention's restrictions "toxic chemicals and their precursors, except where intended for purposes not prohibited under this Convention, as long as the types and quantities are consistent with such purposes." The Convention also includes, in Article II(9), a specific and exclusive listing of "purposes not prohibited under this Convention", which includes, for example, agricultural, industrial, and medical purposes.
Given the General Purpose Criterion's central role in the Convention, it would be surprising to find that the Convention exempted certain categories of toxic chemical from its terms altogether. And, in fact, it did not. The General Purpose Criterion states clearly that it applies to all "toxic chemicals." The Convention's definition of "toxic chemicals", in turn, specifically includes chemicals that cause "temporary incapacitation." Compare this to the definition of "riot control agents" in Article II(7):
"any chemical not listed in a Schedule, which can produce rapidly in humans sensory irritation or disabling physical effects which disappear within a short time following termination of exposure." All chemicals that produce "sensory irritation or disabling physical effects" necessarily produce "temporary incapacitation." Thus, "riot control agents" are by definition a subset of "toxic chemicals", and so are subject to the General Purpose Criterion, including its restrictions as to types and quantities of chemicals. The specific inclusion of "law enforcement and domestic riot control purposes" (emphasis added) in Article II(9)'s list of "purposes not prohibited by the Convention" further underscores the fact that riot control agents are subject to the General Purpose Criterion.
An interpretation of the Convention that excluded riot control agents from the "types and quantities" restrictions of the General Purpose Criterion would also be deeply inconsistent with the purposes of the Convention. As a matter of treaty interpretation, this inconsistency weighs strongly against the validity of the proposed interpretation. As observed above, the proposed interpretation would exempt riot control agents from the Convention's restrictions on development, production, acquisition, stockpiling, retention and transfer, as well as its restrictions as to "types and quantities." Moreover, because "riot control agents" could never be "chemical weapons", they would be exempted from most of the Convention's declaration requirements, which are framed in terms of chemical weapons (Art. III (1) (a) - (d)).
Because this interpretation exempts riot control agents from so many of the Convention's core provisions, it poses the danger of a partial return to the mutual suspicion and security dilemmas that drove chemical weapons acquisition before the Convention. By eroding the norm against the possession of toxic agents for use in warfare, and by encouraging national military establishments to include the use of toxic chemicals in their training and doctrines, this situation might in turn permit a gradual return to the development and production of a broader range of toxic chemicals for military purposes.
To illustrate the application of the Convention's provisions on riot control agents, consider the example, introduced above, of a stockpile of howitzer shells loaded with a toxic chemical meeting Art. II(7)'s definition of a riot control agent. Because a riot control agent causes temporary incapacitation, it is a "toxic chemical" in the sense of Art. II(2). Thus, by the terms of Art. II(1)(a), it is subject to the general purpose criterion articulated in that section. Under the general purpose criterion, a toxic chemical is a "chemical weapon" unless it is "intended for a purpose not prohibited under this Convention, as long as the types and quantities are consistent with such purposes." Turning to the listing of purposes not prohibited by the Convention in Art. II(9), we may proceed through the four purposes not prohibited by the Convention.
The first, "industrial, agricultural, research, medical, pharmaceutical or other peaceful purposes", does not appear to encompass a stockpile of howitzer shells. The second, "protective purposes, namely those purposes directly related to protection against toxic chemicals and to protection against chemical weapons", might conceivably permit a small number of howitzer shells intended for use in training troops to operate in an environment containing toxic chemicals, but would not permit a stockpile of such shells. The third, "military purposes not connected with the use of chemical weapons and not dependent on the use of the toxic properties of chemicals as a method of warfare", appears to be inapplicable. Training, as discussed above, would not require a stockpile of shells, and there is no obvious other military application for a stockpile of howitzer shells that would not amount to use as a method of warfare. Finally, there appears to be no legitimate application of howitzer shells for "law enforcement including domestic riot control purposes", the last of the purposes not prohibited under the Convention.
Thus, a state possessing a stockpile of shells of this type would be required to treat them as chemical weapons, and they would be subject to the Convention's prohibitions on production, acquisition, retention, use, and transfer, and to the Convention's provisions on declaration and destruction, if still in existence when the Convention enters into force.
II. "Law Enforcement": What Law, Where Enforced?
The Convention does not state clearly what law States may enforce when they invoke the Convention's law enforcement provisions, or where and under what circumstances they may enforce that law. This creates considerable scope for confusion. For instance, may a State use toxic chemicals in enforcing its national law in the territory of another State, without the latter State's consent? May it invoke this provision whenever it judges international law has been violated? Further, the Convention does not clarify the relationship between its limited exception for military purposes and its exception for law enforcement purposes. Might an activity that would be barred if considered under the military purposes exception qualify as law enforcement, and so be permitted? For instance, could a State use riot control agents in fighting a civil war, claiming that enemy forces were violating its national law, international law, or both? Paragraph 1 of the proposed guidelines attempts to address these questions by providing reasonably clear standards for when and where States may claim to be engaging in law enforcement. It provides:
The term "law enforcement" in Art. II(9)(d) means actions taken within the scope of a nation's "jurisdiction to enforce" its national law, as that term is understood in international law. When such actions are taken in the context of law enforcement or riot control functions under the authority of the United Nations, they must be specifically authorized by that organization. No act is one of "law enforcement" if it otherwise would be prohibited as a "method of warfare" under Art. II(9)(c).
A. Enforcement of National Law. The proposed guidelines for the enforcement of national law are derived from background rules of international law. "Enforcement" of national law is a term of art in international law; there are reasonably well-defined rules as to the circumstances in which a state has jurisdiction to enforce its national law. Therefore, according to the rules of treaty interpretation, the Convention's references to law "enforcement" should be interpreted in light of this international standard. For example, international law provides that a state may enforce its own law within its own territory or against its own citizens, but that it may not enforce its national law in the territory of another State without that State's consent. The proposed guidelines make clear that the same rules apply to the use of toxic chemicals in law enforcement.
International legal rules also govern States' authority to enforce their laws in areas beyond any State's territorial jurisdiction, as on the high seas, in contested areas, and in the air. For instance, the only State that may ordinarily enforce its laws against a vessel is that vessel's flag State. Again, the same rule applies to the use of toxic chemicals in law enforcement. In a small number of instances, the precise international legal rules governing nations' jurisdiction to enforce national law in areas outside any state's territorial jurisdiction are in some dispute. In those cases, the rules governing nations' ability to use toxic chemicals in law enforcement will be clarified as broader questions as to the scope of national enforcement jurisdiction are resolved.
B. Enforcement of International Law. The Convention also does not state explicitly what sources of law States may enforce in invoking Art. II(9)(c). It seems possible, therefore, that States might wish to invoke international law to justify their "law enforcement" activities. The purposes of the Convention and background norms of international law itself demonstrate that such claims should be permitted in only the narrowest of circumstances. Permitting States to judge and enforce violations of international law themselves, without the consent of the world community, would be an invitation to anarchy. Moreover, States are prohibited by the United Nations Charter from unilaterally using force except in cases of self-defense against an armed attack. The Charter bars states from pursuing international disputes in a manner which "endanger[s] the maintenance of international peace and security", and instead requires them to settle their disputes by "peaceful means." Instead, the Charter provides that both the Security Council and the General Assembly shall have a role in the settlement of international legal disputes. Although the Charter reserves to States the right to use force in self-defense, acts of self-defense are not acts taken to enforce international law. Instead, they are acts of warfare, and are regulated by the Convention's provision on "military purposes not connected with the use of chemical weapons and not dependent on the use of the toxic properties of chemicals as a method of warfare."
Thus, the proposed guidelines contain only a very narrow provision on enforcement of international law. This provision applies only to States acting in the context of "law enforcement or riot control operations under the authority of the United Nations", and, in recognition of the United Nations' ultimate authority over such operations, requires that the United Nations grant "specific" approval for the use of toxic chemicals. The requirement that the approval be "specific" is intended to prevent States from invoking broad statements appearing in a resolution of a United Nations body, such as "all necessary means", as permission to act under this provision.
C. Relationship between "Law Enforcement" and "Methods of Warfare." Article II(9)(c) of the Convention defines the circumstances in which the use of toxic chemicals for military purposes is not prohibited as "military purposes not connected with the use of chemical weapons and not dependent on the use of the toxic properties of chemicals as a method of warfare" (emphasis added). It seems clear that the restriction on using the toxic properties of chemicals as a method of warfare applies with equal force to actions under Art. II(9)(d), the "law enforcement" provision. If it did not, States could circumvent the restriction on using the toxic properties of chemicals as a method of warfare by merely asserting that they were enforcing their national law, for instance in fighting a war on their own territory. Such an outcome would be entirely inconsistent with the purposes of the Convention. For this reason, the proposed guidelines state that "no act is one of 'law enforcement' if it otherwise would be prohibited as a 'method of warfare' under Art. II(9)(c)."
III. The Uses of Toxic Chemicals That Constitute "Methods of Warfare"
The Convention does not prohibit the use of toxic chemicals for "military purposes not connected with the use of chemical weapons and not dependent on the use of the toxic properties of chemicals as a method of warfare." States might claim that they should be permitted to use toxic chemicals, particularly riot control agents, in a range of situations under this provision. The most prominent examples of such applications, are, in the words of United States Executive Order 11850, use "in defensive military modes to save lives." It is not immediately clear when uses justified in such terms would amount to use as a "method of warfare."
The proposed guidelines attempt to answer this question, or at least to clarify the Convention's standard. They provide that:
The uses of toxic chemicals prohibited as "methods of warfare" include any use of toxic chemicals by virtue of their toxic properties against enemy combatants (whether regulars or irregulars), and any use of toxic chemicals by virtue of their toxic properties against noncombatants if designed to advance a specific military objective in war.
This guideline applies different standards to uses of toxic chemicals on combatants and on noncombatants. Uses on enemy combatants are barred outright, irrespective of purpose; uses on noncombatants are barred when they are "designed to advance a specific military objective in war." The standard for uses on noncombatants merely recasts the Convention's "method of warfare" language in terms of the purpose for which the toxic chemicals are used, an approach that occurs throughout the Convention.
Uses on enemy combatants are, however, barred irrespective of purpose. This is so because uses of toxic chemicals on enemy combatants will virtually always be designed to advance some military objective. Even if the State using toxic chemicals intends in part to save the lives of enemy combatants, it also intends to harass or immobilize them, a military objective, and so is using the toxic properties of chemicals as a method of warfare. As to the vanishingly small class of cases in which the objective of harassment might be absent, it is appropriate to set aside these cases in order to avoid encouraging States to make self-serving claims about their intentions towards enemy combatants. It may seem paradoxical that uses on combatants should be held to a higher standard than uses on civilians. In fact, such a difference in standards is not especially anomalous; for instance, states are held to a higher standard in a number of respects in their treatment of prisoners of war than of their own civilians. Note that the distinction between combatants and noncombatants is neither unusually difficult to apply nor new to international law. It is drawn from the First and Second Protocols Additional to the Geneva Conventions.
Since Executive Order 11850's promulgation in 1974, much of the controversy on these questions has focused on the particular situations it cites as permissible military uses of riot control agents. Thus, it is useful to consider the way in which 2 of the proposed guidelines would apply in those situations.
A. "Use of riot control agents in riot control situations in areas under direct and distinct [national] military control, to include controlling rioting prisoners of war."
Such uses of riot control agents would generally be permitted under the Convention and proposed guidelines. First, it should be noted that, as discussed in I above, this use of riot control agents is subject to the General Purpose Criterion, and so must be for one of the purposes listed in Art. II(9). In fact, two of II(9)'s clauses apply. Art. II(9)(c)'s "military purposes" provision applies directly. The prisoners of war are noncombatants, and so the use of riot control agents on them is permissible under 2 of the proposed guidelines unless it is "designed to advance a specific military objective in war." The military objective at stake in controlling rioting prisoners of war is, at best, highly attenuated, so this use is permissible.
Alternately, this use of riot control agents might be justified under Art. II(9)(d)'s "law enforcement" provision. Assuming that the State using riot control agents is operating on its own territory, or is an occupying power, or has secured the consent of the State in whose territory it is operating, it has jurisdiction to enforce its national law. Of course, it may not do so in a way that would amount to use as a "method of warfare" (which, as discussed above, this application does not).
B. "Use of riot control agents in situations in which civilians are used to mask or screen attacks, in the hope of reducing or avoiding civilian casualties"
Such uses of riot control agents would be prohibited. The proposed guidelines very clearly bar the use of all toxic chemicals, including riot control agents, on enemy combatants. Removing civilian cover from military forces, whether or not done for humanitarian reasons, serves a military objective, just as would denying the enemy another form of cover or a tactical advantage.
Situations of the type described above are very rare in any case. Civilians ordinarily avoid combat situations when possible. Moreover, civilians disabled by toxic chemicals will have difficulty finding shelter or otherwise protecting themselves from fire, calling into question the humanitarian benefits of this policy. Finally, a right to use toxic chemicals for "humanitarian" purposes may be misused, either deliberately or inadvertently (thanks to the "fog of war"), posing a significant danger of escalation.
C. "Use of riot control agents in rescue missions in remotely isolated areas, to rescue downed aircrews and passengers or escaping prisoners."
Such action is permitted in some situations and prohibited in others. Again, there is a separate analysis for "law enforcement" and for "military purposes." To begin with "law enforcement", if the "remotely isolated areas" are within the territorial jurisdiction of the State using riot control agents, or within the jurisdiction of a consenting State, then the conditions of 1 of the proposed guidelines are satisfied. Of course, the enforcement action may not depend on the use of the toxic properties of chemicals as a method of warfare, as defined further in 2.
If the "remotely isolated areas" are within the jurisdiction of an unconsenting State, then it is possible that limited use of riot control agents might be permitted under the "military purpose" provision, Art. II(9)(c). Rescuing downed aircrews and passengers and recapturing escaping prisoners by using riot control agents on noncombatants has a broad "military purpose", but is not "designed to advance" any more concrete "military objective." Provided that no riot control agents are used against enemy combatants, their use against noncombatants in such situations may be permitted by the Convention.
D. "Use of riot control agents in rear echelon areas outside the zone of immediate combat to protect convoys from civil disturbances, terrorists, and paramilitary organizations."
Once again, addressing the legality of this use requires an investigation of both the "military purposes" and the "law enforcement" provisions. Each exception seems likely to permit such use in some circumstances. The "law enforcement" exception applies, as in C., above, when the act of enforcement is taken within the nation's jurisdiction to enforce, and when the act does not constitute a "method of warfare."
Under either the "law enforcement" or the "military purposes" provision, we then conduct an inquiry into whether the act involved constitutes a "method of warfare." If the terrorists or paramilitary forces involved are acting as part of a "war", either civil or between States, then they are "combatants", and any use of riot control agents in this manner is prohibited. (It is difficult to define a "war" precisely, but the term generally refers to hostilities on a comparatively large scale between forces with at least some internal organization or discipline.) Otherwise, they are noncombatants, and we must ask whether the use of riot control agents in question is "designed to advance a specific military objective in war." In the case at hand, protecting convoys does have a military objective (maintaining supply lines), but it is sufficiently attenuated to render this use of riot control agents permissible.
IV. Types of Toxic Chemicals Permissible for "Law Enforcement." Although Art. II(9)(d) states that "law enforcement, including domestic riot control" is a "purpose not prohibited by the Convention", it does not state in detail which "toxic chemicals" may be used for this purpose. The Convention does provide one explicit restriction: the exclusion of "law enforcement" purposes from the list of permitted applications of Schedule I chemicals in Verification Annex VI(A)(2) implies that Schedule I chemicals may not be used for "law enforcement" purposes. However, States might assert that the Convention's lack of other restrictions permitted them to develop, produce, acquire, transfer and use for anti-personnel purposes any other toxic chemicals, including chemicals listed on one of the other Schedules, provided that they could claim that they were doing so for "law enforcement purposes."
This claim would be both harmful and incorrect. It would be harmful because it is difficult to monitor States' invocation of the Convention's provision on "law enforcement purposes." The Convention's section on declarations imposes only the weakest of requirements on chemicals intended for law enforcement purposes. Under Article III(1)(e), states must declare the chemical identity of all toxic chemicals that they hold "for riot control purposes." If a State decides that it intends to use a toxic chemical for law enforcement purposes, but not for purposes of riot control, it need not make any declarations with respect to that chemical at all.
The threat of this apparent loophole is particularly acute because a number of States currently have active development programs directed at producing new, technologically advanced disabling chemicals, many of which might be claimed to be intended for law enforcement purposes, but not for riot control. These chemicals are unlikely to be listed on a Schedule; thus, the general purpose criterion is the principal restriction on their development, production, ownership and use. Because States need not disclose the identity of these chemicals, however, it is likely to be difficult or impossible to subject States' claims as to their "purposes" to international scrutiny. States will inevitably assume that other States' claims as to their "purposes" are self- serving and question their sincerity, giving rise to precisely the sort of self-reinforcing cycle of mutual suspicion that the Convention was intended to eliminate. Further, developing countries, many of which may find these technologically advanced chemicals out of their reach, are likely to see the Convention's failure to regulate them as evidence of an unfair double standard, harming the legitimacy of the Convention in their eyes.
Yet these harmful consequences all flow from an incorrect interpretation of the Convention. Certainly, it cannot be the case that every state is permitted to judge for itself whether a particular "purpose" is one "not prohibited by the Convention." If that were so, the general purpose criterion would have little force: States could construe medical, industrial, or law enforcement purposes very broadly, and so evade the Convention's restrictions altogether. Rather, as the General Purpose Criterion states, toxic chemicals "intended for purposes not prohibited under [the] Convention" are not chemical weapons only "as long as the types and quantities are consistent with such purposes." Paragraph 3 of the proposed guidelines attempts to provide a focus for discussions that might resolve this problem. It reads:
A toxic chemical used by virtue of its toxic properties is only of a type consistent with the purpose of law enforcement, in the sense of Article II(1)(a), if it meets the Convention's definition of a "riot control agent" in Article II(7). Thus, such chemicals must be "not listed in a Schedule" and "produce rapidly in humans sensory irritation or disabling physical effects which disappear within a short time following termination of exposure." However, any chemical not on Schedule I may be used in carrying out the sentence of a duly constituted tribunal against a natural person.
Under this proposed guideline, an irritant or paralytic chemical with effects that do not endure beyond a short time would be of a "type" suited to law enforcement purposes, but sarin (which is listed in a Schedule) or any paralytic chemical with long- term side effects would not. Any chemical not on Schedule I could be used to carry out a death sentence imposed by a duly constituted tribunal. The requirement that the sentence of the tribunal be imposed on a natural person bars invocation of this provision to justify the use of toxic chemicals against a national group or ethnic minority.
This interpretation is derived from the purposes of the Convention and from international human rights standards, both recognized instruments of treaty interpretation. Most importantly, the Convention's purposes clearly require that States be precluded from retaining large quantities of toxic chemicals with enduring toxic effects with no other justification than asserted law enforcement purposes, particularly if the chemical identities of these toxic chemicals are unknown to the other parties to the Convention.
Further, applicable human rights standards support the interpretation in 3 of the proposed guidelines. Customary international law rules prohibit the use of chemicals with long-term effects by states on civilian populations, whether or not such use is characterized as "law enforcement", except when such use is in accordance with the decision of a duly constituted tribunal. More generally, the Universal Declaration of Human Rights and Covenant on Civil and Political Rights both bar governments from depriving citizens of their liberty other than after a full and fair trial. The Universal Declaration of Human Rights is widely agreed to have largely passed into customary international law, and the Covenant on Civil and Political Rights has over 100 States Parties and is also strong evidence for international custom. In light of these international human rights standards, the Chemical Weapons Convention's references to "law enforcement" must be understood as entailing some minimal guarantees of fair process. Accordingly, the proposed guidelines embody such a standard.
9 June 1994