United States of America
Practice Relating to Rule 75. Riot Control Agents
The US Field Manual (1956) states:
It is the position of the United States that the Geneva [Gas] Protocol of 1925 does not prohibit the use in war of … riot control agents, which are those agents of a type widely used by governments for law enforcement purposes because they produce, in all but the most unusual circumstances, merely transient effects that disappear within minutes after exposure to the agent has terminated. In this connection, however, the United States has unilaterally renounced, as a matter of national policy, certain uses in war of … riot control agents. The policy and provisions of Executive Order No. 11850 do not, however, prohibit or restrict the use of … riot control agents by US armed forces either (1) as retaliation in kind during armed conflict or (2) in situations when the United States is not engaged in armed conflict. Any use in armed conflict of … riot control agents, however, requires Presidential approval in advance.
The US Rules of Engagement for Vietnam (1971) stated:
Riot control agents will be used to the maximum extent possible. CS agents can be effectively employed in inhabited and urban area operations to flush enemy personnel from buildings and fortified positions, thus increasing the enemy’s vulnerability to allied firepower while reducing the unnecessary danger to civilians and the likelihood of destruction of civilian property.
The US Air Force Pamphlet (1976) restates Executive Order No. 11850 of 8 April 1975 and specifies:
The legal effect of this Executive Order is to reflect national policy. It is not intended to interpret the Geneva [Gas] Protocol of 1925 or change the interpretation of the United States that the Protocol does not restrain the use of riot control agents as such.
The US Air Force Commander’s Handbook (1980) states:
The United States does not regard the Geneva [Gas] Protocol as forbidding use of riot control agents … in armed conflict. However, the United States has, as a matter of national policy, renounced the first use of riot control agents … with certain limited exceptions specified in Executive Order 11850, 8 April 1975. Using … riot control agents … in armed conflict requires Presidential approval.
The US Operational Law Handbook (1993) states:
The following measures are expressly prohibited by the law of war and are not excusable on the basis of military necessity:
i. Using weapons which cause unnecessary suffering, prolonged damage to the natural environment, or poison weapons. This prohibition does not preclude the use of herbicides or riot control agents by US forces in wartime when authorized by the President of the US or his delegate.
The US Naval Handbook (1995) states:
The United States considers that use of riot control agents in armed conflict was not prohibited by the 1925 [Geneva] Gas Protocol. However, the United States formally renounced first use of riot control agents in armed conflict except in defensive military modes to save lives. Uses of riot control agents in time of armed conflict which the United States considers not to be violative of the 1925 [Geneva] Gas Protocol include:
1. Riot control situations in areas under effective U.S. military control, to include control of rioting prisoners of war.
2. Situations in which civilians are used to mask or screen attacks and civilian casualties can be reduced or avoided.
3. Rescue missions involving downed aircrews or escaping prisoners of war.
4. Protection of military supply depots, military convoys, and other military activities in rear echelon areas from civil disturbances, terrorist activities, or paramilitary operations.
Such employment of riot control agents by U.S. forces in armed conflict requires NCA approval.
Use of riot control agents as a “method of warfare” is prohibited by the 1993 Chemical Weapons Convention. However, that term is not defined by the Convention. The United States considers that this prohibition applies in international as well as internal armed conflict but that it does not apply in normal peacekeeping operations, law enforcement operations, humanitarian and disaster relief operations, counter-terrorist and hostage rescue operations, and non-combatant rescue operations conducted outside of such conflicts.
The United States also considers that it is permissible to use riot control agents against other than combatants in areas under direct U.S. military control, including to control rioting prisoners of war and to protect convoys from civil disturbances, terrorists and paramilitary organizations in rear areas outside the zone of immediate combat.
The US Naval Handbook (2007) states:
10.3.2 Riot Control Agents (RCAs)
The Chemical Weapons Convention defines RCAs as any chemical not listed in a schedule that can produce rapidly in humans sensory irritation or disabling physical effects that disappear within a short time following termination of exposure. States agreed not to use RCAs as a “method of warfare”; however, the Convention does not define that term. The United States ratified the Chemical Weapons Convention subject to the understanding that nothing in the Convention prohibited the use of RCAs in accordance with Executive Order 11850.
10.3.2.1 Riot Control Agents in Armed Conflict
Under Executive Order 11850, [“]Renunciation of certain uses in war of chemical herbicides and riot control agents[“], the United States renounced the first use of riot control agents in armed conflict except in defensive military modes to save lives, in situations such as:
1. Riot control situations in areas under effective U.S. military control, to include control of rioting prisoners of war
2. Situations in which civilians are used to mask or screen attacks and civilian casualties can be reduced or avoided
3. Rescue missions involving downed aircrews or escaping prisoners or war
4. Protection of military supply depots, military convoys, and other military activities in rear echelon areas from civil disturbances, terrorist activities, or paramilitary operations.
Such employment of riot control agents by U.S. forces in armed conflict requires presidential approval.
The United States considers that the prohibition on the use of RCAs as a “method of warfare” applies in international and internal armed conflict.
The US Manual on Detainee Operations (2008) states regarding the use of riot-control agents for maintaining order in detention facilities for persons deprived of liberty within the context of an armed conflict: “The use of riot control agents, including OC [oleoresin capsicum] spray, requires specific authorizations by competent authority. Less than lethal weapons have the ability to inflict significant harm to detainees.”
The manual also states within Appendix A on “Detention Facility Riot Control Measures”:
Plans, SOPs [standard operating procedures], and other directives must establish the procedures for authorization and use of lethal and non-lethal force, including riot control agents. Use of riot control agents must be approved for use by appropriate authority IAW [in accordance with] applicable rules for the UOF [use of force].
In 1927, during a debate in the US Senate, an argument against ratification of the 1925 Geneva Gas Protocol was that it outlawed the use of tear gas.
In 1931, during the League of Nations Preparatory Commission for the Disarmament Conference, the US representative, with respect to a memorandum submitted by the United Kingdom, stated:
While lachrymatory gases may serve some useful military purpose, for instance as harassing agencies, it is doubtless well-known to all my colleagues that the greatest use of lachrymatory gas is found, not in military service, but in police work either for controlling mobs, in which use it is certainly far more humane and probably more effective than the use of machine guns, sabres, or even truncheons, or it serves the purpose of effecting the capture of a barricaded criminal without bloodshed or loss of life … I think there would be considerable hesitation on the part of many Governments to bind themselves to refrain from the use in war, against an enemy, of agencies which they have adopted for peace-time use against their own population, agencies adopted on the ground that, while causing temporary inconvenience, they cause no real suffering or permanent disability, and are thereby more clearly humane than the use of weapons to which they were formerly obliged to resort to in times of emergency.
In 1969, during a debate in the First Committee of the UN General Assembly on the question of chemical and bacteriological (biological) weapons, the US representative stated with respect to the then still draft Resolution 2603 (XXIV):
41. … We do not agree with the interpretation which this resolution would place upon international law as embodied in the [1925 Geneva Gas Protocol]. I note that for the last forty years States have recognized the ambiguity of the Geneva Protocol, as to whether it prohibits the use of riot-control agents. They have not been able to resolve this ambiguity, despite several efforts to do so, and here we must respectfully differ with the Swedish delegation with regard to the conclusive – or we would say “inconclusive” – character of the negotiations leading up to the abortive Disarmament Conference of 1933. For if, as [the Swedish delegation] said … of the Geneva Protocol, “States did not doubt the comprehensive nature of the ban”, one must then ask why in the years after 1925 they continued to debate it.
43. We have examined in detail the negotiating histories of the 1899 and 1907 Hague Conventions, the Treaty of Versailles of 1919, the 1922 Washington Treaty, which never entered into force, and the 1925 Geneva [Gas] Protocol, and we have come to the conclusion that the negotiating histories of these treaties support the view that riot-control agents are not covered by the Geneva Protocol, and that, accordingly, [the draft resolution which became UN General Assembly resolution 2603 (XXIV)] incorrectly interprets the generally recognized rules of international law as embodied in the Geneva Protocol.
Executive Order No. 11850, issued by the US President on 8 April 1975, states:
The United States renounces, as a matter of national policy, … first use of riot control agents in war except in defensive military modes to save lives such as:
a) Use of riot control agents in riot control situations in areas under direct and distinct U.S. military control, to include controlling rioting prisoners of war.
b) Use of riot control agents in situations in which civilians are used to mask or screen attacks and civilian casualties can be reduced or avoided.
c) Use of riot control agents in rescue missions in remotely isolated areas, of downed aircrews and passengers, and escaping prisoners.
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.
Section 1. The Secretary of Defense shall take all necessary measures to ensure that the use by the Armed Forces of the United States of any riot control agents … in war is prohibited unless such use has Presidential approval, in advance.
Various sources observed that riot control agents were used in the Vietnam War by the United States and South Vietnamese forces.
In some circumstances, tear gas was allegedly used in conjunction with fragmentation bombs.
An article in a Swedish newspaper stated that VX gas was used against the North Vietnamese army in Cambodia.
At the CDDH, the United States stated, with regard to the asphyxiating, poisonous or other gases: “Opinions differed as to whether tear gas was covered by the Geneva Protocol of 1925.”
In 1980, in a memorandum of law on the “Reported Use of Chemical Agents in Afghanistan, Laos, and Kampuchea”, a legal adviser of the US Department of State stressed:
Although the United States does not regard the prohibition [on first use of chemical weapons] as applying to riot control agents, this view is not shared by the great majority of states (including the Soviets), and they would presumably regard themselves as being entitled to use chemical agents (including lethal agents) in response to use of riot control agents against them.
In 1994, the US President transmitted to the US Senate the findings of his administration’s review of the impact of the 1993 Chemical Weapons Convention on Executive Order No. 11850 concerning US policy on the use of riot control agents in armed conflict. The accompanying message of the President stated:
Article I(5) of the [Chemical Weapons Convention (CWC)] prohibits Parties from using [riot control agents (RCAs)] as a “method of warfare”. That phrase is not defined in the CWC. The United States interprets this provision to mean that:
–The CWC applies only to the use of RCAs in international or internal armed conflict. Other peacetime uses of RCAs, such as normal peacekeeping operations, law enforcement operations, humanitarian and disaster relief operations, counter-terrorist and hostage rescue operations, and non-combatant rescue operations conducted outside such conflicts are unaffected by the Convention.
–The CWC does not apply to all uses of RCAs in time of armed conflict. Use of RCAs solely against noncombatants for law enforcement, riot control, or other noncombatant purposes would not be considered as a “method of warfare” and therefore would not be prohibited. Accordingly, the CWC does not prohibit the use of RCAs in riot control situations in areas under direct U.S. military control, including against rioting prisoners of war, and to protect convoys from civil disturbances, terrorists, and paramilitary organizations in rear areas outside the zone of immediate combat.
–The CWC does prohibit the use of RCAs solely against combatants. In addition, according to the current international understanding, the CWC’s prohibition on the use of RCAs as a “method of warfare” also precludes the use of RCAs even for humanitarian purposes in situations where combatants and noncombatants are intermingled, such as the rescue of downed air crews, passengers, and escaping prisoners and situations where civilians are being used to mask or screen attacks. However, were the international understanding of this issue to change, the United States would not consider itself bound by this position.
In 1996, during hearings on the 1993 Chemical Weapons Convention before the US Senate’s Foreign Relations Committee, the US Secretary of Defense stated:
The [Chemical Weapons Convention (CWC)] does not prohibit the use of [Riot Control Agents (RCAs)] in riot control situations in areas under direct and distinct US military control, to include controlling rioting prisoners of war, and in rear echelon areas outside the zone of immediate combat to protect convoys from civil disturbance, terrorist and paramilitary organizations. The CWC does prohibit the use of RCAs solely against combatants and, according to the understanding of our allies and treaty signatories, even for humanitarian purposes in situations where combatants and non-combatants are intermingled.
At the same hearing, the Joint Staff Director of Strategic Plans and Policy stated: “In peacekeeping operations under Chapter six, Chapter seven UN operations, of course, the provisions of this convention don’t apply, and we would be able to use riot control agents … It’s my understanding that we could use riot control agents in Bosnia.”
In 1998, in a legal review of Oleoresin Capsicum (OC) pepper spray, the Deputy Assistant Judge Advocate General of the US Department of the Navy stated:
Oleoresin Capsicum is not calculated (i.e., designed), nor does it in fact cause unnecessary suffering. It is designed specifically to temporarily incapacitate violent or threatening subjects while reducing human suffering and is in consonance with the [Department of Defense Non-Lethal Weapon] program. Its physiological effects, while relatively painful, are temporary and do not rise to the level of unnecessary suffering contemplated in the prohibition … Provided a military necessity justifies its employment, the principle of unnecessary suffering would not preclude employment of OC in appropriate circumstances.
The OC system contemplated for acquisition and employment by the Marine Corps is specifically designed to limit its effects only to intended targets. The contemplated OC dispersers utilize a target specific stream of ballistic droplets for controlled delivery and minimal cross contamination (i.e., point target delivery), rather than an aerosolized spray which increases the likelihood of unintended subject impact. Provided the weapon is employed in a discriminating manner, the principle of distinction/discrimination presents no prohibition to acquisition and employment of OC in appropriate circumstances.
The second major category of chemicals regulated by the [Chemical Weapons Convention (CWC)] is Riot Control Agents [RCAs] …
While the proscriptions imposed by the CWC on chemical weapons are stated as absolute, the Convention seems to permit employment of RCAs, provided they are not used as a method of warfare. The CWC does not address whether a given substance can be subject to both the restrictions placed on RCAs and those placed on chemical weapons. Subsequent analysis in this memorandum concludes that RCAs are only constrained by the method of warfare restriction, that is, the CWC Treaty establishes a regime for treatment of RCAs separate from the regime dealing with chemical weapons.
… The definition of toxic chemicals
[of the Chemical Weapons Convention] appears broad enough to include many, if not all, RCAs
. Specifically, the use of the term temporary incapacitation
in the definition of toxic chemical
is difficult to distinguish from the term disabling effect
used in the definition of RCAs
. Thus, some contend that RCAs
fall under the CWC’s definition of toxic chemical
. If that is the case, then RCAs become subject to the CWC’s chemical weapon regime as well as the RCA regime. The consequences of such an interpretation are significant. RCAs would then be a chemical weapon
, subject to all the limitations applicable to such weapons, unless they were used for a purpose not prohibited
. This is problematic and would have a major impact on the use of RCAs since the purposes not prohibited
exclusion for use of chemical weapons
is an enumerated and apparently exclusive list of four activities only. Alternatively, if the CWC provides for a regime for RCAs
separate than that for chemical weapons
, then the only limitation on their use is that they may not be employed as a method of warfare
[emphasis in original]
In a footnote on this point, the Deputy Assistant Judge Advocate General stated: “If RCAs were subject to the chemical weapons regime, then the only ‘purpose not prohibited’ that would permit employment of RCAs is article II(9)(d) [of the 1993 Chemical Weapons Convention], the law enforcement exclusion”.
However, Deputy Assistant Judge Advocate General went on to state:
It is apparent … that the nature of the harm caused by RCAs is generally much less severe and that the toxic effects of RCAs are transient. Thus, it is clear from the definition of RCAs that the CWC envisages RCAs to be a relatively benign category of chemicals. The fact that the definition excludes those chemicals listed on Chemical Annex Schedules, many of which are extremely toxic, bolsters this point. While RCAs may well be toxic chemicals, in establishing a separate regime for a particular category of toxic chemicals, RCAs, the CWC has limited the boundaries of this category by narrowly defining the chemicals that qualify as RCAs.
Turning to the 1925 Geneva Gas Protocol, the Deputy Assistant Judge Advocate General stated:
Disagreement swirled around the Protocol’s coverage of RCAs. Since the 1960s, the U.S. has maintained that the Protocol applies only to lethal and incapacitating chemical agents and not to RCAs. The U.S. therefore maintained that RCAs could be used during armed conflict. That view was not universally shared in the international community. The United States’ extensive use of RCAs during the Vietnam War brought the differing interpretations to light. As a matter of national policy, however, the U.S., upon ratifying the Protocol in 1975, renounced the first use of RCAs in war except in defensive military modes to save lives. Nonetheless, the U.S. maintained that RCAs were not chemical weapons covered by the Protocol.
… Some nations, however, expressed concern that “RCAs would constitute an immediate risk and danger if they were allowed to develop into a new generation of non-lethal but effective chemical agents of warfare, causing insurmountable problems in trying to distinguish between ‘real’ and ‘non-lethal’ chemical weapons on the battlefield, as well as ‘real’ and ‘non-lethal’ chemical warfare units.” The result was a compromise in which the U.S. accepted the CWCs Article I (5) prohibition on the use of RCAs as a “method of warfare” in exchange for their categorization outside the chemical weapon regime.
The phrase method of warfare
is not defined in the CWC or in the negotiating record and has been the subject of significant debate in the United States. The Administration view is that United States Armed Forces must be involved in an armed conflict, either international or non-international, to engage in a method of warfare
[emphasis in original]
With respect to Executive Order No. 11850, issued by the US President on 8 April 1975, the Deputy Assistant Judge Advocate General stated:
U.S. ratification of the Chemical Weapons Convention … created a debate regarding the continuing efficacy of [Executive Order] 11850, particularly exceptions (b) and (c) … If a use of RCAs constitutes a “method of warfare” then the CWC prohibits such use as a U.S. treaty obligation under international law. The executive order, however, authorizes use of RCAs, in war in certain situations. Though not explicitly stated, the apparent intent of the Executive Order permits RCA employment against combatants in war in situations like those enumerated in exceptions (b) and (c). Although the CWC does not define the phrase method of warfare, the apparent intent seems to prohibit the uses of RCAs contemplated in exceptions (b) and (c) to [Executive Order] 11850.
This review reiterates that the continuing efficacy of [Executive Order] 11850 is currently an issue of debate. The draft instruction [i.e., the draft of the Chairman of the Joint Chiefs of Staff Instruction (CJCSI) 3110.07A, Nuclear, Biological, and Chemical (NBC) Defence; Riot Control Agents; and Herbicides Annual Review, of 1 March 1998] and its list of permissible uses of RCAs is, however, currently the U.S. military position. Should appropriate U.S. Government authority determine that [Executive Order] 11850 is no longer valid authority, such a decision would only impact the use of RCAs in war … when the U.S. is a party to the conflict. All other uses of RCAs listed in the draft instruction would remain unaffected.
[emphasis in original]
In 1998, a US Department of Defense document discussing the use of chemical agents in the Vietnam War stated that the “use of tear gas, or Riot Control Agents (RCA) as they were sometimes called, was in accordance with US policy at the time”.
In April 2003, following the outbreak of the Iraq War one month earlier, the US Department of Defense held a press briefing to explain the legal concepts behind the US treatment of EPWs (Enemy Prisoners of War) in Iraq. In response to a question about whether the use of tear gas would be a violation of the 1993 Chemical Weapons Convention, the Special Assistant to the Army JAG responded:
The 1993 Chemical Weapons Convention prohibits the use of riot-control agents as a method of warfare. It’s not a precisely defined term. The United States has an executive order that suggests that riot-control agents can be used for defensive purposes to save lives. That’s a very long-standing executive order.
It gives a few examples in there. One is combat search and rescue. The others are rioting prisoners of war. A third example is, if in fact an enemy placed civilians in front of it, to advance on your lines. There is a very careful process for the decision as to whether or not riot control agents may be used on the battlefield, requiring presidential authorization, which may be delegated to the combatant commander. But it’s not something that we do lightly.