Global Policy Forum

States parties to the Convention on Conventional Weapons (CCW) have spent most of 2008 developing a proposal to address the humanitarian impact of cluster munitions. At the third session of the CCW's Group of Governmental Experts in July 2008, states considered the draft text of a protocol to regulate cluster munitions circulated by chair Ambassador Bent Wigotski of Denmark. For Human Rights Watch, this draft protocol is "too little, too late." It is too little because it does not go far enough in tackling the challenges posed by cluster munitions and too late because it falls far short of the Convention on Cluster Munitions (CCM) that 107 states adopted on May 30, 2008. The CCM has been widely recognized as a milestone of international humanitarian law that will help end the civilian harm caused by cluster munitions. It prohibits use, production, transfer, and stockpiling of all cluster munitions. It also requires destruction of stockpiles within eight years and includes post-conflict remedial measures, such as victim assistance, clearance of submunitions within 10 years, and international cooperation and assistance. It will be opened for signature in Oslo, Norway on December 3, 2008.

By contrast, the draft CCW protocol merely attempts to regulate cluster munitions. It includes a transition period and does not represent an absolute ban; therefore, it permits continued use, ensuring that there will be more civilian casualties from these weapons. The draft protocol also contains several exemptions, meaning that its provisions apply only to certain kinds of cluster munitions. Articles 3 and 4, which lay out limited regulations for the use, production, and acquisition of cluster munitions, are particularly problematic. The protocol includes a few positive provisions, such as those on victim assistance, which are very similar to the CCM. Overall, however, the protocol is weak and unacceptable compared to the comprehensive convention.

Human Rights Watch urges all states to sign the CCM this December and to ratify it as soon as possible. States should not view the current CCW draft protocol as an alternative or supplement to the new convention. The draft protocol, which legitimizes the use of cluster munitions rather than prohibiting them outright, is contradictory to, not complementary with, the CCM, and any state that adopts the CCM should not support an instrument that contains a lower standard. States parties to the CCW have discussed cluster munitions for seven years while making virtually no progress on the issue. The current talks, which began after negotiations on the CCM had started, seem to be an effort to respond to the Oslo Process that produced the convention rather than to the humanitarian harm caused by cluster munitions. More than two-thirds of the world's producers, stockpilers, and users of the weapon participated in the ban treaty negotiations, but several major military powers—including China, India, Israel, Pakistan, Russia, and the United States—refused to do so. These states parties to the CCW are out of step with the rest of the world, and the draft protocol they are debating does not reflect the clear direction of the broader international community.

While Human Rights Watch believes that the CCW draft protocol is insufficient and unnecessary, we nonetheless are providing this commentary, which reviews the draft text article by article. The protocol's first paragraph says that states parties seek to "address the humanitarian impact caused by cluster munitions." Any state that truly wishes to do so should focus on the CCM and join the more than 100 adopting states in signing the standard-setting convention.

Article 1: General Provisions and Scope of Application

Beyond expressing the intent of the draft protocol, Article 1 describes its application. It says that it is "without prejudice" to other international humanitarian law that might include stricter obligations. Under Article 30 of the Vienna Convention on the Law of Treaties, when one treaty states that it should not be "considered as incompatible" with another, the rules of the latter prevail. In this case, the phrase "without prejudice" indicates that the protocol would not be considered incompatible with the CCM; therefore, Human Rights Watch believes that the CCM's obligations trump those of the protocol for states party to both instruments. While the implicit recognition of the CCM is important, states could more clearly support the humanitarian goal of minimizing civilian harm by joining only the stronger CCM. Like the CCM, the draft protocol should also apply to "all circumstances," not just those covered by CCW amended Article 1(1-6).

Article 2: Definitions

This article defines the key terms of the draft protocol. Options A and B for the definition of cluster munition are similar to that of the CCM in their technical description of the weapon, but option A is preferable because it borrows widely accepted language from the CCM. To the protocol's mention of submunition, states should add the phrase "weighing less than 20 kilograms" to be consistent with the CCM. Despite its apparent similarity to the CCM, the simplicity of the protocol's definition hides a major flaw. After its basic definition of the weapon, the CCM states what weapons carrying submunitions are not cluster munitions. They are excluded from the definition because they do not function like cluster munitions, which have a broad area effect and leave large quantities of duds. The draft protocol contains no such exclusions but, in the more problematic Article 4, exempts some cluster munitions from the treaty's restrictions. In other words, it makes use of certain types of cluster munitions acceptable in certain situations. As a result, the regulations of the protocol do not apply to all cluster munitions. Human Rights Watch strongly supports the definition of cluster munition victims, which replicates the language in the CCM. Article 2 of the protocol adds definitions of military objective, civilian objects, concentration of civilians, feasible precautions, self-neutralization mechanism, and shelf-life, none of which would be necessary if the protocol's later articles were stronger.

Article 3: Protection of Civilians, Civilian Populations, and Civilian Objects

We find serious shortcomings with this article. The paragraphs that represent new law—3(7) and 3(9)—merely regulate the weapon and do not go far enough to minimize humanitarian harm. States should instead institute a complete ban on use that eliminates all dangers to civilians both during strikes and afterwards. While an absolute ban on cluster munitions best protects civilians, paragraph 7 of the draft protocol represents the most important step forward for the CCW instrument. It prohibits use of cluster munitions in concentrations of civilians. Human Rights Watch believes that attacks in or near populated areas should be presumed to violate existing international humanitarian law because they cannot distinguish between combatants and civilians. Paragraph 7 addresses that concern, at least with regard to casualties at the time of attack. It should, however, refer to attacks "near" as well as "within" concentrations of civilians.

Paragraph 9 copies Article 54 of Protocol I to the 1949 Geneva Conventions; however, its protections must be made stronger if they are to reduce the dud problem. Paragraph 9(1) prohibits attacks on "objects indispensable to the survival of the civilian population." It qualifies that phrase with "for the specific purpose of denying [the objects] for their sustenance value to the civilian population or to the adverse Party." Attacks on agricultural land for other purposes would be permissible; as a result, states could still blanket fields and farms with unreliable submunitions that cause harm to lives and livelihoods. Furthermore, if states are serious about minimizing the harm from duds, in the case of cluster munitions, they should not only broaden paragraph 9(1) but also eliminate the exceptions laid out in paragraphs 9(2) and 9(3). The former allows attacks on protected objects if they support troops; duds launched into those areas could endanger civilians when the soldiers are gone. The latter, which allows states to derogate from the provision in cases of "imperative military necessity," again weakens the prohibition; the CCM allows no derogations to its provisions. The rest of the article reiterates existing international humanitarian law. These paragraphs would not be necessary if stronger provisions specific to cluster munitions were laid out, but, in themselves, they are not problematic.

Article 4: General Prohibitions and Restrictions

Of the three options proposed under Article 4, we believe that only Option C is acceptable. It follows the CCM model, which best controls and stigmatizes the weapon by ensuring that all cluster munitions are prohibited. Option B is stronger than A, but states should reject both options for the following reasons. First, states must reject any type of transition period. Options A and B include a transition period during which states can continue to use, with certain limitations, any cluster munitions they choose, even outdated ones with high dud rates. Neither of these options is satisfactory. If states acknowledge that cluster munitions cause too much humanitarian harm, they cannot at the same time allow their continued use. Second, states must always keep in mind the dual dangers of cluster munitions. Option A focuses primarily on duds, ignoring the harm caused by the weapons' area effect. Paragraph 2(f) says that states parties must "undertake…to the extent possible to improve the accuracy of their cluster munitions." The paragraph does not specify a particular level of accuracy, however, and qualifies the provision with "to the extent possible." The only other provisions that address the area effect are paragraphs (4)(d) and (e), which list guidance systems and a yet-to-be determined number of submunitions as characteristics that cluster munitions must have after the end of the transition period. The article, at least at this point, does not require the characteristics to be cumulative, so it is possible a cluster munition might have neither of them. Paragraph 1 of Option B requires cluster munitions to meet criteria that seek to reduce harm both during attacks and afterwards. Nevertheless, its criteria are not in this draft mandated as cumulative, making it weaker than Option C.

Third, when addressing the dud problem, states must avoid too much reliance on technical fixes, which have proved inadequate in the past. Both Options A and B give states the choice of employing self-destruct, self-deactivation, or self-neutralization devices instead of requiring redundant fail-safe mechanisms. As shown by Israel's use of the M85 submunition in Lebanon in 2006, submunitions with only self-destruct devices can have high dud rates and cause civilian casualties. The draft protocol's alternative requirement that submunitions have a certain reliability rate is insufficient because, as Human Rights Watch investigations and other field research have demonstrated, rates determined by testing do not accurately reflect rates during combat. Fourth, states must create an instrument with strong obligations. Option A is replete with qualifiers such as "as soon as feasible," "to the extent possible," and "in a position to do so," to name a few. Paragraph 2, which focuses on state conduct during the transition period, requires states to "undertake" to do what it lists; the protocol would be stronger if it used "shall." As a result, states parties could avoid most of the provisions laid out in the article, at least during the transition period. Fifth, states should not retain cluster munitions for training purposes, something both Options A and B allow. Although the CCM has a similar provision, Human Rights Watch strongly opposed its inclusion. Deminers do not train on live submunitions, and retention of any such weapons only makes the munitions vulnerable to operational use. While much preferable to Options A and B, Option C would be stronger if paragraph 1, as discussed above, was placed in the definition article and paragraph 2 became its own article. Option C should also add effects-based language akin to that of the CCM. The equivalent provision of the CCM states that the cumulative criteria that appear in Option C are designed "to avoid indiscriminate area effects and the risks posed by unexploded submunitions." The final language of Article 4 affects the application of several other articles, making it even more critical that states strengthen it to meet the standard of the CCM.

Article 5: Storage and Destruction

For this article to be effective, states must accept Option C to Article 4; only that choice will ensure Article 5 applies to the broadest range of cluster munitions, encompassing all those banned by the CCM. States should establish a deadline for stockpile destruction that is eight years or less to meet the standard set by the CCM. States should not allow an extension to the stockpile destruction deadline. The Mine Ban Treaty does not permit such an extension, and that fact has induced many states to expedite their destruction process. If states determine an extension provision is absolutely necessary, they should frame it in such a way that it is only available to states in extraordinary circumstances (i.e., extremely large stockpiles, involving extreme costs for destruction) and should be for a minimum period that is not renewable. They should also require stockpilers requesting extensions to the destruction deadline to provide detailed information, such as the quantity and type of submunitions destroyed and to be destroyed, the duration of the proposed extension, and details about how destruction will be completed.

Article 6: Transfers

As with Article 5, to make this article effective and applicable to the broadest range of cluster munitions, states must accept Option C to Article 4. Otherwise many types of cluster munitions will be exempted. In addition, the only acceptable transfers should be those "for the purpose of destruction."

Article 7: Clearance and Destruction

Article 7 has some good points but overall should be stronger and more detailed, as in the CCM. We welcome paragraph 1, which requires user states to provide assistance to affected states. It draws on Article 3(1) of CCW Protocol V on Explosive Remnants of War and strengthens it a bit by excising the phrase "where feasible." Following the precedent set by CCM Article 4(4), however, states should explicitly apply this provision to the use of cluster munitions that predate the treaty. In addition, as with stockpile destruction, states seeking an extension for clearance should provide detailed information about the status of their progress and their schedule for completion, among other things.

Article 8: Recording, Retaining, and Transmission of Information

Human Rights Watch supports, in theory, the requirement that states parties record and retain information about cluster munition use and that they make such information available to affected states. Such an information transfer greatly facilitates clearance and saves civilian lives. Paragraph 1 of this article should not be necessary, however, because states should prohibit all future use and therefore not have information to record and retain. Paragraph 2 should be preserved, but it should make explicit that it requires transmission of information on use prior to the entry into force of the protocol in order to minimize civilian casualties from existing unexploded submunitions.

Article 9: Protection of Humanitarian Missions and Organizations from the Effects of Cluster Munitions No comment.

Article 10: Victim Assistance

Human Rights Watch strongly supports this article, which almost exactly replicates that of the CCM.

Article 11: Co-operation and Assistance

Human Rights Watch welcomes the inclusion of an article on cooperation and assistance, and its provisions are generally in line with those of the CCM. States should consider the details of the article in comparison to Article 6 of the CCM. For example, paragraph 3 should specify that victim assistance is age and gender sensitive and includes medical care, rehabilitation, and psychological support.

Article 12: Consultations of High Contracting Powers

This article should explicitly allow for the participation of the United Nations, international and regional organizations, the International Committee of the Red Cross, the International Federation of Red Cross and Red Crescent Societies, and relevant nongovernmental organizations as observers to the consultations.

Article 13: Compliance

No comment.

Technical Annex: Best Practices for Testing of Cluster Munitions

There should be no need for the Technical Annex, because the protocol should ban cluster munitions completely and make testing irrelevant. Furthermore, as the annex itself recognizes, testing results "will not always be perfectly representative of the facts on the ground." As Human Rights Watch investigations and other field research have demonstrated, the tested failure rate of a submunition is not an adequate measure of performance in combat conditions and, as explained above, should not be used as a standard for determining if submunitions should be allowed. If states decide they must have such an annex, it should be a binding rather than best practices provision.