103. Initial response of U.S. to ICRC study on Customary International Humanitarian Law with Illustrative Comments (November 3, 2006)

November 3, 2006

Dr. Jakob Kellenberger
International Committee of the Red Cross
19 Avenue de la Paix
CH 1202 Geneva

Dear Dr. Kellenberger:

We write to provide the U.S. Government’s initial reactions to the ICRC’s recent study, entitled Customary International Humanitarian Law (the “Study”).

We welcome the Study’s discussion of this complex and important subject of the customary “international humanitarian law,” and we appreciate the major effort that the ICRC and the Study’s authors have made to assemble and analyze a substantial amount of material. We share the ICRC’s view that knowledge of the rules of customary international law is of use to all parties associated with armed conflict, including governments, those bearing arms, international organizations, and the ICRC. Although the Study uses the term “international humanitarian law,” we prefer the “law of war” or the “laws and customs of war.”[i]

Given the Study’s large scope, we have not yet been able to complete a detailed review of its conclusions. We recognize that a significant number of the rules set forth in the Study are applicable in international armed conflict because they have achieved universal status, either as a matter of treaty law or – as with many provisions derived from the Hague Regulations of 1907 – customary law. Nonetheless, it is important to make clear – both to you and to the greater international community – that, based upon our review thus far, we are concerned about the methodology used to ascertain rules and about whether the authors have proffered sufficient facts and evidence to support those rules. Accordingly, the United States is not in a position to accept without further analysis the Study’s conclusions that particular rules related to the laws and customs of war in fact reflect customary international law.

We will continue our review and expect to provide additional comments or otherwise make our views known in due course. In the meantime, we thought it would be constructive to outline some of our basic methodological concerns and, by examining a few of the rules set forth in the Study, to illustrate how these flaws call into question some of the Study’s conclusions.

This is not intended to suggest that each of our methodological concerns applies to each of the Study’s rules, or that we disagree with every single rule contained in the study -- particular rules or elements of those rules may well be applicable in the context of some categories of armed conflict. Rather, we hope to underline by our analysis the importance of stating rules of customary international law correctly and precisely, and of supporting conclusions that particular rules apply in international armed conflict, internal armed conflict, or both. For this reason, the specific analysis that follows this letter is in certain respects quite technical in its evaluation of both the proffered rule and the evidence that the Study uses to support the rule.

There is general agreement that customary international law develops from a general and consistent practice of States followed by them out of a sense of legal obligation, or opinio juris. Although it is appropriate for commentators to advance their views concerning particular areas of customary international law, it is ultimately the methodology and the underlying evidence on which commentators rely – which must in all events relate to State practice – that must be assessed in evaluating their conclusions.

State practice. Although the Study’s introduction describes what is generally an appropriate approach to assessing State practice, the Study frequently fails to apply this approach in a rigorous way.

  • First, for many rules proffered as rising to the level of customary international law, the State practice cited is insufficiently dense to meet the “extensive and virtually uniform” standard generally required to demonstrate the existence of a customary rule.
  • Second, we are troubled by the type of practice on which the Study has, in too many places, relied. Our initial review of the State practice volumes suggests that the Study places too much emphasis on written materials, such as military manuals and other guidelines published by States, as opposed to actual operational practice by States during armed conflict. Although manuals may provide important indications of State behavior and opinio juris, they cannot be a replacement for a meaningful assessment of operational State practice in connection with actual military operations. We also are troubled by the extent to which the Study relies on non-binding resolutions of the General Assembly, given that States may lend their support to a particular resolution, or determine not to break consensus in regard to such a resolution, for reasons having nothing to do with a belief that the propositions in it reflect customary international law.
  • Third, the Study gives undue weight to statements by non-governmental organizations and the ICRC itself, when those statements do not reflect whether a particular rule constitutes customary international law accepted by States.
  • Fourth, although the Study acknowledges in principle the significance of negative practice, especially among those States that remain non-parties to relevant treaties,2 that practice is in important instances given inadequate weight.
  • Finally, the Study often fails to pay due regard to the practice of specially affected States.3 A distinct but related point is that the Study tends to regard as equivalent the practice of States that have relatively little history of participation in armed conflict and the practice of States that have had a greater extent and depth of experience or that have otherwise had significant opportunities to develop a carefully considered military doctrine. The latter category of States, however, has typically contributed a significantly greater quantity and quality of practice.

Opinio juris. We also have concerns about the Study’s approach to the opinio juris requirement. In examining particular rules, the Study tends to merge the practice and opinio juris requirements into a single test. In the Study’s own words,

it proved very difficult and largely theoretical to strictly separate elements of practice and legal conviction. More often than not, one and the same act reflects both practice and legal conviction. . . . When there is sufficiently dense practice, an opinio juris is generally contained within that practice and, as a result, it is not usually necessary to demonstrate separately the existence of an opinio juris.4

We do not believe that this is an appropriate methodological approach. Although the same action may serve as evidence both of State practice and opinio juris, we do not agree that opinio juris simply can be inferred from practice. Both elements instead must be assessed separately in order to determine the presence of a norm of customary international law. For example, Additional Protocols I and II to the Geneva Conventions contain far-reaching provisions, but States did not at the time of their adoption believe that all of those instruments’ provisions reflected rules that already had crystallized into customary international law; indeed, many provisions were considered ground-breaking and gap-filling at the time. One therefore must be cautious in drawing conclusions as to opinio juris from the practice of States that are parties to conventions, since their actions often are taken pursuant to their treaty obligations, particularly inter se, and not in contemplation of independently binding customary international law norms.5 Even if one were to accept the merger of these distinct requirements, the Study fails to articulate or apply any test for determining when state practice is “sufficiently dense” so as to excuse the failure to substantiate opinio juris, and offers few examples of evidence that might even conceivably satisfy that burden.

We are troubled by the Study’s heavy reliance on military manuals. We do not agree that opinio juris has been established when the evidence of a State’s sense of legal obligation consists predominately of military manuals. Rather than indicating a position expressed out of a sense of a customary legal obligation, in the sense pertinent to customary international law, a State’s military manual often (properly) will recite requirements applicable to that State under treaties to which it is a party. Reliance on provisions of military manuals designed to implement treaty rules provides only weak evidence that those treaty rules apply as a matter of customary international law in non-treaty contexts. Moreover, States often include guidance in their military manuals for policy, rather than legal, reasons. For example, the United States long has stated that it will apply the rules in its manuals whether the conflict is characterized as international or non-international, but this clearly is not intended to indicate that it is bound to do so as a matter of law in non-international conflicts. Finally, the Study often fails to distinguish between military publications prepared informally solely for training or similar purposes and those prepared and approved as official government statements. This is notwithstanding the fact that some of the publications cited contain a disclaimer that they do not necessarily represent the official position of the government in question.

A more rigorous approach to establishing opinio juris is required. It is critical to establish by positive evidence, beyond mere recitations of existing treaty obligations or statements that as easily may reflect policy considerations as legal considerations, that States consider themselves legally obligated to follow the courses of action reflected in the rules. In this regard, the practice volumes generally fall far short of identifying the level of positive evidence of opinio juris that would be necessary to justify concluding that the rules advanced by the Study are part of customary international law and would apply to States even in the absence of a treaty obligation.

Formulation of rules. The Study contains several other flaws in the formulation of the rules and the commentary. Perhaps most important, the Study tends to over-simplify rules that are complex and nuanced. Thus, many rules are stated in a way that renders them overbroad or unconditional, even though State practice and treaty language on the issue reflect different, and sometimes substantially narrower, propositions. Although the Study’s commentary purports to explain and expand upon the specifics of binding customary international law, it sometimes does so by drawing upon non-binding recommendations in human rights instruments, without commenting on their non-binding nature, to fill perceived gaps in the customary law and to help interpret terms in the law of war. For this reason, the commentary often compounds rather than resolves the difficulties presented by the rules, and it would have been useful for the Study’s authors to articulate the weight they intended readers to give the commentary.

Implications. By focusing in greater detail on several specific rules, the attachment illustrates how the Study’s methodological flaws undermine the ability of States to rely, without further independent analysis, on the rules the Study proposes. These flaws also contribute to two more general errors in the Study that are of particular concern to the United States:

· First, the assertion that a significant number of rules contained in the Additional Protocols to the Geneva Conventions have achieved the status of customary international law applicable to all States, including with respect to a significant number of States (including the United States and a number of other States that have been involved in armed conflict since the Protocols entered into force) that have declined to become a party to those Protocols; and

· Second, the assertion that certain rules contained in the Geneva Conventions and the Additional Protocols have become binding as a matter of customary international law in internal armed conflict, notwithstanding the fact that there is little evidence in support of those propositions.

We would like to reiterate our appreciation for the ICRC’s continued efforts in this important area, and hope that the material provided in this letter and in the attachment will initiate a constructive, in-depth dialogue with the ICRC and others on the subject.


John B. Bellinger, III William J. Haynes II

Legal Adviser General Counsel

U.S. Department of State U.S. Department of Defense

[1] As the Study itself indicates, the field has traditionally been called the “laws and customs of war.” Accordingly, we will use this term, or the term “law of war,” throughout. J.-M. Haenckerts and L. Doswald-Beck, Customary International Humanitarian Law, Vol. I, p. xxv (Cambridge 2005) (hereinafter, “Study”).

2 Study, Vol. I, p. xliv (indicating that contrary practice by States not parties to treaties that contain provisions similar to the rule asserted “has been considered as important negative evidence”).

3 As the Study notes (Vol. I, p. xxxviii), the International Court of Justice has observed that “an indispensable requirement” of customary international law is that “State practice, including that of States whose interests are specially affected, should have been both extensive and virtually uniform in the sense of the provision invoked; -- and should moreover have occurred in such a way as to show a general recognition that a rule of law or legal obligation is involved.” North Sea Continental Shelf Cases (Federal Republic of Germany v. Denmark; Federal Republic of Germany v. Netherlands, [1969] I.C.J. 4, 43 (emphasis added). In this context, the Study asserts, this principle means that “[w]ith respect to any rule of international humanitarian law, countries that participated in an armed conflict are ‘specially affected’ when their practice examined for a certain rule was relevant to that armed conflict.” Study, Vol. I, p. xxxix. This rendering dilutes the rule and, furthermore, makes it unduly provisional. Not every State that has participated in an armed conflict is “specially affected”; such States do generate salient practice, but it is those States that have a distinctive history of participation that merit being regarded as “specially affected.” Moreover, those States are not simply “specially affected” when their practice has, in fact, been examined and found relevant by the ICRC. Instead, specially affected States generate practice that must be examined in order to reach an informed conclusion regarding the status of a potential rule. As one member of the Study’s Steering Committee has written, “The practice of ‘specially affected states’ – such as nuclear powers, other major military powers, and occupying and occupied states – which have a track record of statements, practice and policy, remains particularly telling.” Theodore Meron, The Continuing Role of Custom in the Formation of International Humanitarian Law, 90 Am. J. Int’l L. 238, 249 (1996).

4 Study, Vol. I, p. xl.

5 Even universal adherence to a treaty does not necessarily mean that the treaty’s provisions have become customary international law, since such adherence may have been motivated by the belief that, absent the treaty, no rule applied.

Illustrative Comments on Specific Rules in the Customary International Humanitarian Law Study[1]

Rule 31

Rule 31 states: “Humanitarian relief personnel must be respected and protected.”

The United States consistently has supported and facilitated relief efforts in armed conflicts around the world, and is keenly aware of the critical role humanitarian relief personnel (“HRP”) play in bringing food, clothing, and shelter to civilians suffering from the impact of such conflicts. It is clearly impermissible intentionally to direct attacks against HRP as long as such personnel are entitled to the protection given to civilians under the laws and customs of war.

Rule 31, however, sets forth a much broader proposition without sufficient evidence that it reflects customary international law. The Study fails to adduce a depth of operational State practice to support that rule. Had it examined recent practice, moreover, its discussion might have been more sensitive to the role of State consent regarding the presence of such personnel (absent a UN Security Council decision under Chapter VII of the UN Charter) and the loss of protection if such personnel engage in particular acts outside the terms of their mission. The Study summarily dismisses the role of State consent regarding the presence of HRP but fails to consider whether a number of the oral statements by States and organizations that it cites actually reflected situations in which HRP obtained consent and were acting consistent with their missions.[2] To be clear, these qualifications do not suggest that HRP who have failed to obtain the necessary consent, or who have exceeded their terms of mission short of taking part in hostilities, either in international or internal armed conflicts, may be attacked or abused. Rather, it would be appropriate for States to take measures to ensure that those HRP act to secure the necessary consent, conform their activities to their terms of mission, or withdraw from the State. Nevertheless, a proposition that fails to recognize these qualifications does not accurately reflect State practice and opinio juris.

Relevant treaty provisions. Treaty provisions on the treatment of HRP guide the current practice of many States, and clearly articulate limits to the obligation asserted by rule 31:

  • Article 71(1) of Additional Protocol I (“AP I”) requires that HRP obtain the consent of the State in which they intend to operate.[3] Article 71(4) prohibits HRP from exceeding the “terms of their mission” and permits a State to terminate their mission if they do so. Even Article 17(2) of AP I, which the Study cites in support of a State’s obligation to protect aid societies, describes a situation in which consent almost certainly would be present, since a State that appeals to an aid society for assistance effectively is providing advance consent for that society to enter its territory.
  • The Convention on the Safety of United Nations and Associated Personnel, which places an obligation on States Parties to take appropriate measures to ensure the safety and security of UN and associated personnel, applies to situations in which UN personnel are in the host State with the host State’s consent, since Article 4 requires the UN and the host State to conclude an agreement on the status of the UN operation.[4]
  • Article 12 of Amended Protocol II to the Convention on Conventional Weapons (“CCW Amended Protocol II”), which addresses States Parties’ obligations to protect certain humanitarian missions from the effects of mines and other devices, states that “this Article applies only to missions which are performing functions in an area with the consent of the High Contracting Party on whose territory the functions are performed.”[5] The Article continues, “Without prejudice to such privileges and immunities as they may enjoy . . . personnel participating in the forces and missions referred to in this Article shall: . . . refrain from any action or activity incompatible with the impartial and international nature of their duties.”[6]
  • The Fourth Geneva Convention likewise contains both a consent and a “terms of mission” requirement for HRP. Article 10 states that “[t]he provisions of the present Convention constitute no obstacle to the humanitarian activities which the [ICRC] or any other impartial humanitarian organization may, subject to the consent of the Parties to the conflict concerned, undertake for the protection of civilian persons and for their relief.”[7] (emphasis added). Article 9 of the First, Second, and Third Geneva Conventions contain virtually identical provisions.
  • Additional Protocol II (“AP II”) does not contain provisions relating directly to the acts of HRP themselves, but Article 18 states that relief actions require the consent of the High Contracting Party in whose territory the HRP may wish to operate and must be limited to actions of an “exclusively humanitarian and impartial nature.”
  • The Statute of the International Criminal Court (“Rome Statute”) includes as a war crime the act of “[i]ntentionally directing attacks against personnel [] involved in a humanitarian assistance or peacekeeping mission in accordance with the Charter of the United Nations, as long as they are entitled to the protection given to civilians or civilian objects under the international law of armed conflict.”[8] The Commentary to the ICC Statute states, with regard to this provision, that “[t]he humanitarian assistance should also receive the consent of the parties to the conflict the territory of which it must pass or in which it carries out its tasks.”[9]

Despite the fact that these treaties clearly qualify State obligations regarding HRP, rule 31 lacks any such qualifications. Because the practice of States Parties to treaties presumptively tracks their treaty prerogatives and obligations, we would expect that, to justify omission of these qualifications, the Study would have provided particularly strong evidence of State practice that was inconsistent with them. However, the Study simply concludes that “the overwhelming majority of practice does not specify this condition [of consent],” even after acknowledging that the protection of HRP under the Additional Protocols “applies only to ‘authorised’ humanitarian personnel as such.”[10]

The role of State consent. Much of the practice on which the Study bases its conclusion that State consent is irrelevant is ambiguous or off-point, and in any event, the Study’s analysis lacks sufficient attention to detail and context. For instance, peacekeeping implementation agreements such as those among parties to the conflict in Bosnia and Herzegovina, in which each side undertook to provide security assurances to the ICRC, may be seen as a grant of advance consent for the presence of ICRC personnel in the territory of each party.[11] (If the States objected to the presence of the ICRC, they would not have agreed to provide it with security assurances.) The Study relies on other examples of State discussions of the protection of HRP that specifically allude to the State’s support for the Geneva Conventions and their Additional Protocols;[12] as noted above, however, both the Geneva Conventions and the Additional Protocols reflect the need for HRP to obtain State consent.

The Study cites only seven military manuals, all from States Parties to AP I. The cited excerpts from these manuals offer no indication that these States reject the role of consent. Australia’s and France’s military manuals simply state that HRP are given special protection, but this does not explain the scope of and preconditions for a State’s obligations.[13] Only one State’s manual (Sweden’s) states the view that Article 71(2) has achieved the status of customary international law, and it is not clear from the excerpt whether Sweden believes that other paragraphs of Article 71 (including the consent provision in paragraph (1)) also are customary international law.[14] Indeed, the role of consent may be so commonly understood that States, in discussing this issue, simply assume that HRP will obtain it, particularly given the strong incentives for them to do so. As for many of the UN Security Council resolutions cited as State practice supporting rule 31, almost all of the peacekeeping operations from which these resolutions stem were established with the consent of the host governments or under the Security Council’s Chapter VII authority. Thus, although the resolutions may not themselves recite a condition of consent, consent almost always was a condition precedent – save in the case of Security Council action under Chapter VII, which is plainly an exceptional circumstance with respect to state sovereignty.[15]

Significant examples of the operational practice of States in this area – which were not included in the Study – are very different from that described by the Study in that they evidence the critical role of State consent. For example, the Civil Military Operations Center and the Humanitarian Operations Center, employed by U.S. and coalition forces in conflicts that include Bosnia, Kosovo, and Afghanistan, required humanitarian relief organizations to coordinate their movements with the coalition forces, in order for those forces to support the organizations’ efforts and to ensure their members’ safety.[16] Fuller consideration of operational practice undoubtedly would have provided the Study’s authors valuable, necessary information.

Terms of mission limitations. Rule 31 also disregards the obvious fact that HRP who commit acts that amount to direct participation in the conflict are acting inconsistent with their mission and civilian status and thus may forfeit protection. The Geneva Conventions and AP I both recognize, implicitly or explicitly, that during such time as a civilian takes direct part in hostilities, he or she may be targeted. As noted above, to support a rule that ignores the “terms of mission” condition, we would expect the Study to provide strong evidence of State practice that ignores States’ prerogatives under relevant treaties to provide protection only for HRP who are providing humanitarian relief. But the Study has not provided such evidence. The Study also fails to provide evidence of opinio juris regarding such practice.

Much of the practice cited in the Study actually supports the condition that HRP must work within the terms of their mission. For instance, Canada’s cited manual refers to the work of HRP themselves as protected, and, with regard to non-governmental organizations, notes that NGOs are to be respected “upon recognition that they are providing care to the sick and wounded.”[17] The Dutch manual uses the more precise term “personnel engaged in relief activities,” which may be read as reflecting the “terms of mission” requirement.[18] The Study cites the fact that India provides relief personnel the same protection as medical and religious personnel,[19] but the latter categories of personnel lose their protection from direct attack if they engage in acts harmful to the enemy or directly participate in hostilities. The Report on the Practice of Jordan states that Jordan has “always assumed [sic] the safety of those who are engaged in humanitarian action.”[20] This, too, fails to support the proposed rule, as it focuses on the actual work of HRP and is silent about the protections Jordan gives HRP who act outside their missions’ terms. Finally, the Study cites the EU Presidency as saying, “[D]uring armed conflicts, the security of humanitarian personnel was frequently not respected.”[21] The only reasonable conclusion to draw from this statement is that State practice is inconsistent with the described rule.

These limitations in treaty provisions, military manuals, and State practice are not inadvertent, but reflect a concerted distinction borne of legitimate State and military security concerns, making it very unlikely that States would acquiesce in the overbroad principle depicted in the rule. For example, during the 1982 Israeli incursion into Lebanon, Israel discovered ambulances marked with the Red Crescent, purportedly representing the Palestinian Red Crescent Society, carrying able-bodied enemy fighters and weapons. This misconduct reportedly was repeated during the 2002 seizure of Bethlehem’s Church of the Nativity by members of the terrorist al Aqsa Martyrs Brigade.[22] If the ambulance drivers in these examples were considered to be HRP and actually were helping fighters in a conflict, Israel would be precluded from taking action under rule 31 as written. Military commanders also have had to worry about individuals falsely claiming HRP status, as happened in Afghanistan when some members of Al Qaeda captured while fighting claimed to be working for a humanitarian relief organization. These examples demonstrate why States, in crafting treaty provisions on this topic, have created a “terms of mission” condition for HRP in a way that rule 31 fails to do.

Opinio juris. According to the Study, a number of States view themselves as having a legal obligation to protect HRP as a matter of customary international law. The meaning and soundness of certain cited examples are at best unclear, however. For instance, the Study cites Nigeria and Rwanda as asserting that they are legally obligated to protect HRP from the effects of military operations, even in the absence of a treaty obligation.[23] Without citations to the actual language, and without context, it is not clear whether these States were asserting that they took this view even in the absence of State consent and in situations in which HRP were acting outside their mission. The Study also quotes Zimbabwe’s submission that it regards the Geneva Conventions’ guarantees relating to the activities of relief personnel as part of customary international law, but, as noted above, those Conventions reflect the importance of State consent.[24] Finally, with regard to the Report on U.S. Practice stating that the United States believes that “unjustified attacks on international relief workers are also violations of international humanitarian law” (emphasis added), nothing in this statement undercuts the fact that matters may be different when HRP are acting as combatants, nor does it speak at all to the question of State consent.[25]

Non-international armed conflicts. Although the Study asserts that rule 31 applies in both international and non-international armed conflict, the Study provides very thin practice to support the extension of rule 31 to non-international armed conflicts, citing only two military manuals of States Parties to AP II and several broad statements made by countries such as the United Kingdom and United States to the effect that killing ICRC medical workers in a non-international armed conflict was “barbarous” and contrary to the provisions of the laws and customs of war.[26] The Study contains little discussion of actual operational practice in this area, with citations to a handful of ICRC archive documents in which non-state actors guaranteed the safety of ICRC personnel. Although AP II and customary international law rules that apply to civilians may provide protections for HRP in non-international armed conflicts, the Study offers almost no evidence that rule 31 as such properly describes the customary international law applicable in such conflicts.

Summary. We do not believe that rule 31, as drafted, reflects customary international law applicable to international or non-international armed conflicts. The rule does not reflect the important element of State consent or the fact that States’ obligations in this area extend only to HRP who are acting within the terms of their mission – that is, providing humanitarian relief. To the extent that the authors intended to imply a “terms of mission” requirement in the rule, the authors illustrated the difficulty of proposing rules of customary international law that have been simplified as compared to the corresponding treaty rules.

* * *

Rule 45

The first sentence of rule 45 states: “The use of methods or means of warfare that are intended, or may be expected, to cause widespread, long-term and severe damage to the natural environment is prohibited.”

Protection of the environment during armed conflict obviously is desirable as a matter of policy, for reasons that include issues of civilian health, economic welfare, and ecology. The following discussion should not be interpreted as opposing general consideration, when appropriate and as a matter of policy, of the possible environmental implications of an attack. Additionally, it is clear under the principle of discrimination that parts of the natural environment cannot be made the object of attack unless they constitute military objectives, as traditionally defined, and that parts of the natural environment may not be destroyed unless required by military necessity.

Nevertheless, the Study fails to demonstrate that rule 45, as stated, constitutes customary international law in international or non-international armed conflicts, either with regard to conventional weapons or nuclear weapons.[27] First, the Study fails to assess accurately the practice of specially affected States, which clearly have expressed their view that any obligations akin to those contained in rule 45 flow from treaty commitments, not from customary international law. (We disagree with the Study’s conclusion that France, the United Kingdom, and the United States are not among those specially affected with regard to environmental damage flowing from the use of conventional weapons, given the depth of practice of these States as a result of their participation in a significant proportion of major international armed conflicts and peacekeeping operations around the globe during the twentieth century and to the present.) Second, the Study misconstrues or overstates some of the State practice it cites. Third, the Study examines only limited operational practice in this area and draws flawed conclusions from it.

Specially affected States. The Study recognizes that the practice of specially affected States should weigh more heavily when assessing the density of State practice,[28] but fails to assess that practice carefully. France and the United States repeatedly have declared that Articles 35(3) and 55 of AP I, from which the Study derives the first sentence of rule 45, do not reflect customary international law. In their instruments of ratification of the 1980 CCW, both France and the United States asserted that the preambular paragraph in the CCW treaty, which refers to the substance of Articles 35(3) and 55, applied only to States that have accepted those articles.[29] The U.S. State Department Principal Deputy Legal Adviser stated during a conference in 1987 that the United States considered Articles 35 and 55 to be overly broad and ambiguous and “not a part of customary law.”[30] Rather than taking serious account of such submissions, the Study instead places weight on evidence of far less probative value. The U.S. Army JAG Corps Operational Law Handbook, which the Study cites for the proposition that the United States believes that the provision in rule 45 is binding, is simply an instructional publication and is not and was not intended to be an authoritative statement of U.S. policy and practice.[31] Nor is the U.S. Air Force Commander’s Handbook, which the Study also cites as authority.[32]

In addition to maintaining that Articles 35(3) and 55 are not customary international law with regard to the use of weapons generally, specially affected States possessing nuclear weapon capabilities have asserted repeatedly that these articles do not apply to the use of nuclear weapons. For instance, certain specially affected States such as the United States, the United Kingdom, Russia, and France so argued in submissions to the International Court of Justice (“ICJ”).[33] The United Kingdom’s military manual specifically excepts from the limitation in Article 35(3) the use of nuclear weapons against military objectives.[34] In a report summarizing the Conference that drafted Additional Protocol I, the United States noted:

During the course of the Conference there was no consideration of the issues raised by the use of nuclear weapons. Although there are several articles that could seem to raise questions with respect to the use of nuclear weapons, most clearly, article 55 on the protection of the natural environment, it was the understanding of the United States Delegation throughout the Conference that the rules to be developed were designed with a view to conventional weapons and their effects and that the new rules established by the Protocol were not intended to have any effects on, and do not regulate or prohibit the use of nuclear weapons. We made this understanding several times during the Conference, and it was also stated explicitly by the British and French Delegations. It was not contradicted by any delegation so far as we are aware.[35]

The Conference Record from 1974, reflecting earlier work on the text that became AP I, records the United Kingdom’s view on the issue: “[The UK] delegation also endorsed the ICRC’s view, expressed in the Introduction to the draft Protocols, that they were not intended to broach problems concerned with atomic, bacteriological or chemical warfare. . . . It was on the assumption that the draft Protocols would not affect those problems that the United Kingdom Government had worked and would continue to work towards final agreement on the Protocols.”[36] In acceding to AP I, both France and the United Kingdom stated that it continued to be their understanding that the Protocol did not apply generally to nuclear weapons. For instance, the United Kingdom stated, “It continues to be the understanding of the United Kingdom that the rules introduced by the Protocol apply exclusively to conventional weapons

. . . . In particular, the rules so introduced do not have any effect on and do not regulate or prohibit the use of nuclear weapons.”[37]

The Study’s summary states: “It appears that the United States is a ‘persistent objector’ to the first part of this rule. In addition, France, the United Kingdom and the United States are persistent objectors with regard to the application of the first part of this rule to the use of nuclear weapons.”[38] However, the weight of the evidence – including the fact that ICRC statements prior to and upon conclusion of the Diplomatic Conference acknowledged this as a limiting condition for promulgation of new rules at the Conference; that specially affected States lodged these objections from the time the rule first was articulated; and that these States have made them consistently since then – clearly indicates that these three States are not simply persistent objectors, but rather that the rule has not formed into a customary rule at all.

General evidence of State practice and opinio juris. Other practice included in the Study fails to support or undercuts the customary nature of rule 45. This includes examples of States consenting to the application of Articles 35(3) and 55;[39] a State expressing a concern that opposing forces were directing attacks against its chemical plants, without asserting that such attacks would be unlawful;[40] the ICJ indicating in 1996 that Article 35(3) constrained those States that subscribed to AP I, and thus indicating that the Article is not customary international law;[41] draft codes and guidelines issued by international organizations and not binding by their terms;[42] and statements that could just as easily be motivated by politics as by a sense of legal obligation. Some cited practice makes specific reference to a treaty as the basis for obligations in this area. In 1992, in a memorandum annexed to a letter to the Chairman of the Sixth Committee of the UN General Assembly, the United States and Jordan stated that Article 55 of AP I requires States Parties to “take care in warfare to protect the natural environment against widespread, long-term and severe damage.” That is, the United States and Jordan described the rule as a treaty-based, rather than customary, obligation.[43] Israel’s Practice Report, which states that Israeli Defense Forces do not use or condone methods or means of warfare that rule 45 covers, contains no suggestion that Israel has adopted this policy out of a sense of legal obligation.[44] With regard to the twenty State military manuals the Study cites (all but one of which are from States Parties to AP I), the Study offers no evidence that any of these nineteen States Parties included such a provision in their manuals out of a sense of opinio juris, rather than on the basis of a treaty obligation. In sum, none of the examples given clearly illustrates unequivocal support for the rule, either in the form of State practice or of opinio juris.

Domestic criminal laws. The Study lists various States’ domestic criminal laws on environmental damage, but some of those laws flow from the obligation in Article 85 of AP I to repress breaches of the Protocol. Certain other States’ laws criminalize a broad crime termed “ecocide,” but most of the cited provisions fail to make clear whether this crime would apply to acts taken in connection with the use of military force. As noted above, a number of States (including Australia, Burundi, Canada, Congo, Georgia, Germany, Netherlands, New Zealand, Trinidad, and the United Kingdom) have incorporated ICC Article 8(2)(b)(iv) into their criminal codes, but the ICC provision prohibits the use of the weapons described in rule 45 only in those cases in which their use “would be clearly excessive in relation to the concrete and direct overall military advantage anticipated.”[45] These domestic criminal provisions clearly do not support the broader statement in rule 45, which would preclude States from taking into account the principles of military necessity and proportionality. Finally, the Study offers almost no evidence that any of these States has enacted criminal laws prohibiting this activity out of a sense of opinio juris. The fact that a State recently criminalized an act does not necessarily indicate that the act previously was prohibited by customary international law; indeed, a State may have criminalized the act precisely because, prior to its criminalization in domestic law, it either was not banned or was inadequately regulated.

Operational practice. The Study examines only a limited number of recent examples of practice in military operations and draws from these examples the conclusion that “[p]ractice, as far as methods of warfare . . . are concerned, shows a widespread, representative and virtually uniform acceptance of the customary law nature of the rule found in Articles 35(3) and 55(1)” of AP I.[46] However, the cited examples are inapposite, as none exhibited the degree of environmental damage that would have brought rule 45 into play. Rather than drawing from that the conclusion that the underlying treaty provisions on which the rule is based are too broad and ambiguous to serve as a useful guideline for States, as the United States long has asserted, the Study assumes that the failure to violate the rule means that States believe it to be customary law. It is notable that, following Iraq’s attacks on Kuwait’s oil fields, most international criticism focused on the fact that these attacks violated the doctrines of military necessity and proportionality.[47] Most criticism did not assert potential violations of customary rules pertaining to environmental damage along the lines of rule 45.[48] The Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia noted that “it would appear extremely difficult to develop a prima facie case upon the basis of these provisions [of AP I], even assuming they were applicable.”[49] It may be the case that rule 45 as drafted, like the treaty provisions on which it is based, sets such a limited and imprecise boundary on action as not to function as a rule at all.

Non-international armed conflicts. For all of the reasons that the Study fails to offer sufficient evidence that the provision in rule 45 is a customary rule in international armed conflict, the Study fails to make an adequate case that the rule is customary international law applicable to non-international armed conflicts. (The Study itself acknowledges that the case that rule 45 would apply in non-international conflicts is weaker.[50]) The fact that a proposal by Australia to include a provision like Article 35(3) in AP II failed further undercuts the idea that rule 45 represents a rule of customary international law in non-international armed conflicts.[51]

Summary. States have many reasons to condemn environmental destruction, and many reasons to take environmental considerations into account when determining which military objectives to pursue. For the reasons stated, however, the Study has offered insufficient support for the conclusion that rule 45 is a rule of customary international law with regard to conventional or nuclear weapons, in either international or non-international armed conflict.

* * *

Rule 78

Rule 78 states: “The anti-personnel use of bullets which explode within the human body is prohibited.”

Although anti-personnel bullets designed specifically to explode within the human body clearly are illegal, and although weapons, including exploding bullets, may not be used to inflict unnecessary suffering, rule 78, as written, indicates a broader and less well-defined prohibition. The rule itself suffers from at least two problems. First, it fails to define which weapons are covered by the phrase “bullets which explode within the human body.” To the extent that the Study intends the rule to cover bullets that could, under some circumstances, explode in the human body (but were not designed to do so), State practice and the ICRC’s Commentary on the 1977 Additional Protocol reflect that States have not accepted that broad prohibition. Second, there are two types of exploding bullets. The first is a projectile designed to explode in the human body, which the United States agrees would be prohibited. The second is a high-explosive projectile designed primarily for anti-materiel purposes (not designed to explode in the human body), which may be employed for anti-materiel and anti-personnel purposes. Rule 78 fails to distinguish between the two. If, as the language suggests, the Study is asserting that there is a customary international law prohibition on the anti-personnel use of anti-material exploding bullets, the Study has disregarded key State practice in this area. Third, the Study extrapolates the rule to non-international conflicts without a basis for doing so.

Bullets covered. With regard to which weapons are covered by the phrase “bullets which explode within the human body,” the language in rule 78 appears to use an effects-based test, and in doing so fails to distinguish between projectiles that almost always detonate within the human body, including those specifically designed to do so; projectiles that foreseeably could detonate within the human body in their normal use; and projectiles that in isolated or rare instances outside their normal use might detonate within the human body. Although there are important practical differences among these types of munitions – and, more generally, between munitions designed to explode within the human body and those designed for other, lawful purposes – the language of the rule suggests that the Study considers all three categories in applying this effects-based test to be illegal. If so, there is no evidence that States have accepted this standard; if States have accepted a rule in this area, it is only with regard to the first category of projectiles – those designed to explode within the human body. Indeed, the Study concedes, “The military manuals or statements of several States consider only the anti-personnel use of such projectiles to be prohibited or only if they are designed to explode upon impact with the human body.”[52] The Study, however, ignores the significance of design in its formulation of rule 78.[53]

The ICRC put forward an effects-based standard at the Second CCW Review Conference in 2001, in proposing that CCW States Parties consider negotiating a protocol that would prohibit the anti-personnel use of bullets that explode within the human body. Although the Study notes the ICRC’s own submission to the Review Conference,[54] it fails to note that States Parties did not choose to pursue a protocol or other instrument on this issue. The ICRC proffered this same standard in the now-withdrawn “superfluous injury or unnecessary suffering” (“SIrUS”) project. Because of its use of this “effects-based” (rather than design-based) standard, the Study’s commentary also brings into the discussion certain weapons that we do not consider to fall within the category of bullets that explode within the human body. The statement in the commentary to rule 78 that “certain 12.7mm bullets exploded in human tissue stimulant” appears to be an effort to include in the category of bullets that explode within the human body the 12.7mm Raufoss multi-purpose ammunition.[55] The Study’s statement refers to a 1998 ICRC test that subsequently proved flawed in its methodology, results, and conclusions in a 1999 re-test at Thun, Switzerland, of which ICRC members were observers.[56] The published conclusions of the participants in the re-test did not support the ICRC conclusion that this ammunition should be considered to be the type that explodes in the human body, yet the Study does not mention this 1999 re-test.

Uses covered. The rule as written suggests a total ban on all instances in which exploding bullets may be used against personnel, but State practice does not support this. Efforts to restrict the use of exploding bullets date back to the 1868 Declaration Renouncing the Use, in Time of War, of Explosive Projectiles Under 400 Grammes Weight (the “St. Petersburg Declaration”).[57] This Declaration banned the use of exploding bullets in international armed conflict between the States Parties. Only seventeen government representatives, however, signed the St. Petersburg Declaration, with two other States, Baden and Brazil, acceding in 1869. Despite the Study’s assertion that the St. Petersburg Declaration represented the practice of “most of the States in existence at that time,”[58] it actually represented that of less than half of the States then in existence.[59] Furthermore, only one State has acceded to the St. Petersburg Declaration since 1869.[60]

Since the St. Petersburg Declaration, there has been considerable State practice involving the anti-personnel use of exploding bullets, despite the ICRC’s statement that governments have “adhered” to the Declaration. Two participants in the ICRC-hosted 1974 Lucerne Meeting of Experts on certain weapons conventional weapons concluded:

At present it is widely held that in view of the development in weapons

technology and state practice the St. Petersburg Declaration cannot be

interpreted literally, or in any case that it has not as such become declaratory

of customary international law…. [T]he prohibition contained in it serves to

illustrate the principle prohibiting the causing of unnecessary suffering, at

least as it was contemplated in 1868.[61]

U.S. legal reviews have detailed State practice contrary to the ICRC’s statement and consistent with the conclusion contained in the above quotation. The ICRC fails to cite this contrary practice in its summary of those U.S. legal reviews.[62] The 1923 Hague Draft Rules of Air Warfare (the “Air Rules”), which explicitly superseded the St. Petersburg Declaration with regard to explosive projectiles, established an exception to the broad ban on explosive bullets for explosive projectiles used “by or against an aircraft.”[63] Although the Study refers to the Air Rules, it does not note that this exception to the total ban on use of exploding bullets permits their use by aircraft without categorical target restrictions, i.e., permits such use for anti-material or anti-personnel use. Since States developed the Air Rules, States widely have employed bullets that may detonate on impact with materiel for both anti-materiel and anti-personnel purposes.[64] Such ammunition was in common use by all States that participated in World War II, and in conflicts thereafter – including in widespread aircraft strafing of enemy forces, a practice common to every conflict since World War I in which aircraft were employed. The considerable State practice involving the use of such anti-materiel weapons against forces are indications that rule 78’s apparently total prohibition on the anti-personnel use of exploding bullets does not reflect customary international law.

The practice the Study cites does not support a rule banning the use of exploding bullets against personnel in all circumstances. The Study includes in Volume II examples from the military manuals of eleven countries, only six of which contain unqualified bans on exploding bullets;[65] the legislation of six countries, only three of which provide additional support for the rule as stated;[66] statements made by several States at diplomatic conferences, most of which are ambiguous;[67] and the reported operational practice of only two States.[68] Among all these sources, at most two cite customary international law as the legal basis for regulations on the use of exploding bullets.[69] Even disregarding the existence of contrary State practice, this body of evidence is insufficient to establish the customary nature of the rule as stated.

The examples of operational practice adduced by the Study are particularly questionable. The Report on the Practice of Indonesia states only that exploding bullets are reported as prohibited in Indonesia, an unconfirmed example of State practice.[70] The Report on the Practice of Jordan states only that Jordan “does not use, manufacture or stockpile explosive bullets,” but does not state whether it does so out of a sense of legal obligation under customary or treaty law, or whether it simply chooses not to do so due to policy or practical concerns.[71] In general, the Study fails to recognize that different militaries have different requirements, and that a State may decide not to use exploding ammunition for military rather than legal reasons.

The only example of actual battlefield behavior cited by the Study in support of rule 78 is an accusation by the Supreme Command of the Yugoslav People’s Army (“JNA”) of the Socialist Federal Republic of Yugoslavia that Slovene forces used exploding bullets.[72] It is unclear whether the bullets were used by ground forces against other ground forces, by airplanes against personnel, or in some other way. Most important, due to the use of ellipses in the Study, it is unclear whether the alleged behavior by Slovene forces was criticized as being “prohibited under international law” due to the anti-personnel use of exploding bullets per se or, rather, criticized as being used against “members and their families” (emphasis added) – allegations that, if true, would state a violation of other tenets of international law. It is thus difficult to determine whether this example supports the broad rule postulated by the Study, or a narrower rule restricting certain anti-personnel uses of exploding bullets.

Non-international armed conflict. The Study also asserts that rule 78 is a norm of customary international law applicable in non-international armed conflicts. The Study, however, provides scant evidence to support this assertion. The St. Petersburg Declaration refers only to international armed conflict between States Party to the Declaration; the Declaration does not mention internal conflict. In fact, the Study’s only evidence of opinio juris in this regard is the failure, in military manuals and legislation cited previously, to distinguish between international and non-international armed conflict. Since governments normally employ, for practical reasons unrelated to legal obligations, the military ammunition available for international armed conflict when engaged in non-international armed conflict, and since there is ample history of the use of exploding bullets in international armed conflict, the Study’s claim that there is a customary law prohibition applicable in non-international armed conflict is not supported by examples of State practice. Furthermore, this analysis fails to account for the military manual of the UK, cited in the Study, which prohibits the use of exploding bullets directed solely at personnel only in international armed conflict.[73]

Summary. Virtually none of the evidence of practice cited in support of rule 78 represents operational practice; the Study ignores contrary practice; and the Study provides little evidence of relevant opinio juris. The evidence in the Study of restrictions on the use of exploding bullets supports various narrower rules, not the broad, unqualified rule proffered by the Study. Thus, the assertion that rule 78 represents customary international law applicable in international and non-international armed conflict is not tenable.

* * *

Rule 157

Rule 157 states: “States have the right to vest universal jurisdiction in their national courts over war crimes.”

Impunity for war criminals is a serious problem that the United States consistently has worked to alleviate. From the Second World War to the more recent crises in the former Yugoslavia and Rwanda, the United States has contributed substantially towards ensuring accountability for war crimes and other international crimes. Efforts to address the problem of accountability have, logically, focused on ensuring that there are appropriate fora to exercise jurisdiction over the most serious violations of international law.[74] One part of this solution is to ensure that those committing such offenses cannot find safe havens, by requiring States Parties to various treaties to reduce jurisdictional hurdles to their prosecution. For example, Article 146 of the Fourth Geneva Convention requires all States Parties to extradite or prosecute an individual suspected of a grave breach, even when a State lacks a direct connection to the crime. The Study, however, does not offer adequate support for the contention that rule 157, which is stated much more broadly, represents customary international law.

Clarity of the asserted rule. If rule 157 is meant to further the overall goal of the Study to “be helpful in reducing the uncertainties and the scope for argument inherent in the concept of customary international law,”[75] it must have a determinate meaning. The phrase “war crimes,” however, is an amorphous term used in different contexts to mean different things. The Study’s own definition of this term, laid out in rule 156, is unspecific about whether particular acts would fall within the definition. For the purpose of these comments, we assume that the “war crimes” referred to in rule 157 are intended to be those listed in the commentary to rule 156. These acts include grave breaches of the Geneva Conventions and AP I, other crimes prosecuted as “war crimes” after World War II and included in the Rome Statute, serious violations of Common Article 3 of the Geneva Conventions, and several acts deemed “war crimes” by “customary law developed since 1977,” some of which are included in the Rome Statute and some of which are not.[76]

Assuming this to be the intended scope of the rule, we believe there are at least three errors in the Study’s reasoning regarding its status as customary international law. First, the Study fails to acknowledge that most of the national legislation cited in support of the rule uses different definitions of the term “war crimes,” making State practice much more diverse than the Study acknowledges. Second, the State practice cited does not actually support the rule’s definition of universal jurisdiction. Whereas rule 157 envisions States claiming jurisdiction over actions with no relation to the State, many of the State laws actually cited invoke the passive or active personality principle, the protective principle, or a territorial connection to the act before that State may assert jurisdiction. Furthermore, the Study cites very little evidence of actual prosecutions of war crimes not connected to the forum state (as opposed to the mere adoption of legislation by the States).[77] Third, the Study conflates actions taken pursuant to treaty obligations with those taken out of a sense of customary legal obligation under customary international law. These errors undermine the Study’s conclusion that rule 157 constitutes customary international law.

Diverse understandings of “war crimes.” The national legislation cited in the commentary to rule 157 employs a variety of definitions of “war crimes,” only a few of which closely parallel the definition apparently employed by the Study, and none that matches it exactly.[78] Much of the legislation cited does not precisely define “war crimes” and therefore cannot be relied on to support the rule. Although the military manuals of Croatia, Hungary, and Switzerland, among others, appear to define “war crimes” as “grave breaches,” the lack of specificity leaves the intended meaning ambiguous.[79] Even among the few States that employ a definition of “war crimes” similar to that in rule 156, no State definition mirrors the Study’s definition precisely. Canada, for example, includes “grave breaches” of the Geneva Conventions and Additional Protocol I, violations of the Hague Convention, violations of “the customs of war,” and possibly certain violations of AP II, but, unlike the Study, does not specifically include “serious violations of Common Article 3 of the Geneva Conventions” in its definition. Furthermore, the Study does not assert that Canada’s conception of “violations of the customs of war” matches that of the Study.[80] It is therefore evident, simply by the diversity of definitions of “war crimes” employed by various States, that State practice does not support the contention that States, as a matter of customary international law, have the right to vest universal jurisdiction in their national courts over the full set of actions defined by the Study as “war crimes.”

Exercise of universal jurisdiction over only limited acts. Although the Study cites legislation from more than twenty States that supposedly demonstrates the customary nature of rule 157, not one State claims jurisdiction over all the acts cited in rule 156 as “war crimes” in the absence of a State connection to the act, whether it be territorial or based on the active personality, passive personality, or protective principles.[81] The domestic legislation of a number of States, including Australia, Belgium, Colombia, Cyprus, and Zimbabwe, only asserts universal jurisdiction over grave breaches of the Geneva Conventions and AP I.[82] Other domestic legislation is focused even more narrowly: the legislation of Barbados, Botswana, Singapore, and Uganda, for instance, only asserts universal jurisdiction over grave breaches of the Geneva Conventions.[83] Further, many of the military manuals cited (including those of Belgium, France, South Africa, Spain, Sweden, and Switzerland) only refer to universal jurisdiction in the context of “grave breaches,” not “war crimes” more generally.[84]

Lack of “pure” universal jurisdiction. Additionally, several of the examples of State practice in the Study are not evidence of States vesting pure universal jurisdiction in their national courts over a set of offenses. Bangladesh’s relevant criminal legislation, for instance, only grants jurisdiction over acts occurring in Bangladesh.[85] The Netherlands’ military manual states that its law “has not entirely incorporated the principle of universality . . . . It requires that the Netherlands be involved in an armed conflict.”[86] Other States provide for universal jurisdiction only for a subset of acts within their various definitions of “war crimes.” France vests universal jurisdiction in its courts over serious violations of international humanitarian law only in cooperation with the International Criminal Tribunals for the Former Yugoslavia and for Rwanda, and relies on the territoriality, active, and passive personality principles for all other war crimes.[87] Likewise, Australia vests universal jurisdiction in its national courts over “grave breaches” of the Geneva Conventions and Additional Protocol I, but requires active personality in order to exercise jurisdiction over other war crimes.[88] Finally, the Study cites several law of war treaties that do not actually illustrate cases in which States Parties agreed to establish universal jurisdiction. For instance, Amended Protocol II to the CCW and the 1997 Ottawa Convention contemplate a territorial link between the State Party and the wrongful act.[89]

Limited practice of prosecutions. Furthermore, although the Study lists more than twenty States that have enacted or have drafted legislation apparently vesting universal jurisdiction in their national courts over “war crimes,” the Study cites a mere nineteen instances in which State courts supposedly have exercised universal jurisdiction over “war crimes.”[90] Of these nineteen, two are not on point because the defendants were not accused of “war crimes,” but of either genocide or genocide and crimes against humanity, respectively.[91] In another case cited in the Study, the government of Australia claimed jurisdiction based on the protective principle of national interest; the court based its decision on the plain language of a criminal statute and explicitly rejected the need to consider whether universal jurisdiction was applicable.[92] Additionally, in one Dutch case, the victims of the war crimes were Dutch citizens; consequently, the Dutch court based its jurisdiction on the passive personality principle, not on the basis of universal jurisdiction.[93]

If one puts these four inapposite cases aside, the remaining fifteen cases cited by the Study offer only weak evidence in support of rule 157. In six of these cases, States explicitly claimed jurisdiction based not on customary rights but on rights and obligations conferred in treaties, primarily under Article 146 of the Fourth Geneva Convention.[94] The nine cases in which States claimed jurisdiction based on customary rights come from only six States: Belgium, Canada, Israel, the Netherlands, Switzerland, and the United Kingdom.[95] The practice of six States is very weak evidence of the existence of a norm of customary international law. This body of practice is insufficiently dense to evidence a customary right of States to claim jurisdiction over the broad array of actions listed in rule 156, and is further weakened when one examines the facts of those cases. Indeed, in many of these cases, States were prosecuting acts that had been committed before the Geneva Conventions were adopted, but that ultimately were considered grave breaches in the Conventions.[96] Thus, although the prosecuting States were not in a position to rely on their treaty obligations as a basis for their prosecutions, the acts at issue effectively were grave breaches. These cases, therefore, should not be construed as supporting a customary right to claim jurisdiction over most of the acts listed in rule 156 as “war crimes” based on universality.

Opinio juris. Finally, and significantly, the Study fails to demonstrate that sufficient opinio juris exists to declare rule 157 customary international law. National legislation vesting universal jurisdiction over particular acts evidences the view of that State that it has the right to exercise such jurisdiction, but does not indicate whether that view is based on customary law or treaty law.[97] Among the evidence cited by the Study, at most nine States express a definitive opinio juris as to the customary nature of the right to vest universal jurisdiction (with the majority of those nine having never exercised this jurisdiction).[98] The majority of States that have adopted legislation make explicit in their laws that universal jurisdiction is based on prerogatives gained through treaties, not through customary international law. For example, the Geneva Conventions Act of Barbados provides that “a person who commits a grave breach of any of the Geneva Conventions of 1949 … may be tried and punished by any court in Barbados that has jurisdiction in respect of similar offences in Barbados as if the grave breach had been committed in Barbados.”[99] The legislation of France, Ireland, and Spain,[100] among others, also makes explicit that claims of universal jurisdiction stem from treaty law. Given the salience of treaty obligations in these and other instances, it is inappropriate to assume that the remaining States – those that do not explicitly state the legal basis for their legislation – do so out of a sense of entitlement arising from customary international law.

Summary. The State practice cited is insufficient to support a conclusion that the broad proposition suggested by rule 157 has become customary: examples of operational practice are limited to a handful of instances; a significant number of the examples do not support the rule; and the cited practice utilizes definitions of “war crimes” too divergent to be considered “both extensive and virtually uniform.”[101] Moreover, the Study offers limited evidence of opinio juris to support the claim that rule 157 is customary.

* * *

Closing observation

We have selected these rules from various sections of the Study, in an attempt to review a fair cross-section of the Study and its commentary. Although these rules obviously are of interest to the United States, this selection should not be taken to indicate that these are the rules of greatest import to the United States or that an in-depth consideration of many other rules will not reveal additional concerns.

[1] J.-M. Haenckerts and L. Doswald-Beck, Customary International Humanitarian Law (Cambridge 2005) (hereinafter, “Study”).

[2] Indeed, the authors of the Study may have intended to use the phrase “humanitarian relief personnel” as shorthand for “humanitarian relief personnel, when acting as such.” However, the rule as written does not say this, even though rule 33, which is closely related to rule 31, reflects the fact that the protection for peacekeepers attaches only as long as they are entitled to the protection given civilians under international humanitarian law.

[3] As Yoram Dinstein notes, “In keeping with Article 71(2) of Protocol I, personnel participating in the transportation and distribution of relief consignments must be protected. However, Article 71(1) underscores that the participation of such personnel in the relief action is subject to the approval of the Party in whose territory they carry out their duties.” The Conduct of Hostilities under the Law of International Armed Conflict 149 (Cambridge 2004) (hereinafter, “Dinstein”).

[4] By its terms, the Convention does not apply to enforcement action that the Security Council takes under Chapter VII of the UN Charter.

[5] CCW Amended Protocol II, Article 12(1)(a), 35 ILM (1996) 1206-17.

[6] Id. at Article 12(7)(b).

[7] Pictet’s Commentary on the Fourth Geneva Convention notes, “In theory, all humanitarian activities are covered

. . . subject to certain conditions with regard to the character of the organization undertaking them, the nature and objects of the activities concerned and, lastly, the will of the Parties to the conflict.” Commentary, IV Geneva Convention 96, Pictet, ed. (ICRC 1960) (hereinafter, “Pictet”). It continues, “All these humanitarian activities are subject to one final condition – the consent of the Parties to the conflict. This condition is obviously harsh but it might almost be said to be self-evident.” Pictet at 98. As discussed herein, we do not believe that this condition has disappeared since Pictet produced this Commentary.

[8] Rome Statute of the International Criminal Court, Article 8(2)(e)(iii), 37 ILM 999, 1008-09 (1998).

[9] Michael Cottier, War Crimes, in Commentary on the Rome Statute of the International Criminal Court 190, Triffterer, ed. (Nomos Verlagsgesellschaft 1999) (italics in original).

[10] Study, Vol. I, p. 109.

[11] Study, Vol. II, p. 589, paras. 5-6.

[12] Study, Vol. II, p. 589, para. 8 (citing the Ground Rules for Operation Lifeline Sudan), and p. 593, para. 39 (stating that Zimbabwe regards relevant provisions of the Geneva Conventions “as part of international customary law”).

[13] Study, Vol. II, pp. 589-90, paras. 13 (Australia) and 15 (France).

[14] Study, Vol. II, p. 590, para. 17.

[15] See, e.g., Study, Vol. II, p. 593-96, paras. 41-45, 47-62.

[16] See generally U.S. Joint Publication 3-07.6, Joint Tactics, Techniques, and Procedures for Foreign Humanitarian Assistance.

[17] Study, Vol. II, p. 590, para. 14. Furthermore, the manual cited by the Study is in fact a training manual designed to “briefly outline [] the Code of Conduct applicable to all Canadian personnel taking part in all military operations other than Canadian domestic operations.” Code of Conduct for Canadian Forces Personnel, Office of the Judge Advocate General, Canadian Ministry of National Defense, B-GG-005-027/AF-023 (undated), p. 1-1. It is not an official representation of Canada’s opinio juris concerning the laws and customs of war; instead, it repeatedly stresses that it is a simplification of applicable laws meant to aid in training.

[18] Study, Vol. II, p. 590, para. 16.

[19] Study, Vol. II, p. 591, para. 27.

[20] Study, Vol. II, p. 592, para. 30.

[21] Study, Vol. II, p. 602, para. 111.

[22] Similarly, on March 27, 2002, Israeli Defense Forces arrested a driver of a Red Crescent ambulance and seized an explosives belt and other explosive charges from the ambulance. The driver admitted that a terrorist leader had given him explosives to transport to terrorist operatives in Ramallah. See http://www.mfa.gov.il/mfa/government/communiques/2002/apprehension%20of%20ambulance%20harboring%20a%20wanted%20terro.

[23] Study, Vol. II, p. 592, paras. 33 (Nigeria) and 34 (Rwanda).

[24] Study, Vol. II, p. 593, para. 39.

[25] Study, Vol. II, p. 613, para. 181.

[26] See, e.g., Study, Vol. II, p. 612, paras. 178 (United Kingdom) and 180 (United States).

[27] This discussion focuses on only the first sentence in rule 45.

[28] Study, Vol. I, p. xxxviii.

[29] The Study includes these statements in Vol. II, p. 878, paras. 152 and 153.

[30] Remarks of Michael J. Matheson, Principal Deputy Legal Adviser, U.S. Department of State, The Sixth Annual American Red Cross-Washington College of Law Conference on International Humanitarian Law: A Workshop on Customary International Law and the 1977 Protocols Additional to the 1949 Geneva Conventions, 2 Am. U. J. Int’l L. and Pol’y 424, 436 (1987). One of the U.S. concerns has been that Articles 35(3) and 55 fail to acknowledge that use of such weapons is prohibited only if their use is clearly excessive in relation to the concrete and direct overall military advantage anticipated. The Study purposefully disregards this objection, even as to the contours of the customary rule. As the commentary states, “[T]his rule is absolute. If widespread, long-term and severe damage is inflicted, or the natural environment is used as a weapon, it is not relevant to inquire into whether this behaviour or result could be justified on the basis of military necessity or whether incidental damage was excessive.” Study, Vol. I, p. 157.

An example illustrates why States – particularly those not party to AP I – are unlikely to have supported rule 45. Suppose that country A has hidden its chemical and biological weapons arsenal in a large rainforest, and plans imminently to launch the arsenal at country B. Under such a rule, country B could not launch a strike against that arsenal if it expects that such a strike may cause widespread, long-term, and severe damage to the rainforest, even if it has evidence of country A’s imminent launch, and knows that such a launch itself would cause environmental devastation. Indeed, one of the Study’s authors has recognized elsewhere that the value of the military objective is relevant to an analysis of when an attack that will cause harm to the environment is permitted. See L. Doswald-Beck, International Humanitarian Law and the Advisory Opinion of the International Court of Justice on the Legality of the Threat or Use of Nuclear Weapons, 316 IRRC 35, 52 (1997).

[31] Study, Vol. II, p. 883, para. 186.

[32] Study, Vol. II, p. 882-83, para. 185.

[33] Letter dated June 20, 1995 from the Acting Legal Adviser of the Department of State, together with the Written Statement of the Government of the United States, p. 25-28; Letter dated June 16, 1995 from the Legal Adviser of the Foreign and Commonwealth Office of the United Kingdom of Great Britain and Northern Ireland, together with Written Statement of the Government of the United Kingdom, p. 40-46; Letter dated June 19, 1995 from the Ambassador of the Russian Federation, together with Written Statement of the Government of Russia, p. 10-11;

Lettre en date du 20 juin 1995 du Ministre des affaires étrangères de la République française, accompaignée de l’exposé écrit du Gouvernement de la République française, p. 31-33.

[34] Study, Vol. II, p. 882, para. 184.

[35] Digest of U.S. Practice, 1977, p. 919.

[36] Official Records of the Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law Applicable in Armed Conflicts, Vol. 5-6 (1977), p. 134.

[37] Statement of the United Kingdom, January 28, 1998, revised July 2, 2002. See also statement of France, April 11, 2001.

[38] We note that the Study raises doubts about the continued validity of the “persistent objector” doctrine. Study, Vol. I, p. xxxix. The U.S. Government believes that the doctrine remains valid.

[39] Study, Vol. II, p. 879, paras. 157 and 158.

[40] Study, Vol. II, p. 887-88, paras. 224 and 225. See also p. 900, para. 280 (CSCE committee drew attention to shelling that could result in harm to the environment, without indicating that such attacks were unlawful).

[41] Study, Vol. II, p. 900-01, para. 282.

[42] Study, Vol. II, p. 878 (para. 156), p. 879 (para. 160), p. 898 (paras. 273 and 274), p. 898-99 (para. 275), and p. 902 (para. 289).

[43] Study, Vol. II, p. 891, para. 244.

[44] Study, Vol. II, p. 890, para. 241.

[45] Rome Statute of the International Criminal Court, Article 8(2)(b)(iv).

[46] Study, Vol. I, p. 154.

[47] See Yoram Dinstein, Protection of the Environment in International Armed Conflict, 5 Max Planck UNYB 523, 543-46 and notes (2001) (discussing the illegality of Iraq’s acts but noting that “many scholars have adhered to the view that – while the damage caused by Iraq was undeniably widespread and severe – the ‘long term’ test (measured in decades) was not satisfied”).

[48] These attacks, of course, violated provisions of the law of armed conflict, particularly those relating to military necessity. The U.S. Government, in concurring in the opinion of the conference of international experts, convened in Ottawa, Canada from July 9-12, 1991, found that Iraq’s actions violated, among other provisions, Article 23(g) of the Annex to the 1907 Hague Convention IV and Article 147 of the Fourth Geneva Convention. See Letter dated March 19, 1993 From the Deputy Permanent Representative of the United States of America to the United Nations Addressed to the President of the Security Council, S/25441, p. 15.

[49] Final Report to the Prosecutor by the Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia (June 13, 2000), para. 15.

[50] Study, Vol. I, p. 156-57.

[51] Study, Vol. II, pp. 877-78, para. 150.

[52] Study, Vol. I, p. 273.

[53] Germany’s military manual recognizes a prohibition on those exploding bullets “which can disable only the individual directly concerned but not any other persons.” (emphasis added) (Study, Vol. II, p. 1788, para. 13). A U.S. legal review states that “an exploding projectile designed exclusively for antipersonnel use would be prohibited, as there is no military purpose for it.” (emphasis added) (id. at 1791, para. 35).

[54] Study, Vol. II, p. 1794, para. 47.

[55] Study, Vol. I, p. 273.

[56] In part, the 1998 test was flawed because it was set up in a way that was contrary to the principle that “in looking at small caliber weaponry, it is necessary to look not just at the bullet but at the entire means of delivery and the context in which the weapon will be used.” Christopher Greenwood, “Legal Aspects of Current Regulations.” Keynote speech at Third International Workshop on Wound Ballistics, Thun, Switzerland, March 28-29, 2001.

[57] 1 Am. J. Int’l L. (1907) Supp. 95-96.

[58] Study, Vol. I, p. 272.

[59] Of all the independent States in the Western Hemisphere, only Brazil acceded to the St. Petersburg Declaration. Additionally, none of the African or East Asian States in existence at the time acceded to the Declaration.

[60] Estonia acceded in 1991.

[61] Pertti Joenniemi and Allan Rosas, International Law and the Use of Conventional Weapons (1975), at 30.

[62] Study, Vol. II, p. 1791, para. 35.

[63] Hague Draft Rules of Air Warfare, 17 Am. J. Int’l L. (1923) Supp. 245-60, Ch. IV, Art. 18.

[64] The 2000 update of the 1998 U.S. legal review of the 12.7mm Raufoss Multi-purpose ammunition, other sections of which are cited by the Study (Study Vol. II, p. 1791, para. 35), lists widespread use of high-explosive or high-explosive-incendiary projectiles weighing less than 400 grams, many of which may have tended to detonate on impact or within the human body. Although the Study cites this review, it does not provide the full picture of the Study’s finding in that it omits this compilation of State practice.

[65] The Study cites military manuals of Australia, Belgium, Canada, France, Germany, Italy, New Zealand, Russia, Spain, the United Kingdom, and the United States. (Study, Vol. II, p. 1788-89, paras. 8-20.) Of these, Germany’s clearly opposes the rule as written, and France’s, Italy’s, and the United Kingdom’s offer inconclusive support. The U.S. Air Force Pamphlet, also cited for rule 157, bears a disclaimer that states, “This pamphlet is for the information and guidance of judge advocates and others particularly concerned with international law requirements applicable during armed conflict. It furnishes references and suggests solutions to a variety of legal problems but it is not directive in nature. As an Air Force pamphlet, it does not promulgate official U.S. Government policy although it does refer to U.S., DoD and Air Force policies.” The U.S. Air Force Pamphlet therefore cannot be considered a useful example of State practice.

[66] Legislation of Andorra, Australia, Ecuador, Italy, the Netherlands, and Yugoslavia. Ecuador’s legislation bans only the use of exploding bullets by its National Civil Police, and Italy’s includes an exception for “air or anti-air systems.” The Study notes that the 1945 Australian war crimes act prohibited “exploding bullets.” Study, Vol. II, p. 1790, paras. 21-26. The Study makes no reference to a 2001 Australian legal review of the 12.7mm Raufoss Multipurpose projectile, which concluded that munition was legal. Defence Legal Office, Defence Corporate Support, Australian Ministry of Defence, Memorandum CS 97/23/23431 (January 23, 2001), Subject: Legal Review of the 12.7mm Ammunition Produced by NAMMO. The Australian legal review was the subject of a presentation at the ICRC’s Expert Meeting on Legal Reviews of Weapons and the SIrUS Project, held at Jongny-sur-Vevey from January 29 – 31, 2001.

[67] Statements made by Brazil and Colombia do not support the assertion that the rule as written is customary, but rather express support for the prohibition of exploding bullets in some context. Study, Vol. II, p. 1790, paras. 28-29. The Study also includes statements by Norway and the UK made at the Second CCW Review Conference (2001) and before the ICJ in the Nuclear Weapons Case (1995), respectively. See id. at p. 1791, para. 32-33 (Norway) and para. 34 (UK). The Norwegian statement to the ICRC reflects Norway’s view that one must consider a number of factors, including intended use, when assessing the legality of a weapon; the UK statement appears to be a description of what the St. Petersburg Declaration provides.

[68] The Study sets forth only three purported examples of operational practice: the Report on the Practice of Indonesia (Vol. II, p. 1791, para. 30); the Report on the Practice of Jordan (id. at para. 31); and a statement by the Yugoslav Army (id. at p. 1792, para. 37). The Report on Indonesia does not actually appear to evidence operational practice; rather, it simply states what applicable law is in Indonesia.

[69] These are the military manuals of Germany and, arguably, the Penal Code of Yugoslavia.

[70] Study, Vol. II, p. 1791, para. 30.

[71] Id., para. 31.

[72] “The authorities and Armed Forces of the Republic of Slovenia are treating JNA as an occupation army; and are in their ruthless assaults on JNA members and their families going as far as to employ means and methods which were not even used by fascist units and which are prohibited under international law…. They are … using explosive bullets.” Study, Vol. II, p. 1792, para. 37.

[73] Study, Vol. II, p. 1789, para. 19.

[74] The Geneva Conventions and AP I incorporate elements that reflect these efforts.

[75] Study, Vol. I, p. xxix.

[76] Study, Vol. I, p. 574-603.

[77] “[I]t should be stressed that custom-generating practice has always consisted of actual acts of physical behaviour and not of mere words, which are, at most, only promises of a certain conduct. The frequent confusion seems to result from the fact that verbal acts, such as treaties, resolutions or declaration, are of course also acts of behaviour in the broad sense of the term and they may in certain cases also constitute custom-generating practice, but only as regards the custom of making such verbal acts, not the conduct postulated in them.” K. Wolfke, Some Persistent Controversies Regarding Customary International Law, 24 Netherlands Y.B. Int’l L. 1 (1993).

M. Cherif Bassiouni has discussed the limited practice of States invoking universal jurisdiction to prosecute various international crimes. He notes, “No country has universal jurisdiction for all these crimes [genocide, war crimes, crimes against humanity, piracy, slavery, torture, and apartheid]. It is therefore difficult to say anything more than universal jurisdiction exists sparsely in the practice of states and is prosecuted in only a limited way.” M. Cherif Bassiouni, Universal Jurisdiction for International Crimes: Historical Perspectives and Contemporary Practice, 42 Va. J. Int’l L. 81, 136 n.195 (2001).

[78] Study, Vol. II, p. 3894-3912, paras. 163-245.

[79] Study, Vol. II, p. 3858, para. 22 (Croatia), p. 3859, para. 28 (Hungary), and p. 3861, para. 38 (Switzerland). The Study also includes a number of citations to State laws and manuals that do not include law of war offenses, but rather refer to provisions such as “other punishable acts against human rights” (Costa Rica, p. 3899, para. 182); “crimes against humanity, human dignity or collective health or prosecutable under international treaties” (Cuba, p. 3899, para. 184); and the substance of Articles 64 and 66 of GC IV related to the trial of civilians in occupied territory (Argentina, p. 3894, para. 163).

[80] Study, Vol. II, p. 3858, para. 20; see also id., p. 3864-65, paras. 51-52.

[81] See Study, Vol. I, p. 604 n. 194 (listing states). This discussion is not intended to suggest that the U.S. Government believes that the Study has shown conclusively the customary nature of Rule 156.

[82] Study, Vol. II, p. 3895, para. 166 (Australia); p. 3896, para. 172 (Belgium); p. 3898, para. 180 (Colombia); p. 3899, paras. 185-86 (Cyprus); p. 3912, para. 245 (Zimbabwe).

[83] Study, Vol. II, p. 3896, para. 170 (Barbados); id., para. 174 (Botswana); p. 3908, para. 227 (Singapore); and p. 3910, para. 236 (Uganda).

[84] Study, Vol. II, p. 3888, para. 145 (Belgium); p. 3889, para. 148 (France); p. 3890, para. 153 (South Africa); p. 3890, para. 154 (Spain); p. 3890-91, para. 155 (Sweden); and p. 3891, para. 156 (Switzerland).

[85] Study, Vol. II, p. 3959-60, para. 397.

[86] Study, Vol. II, p. 3889, para. 150.

[87] Study, Vol. II, p. 3900-01, paras. 192-95.

[88] Study, Vol. II, p. 3894-95, paras. 165-66.

[89] Study, Vol. II, p. 3885, paras. 132-33.

[90] Although Volume II of the Study contains references to twenty-seven cases, the Study does not assert that eight of these cases are examples of States exercising universal jurisdiction over war crimes. For example, the Musema case appears to be a situation in which Switzerland simply determined that dual criminality existed in Switzerland with regard to the offense for which the ICTR sought the defendant.

[91] The Munyeshyaka case in France and the Demjanjuk case in the United States (which subsequently was overturned on unrelated grounds). In the Demjanjuk case, the Israeli arrest warrant on which the extradition request was based charged that Demjanjuk had operated the gas chambers in Treblinka “with the intention of destroying the Jewish people [i.e., genocide] and to commit crimes against humanity.” Demjanjuk v. Petrovsky, 776 F.2d 571, 578 (6th Cir. 1985). For the Munyeshyaka case, see Study, Vol. II, p. 3915, para. 253.

[92] The Polyukhovich case. The majority opinion stated, “It is enough that Parliament’s judgment is that Australia has an interest or concern. It is inconceivable that the court could overrule Parliament’s decision on that question. That Australia has such an interest or concern in the subject matter of the legislation here, stemming from Australia’s participation in the Second World War, goes virtually without saying . . . . It is also unnecessary to deal with the alternative submission that the law is a valid exercise of the power because it facilitates the exercise of universal jurisdiction under international law.” 91 ILR 13-14 (1991).

[93] The Rohrig and Others case. “Article 4 of the Decree on Special Criminal Law [that the defendants were charged with violating] was, however, in accordance with international law as being based on the principle of ‘passive nationality’ or ‘protection of national interests.’” 17 ILR 393, 396 (1950).

[94] See Study, Vol. II, p. 3914, para. 251 (Sariæ), p. 3914-15, para. 252 (Javor), p. 3915-16, para. 254 (Djajiæ), p. 3916-17, para. 255 (Jorgiæ), p. 3917, para. 256 (Sokoloviæ) and para. 257 (Kusljiæ). The prosecution in the Sokoloviæ and Kusljiæ cases successfully argued that crimes committed by the accused (Bosnian nationals) in Bosnia and Herzegovina were part of an international armed conflict, and that obligations under Article 146 of the Geneva Conventions (relating to grave breaches) therefore were applicable. It follows that this arguably strained reliance on the Geneva Conventions denotes a hesitance to claim a right to universal jurisdiction under customary international law. In addition, the German Penal Code permitted its domestic courts to exercise jurisdiction over grave breaches “if this was provided for in an international treaty binding on Germany.” Thus, the German law explicitly looks to the existence of a treaty permitting the exercise of such jurisdiction, and does not rely on any customary international legal “right.”

[95] These are, from Belgium: Public Prosecutor v. Higaniro (Four from Butare case) and Public Prosecutor v Ndombasi, which led to the Case Concerning the Arrest Warrant of 11 April 2000 (2002 I.C.J. 3); from Canada: the Finta case; from Israel: the Eichmann case; from the Netherlands: the Knesevic case and the Ahlbrecht case – the latter of which concerned acts committed in occupied Holland and therefore is not a clear example of the invocation of universal jurisdiction; from Switzerland: the Grabez case and the Niyonteze case; and from the United Kingdom: the Sawoniuk case. For the Ahlbrecht case, see 14 ILR 196 (1947).

[96] In the Finta, Ahlbrecht, Sawioniuk, and Eichmann cases, the only “war crimes” of which the defendants were accused would have constituted grave breaches of the Fourth Geneva Convention, including forced deportation and murder of protected persons, if that Convention had been in effect at the time they were committed. See Regina v. Finta, 69 O.R. (2d) 557 (Canadian High Ct. of Justice 1989), 14 ILR 196, 2 Cr App Rep 220 (UK Court of Appeal, Criminal Division 2000), and 36 ILR 5, respectively.

[97] The Geneva Conventions, for instance, require States Parties to “enact any legislation necessary to provide effective penal sanctions for persons committing, or ordering to be committed, any of the grave breaches of the present Convention defined in the following Article.” See, e.g., 1949 Geneva Convention IV Relative to the Protection of Civilian Persons in Time of War, Art. 146.

[98] These States are Australia, Azerbaijan, Bangladesh, Belarus, Belgium, Ecuador, Switzerland, Tajikistan, and possibly New Zealand.

[99] Study, Vol. II, p. 3896, para. 170.

[100] Study, Vol. II, p. 3901, para. 194 (France); p. 3902-03, para. 202 (Ireland); p. 3909, para. 229 (Spain).

[101] North Sea Continental Shelf Cases, [1969] I.C.J. Reports at 43.

[i] As the Study itself indicates, the field has traditionally been called the “laws and customs of war.” Accordingly, we will use this term, or the term “law of war,” throughout. J.-M. Haenckerts and L. Doswald-Beck, Customary International Humanitarian Law, Vol. I, p. xxv (Cambridge 2005) (hereinafter, “Study”).