The Empire Strikes Back – Debating the Origins of the Customary Laws of War

by Duncan Hollis

In 2005, the International Committee for the Red Cross (ICRC) released a VERY long study on customary international humanitarian law. It is an impressive work, spanning more than 5000 pages, presenting the detailed and expansive views of an organization uniquely situated to know what rules have emerged binding states as a matter of custom, separate and apart from their treaty commitments under the laws of war. But, as John Bellinger alluded during his guest blogging stint with us, the ICRC study was not the last word on the subject. Late last year, Bellinger and DOD General Counsel, William Haynes, sent the ICRC a letter, providing the official U.S. reaction to the Study. You can now access it here. Regardless of whether you agree or disagree with U.S. views on the laws of war, this is an important letter. International lawyers should read it for at least two reasons.

First, read the letter on the merits. It records U.S. disagreement with some of the ICRC study’s general and specific conclusions. Thus, it rejects the study’s 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 those that did not join the Protocols (e.g., the United States). The U.S. response also questions the ICRC 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.” More specifically, the United States attaches to its letter a series of critiques of the ICRC’s characterization of rules on (i) respect due humanitarian relief personnel, (ii) the prohibition of anti-personnel bullets that explode in the human body, (iii) the right of states to vest universal jurisdiction in national courts over war crimes; and (iv) the very existence of a rule prohibiting methods or means of warfare causing widespread, long-term or severe damage to the natural environment. Here’s a taste of the U.S. views:

Virtually none of the evidence of practice cited in support of rule 78–[prohibiting use of bullets that explode within the human body]–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.

Since the United States engages in armed conflict more than other states (think Iraq, Afghanistan, Somalia), its views have obvious practical implications for the conduct of armed conflict. They may also effect how other states react to the ICRC study, and—more importantly—the content of customary international humanitarian law.

Second, read the letter for its methodology. Whether you agree with the U.S. letter (or the ICRC study) may well turn on when you would admit the existence of customary international law. Indeed, the U.S.-ICRC differences are fundamentally about method—how customary international law comes into existence. The exchange reveals many, if not most, of the on-going debates about the formation of customary international law.

Now, neither the United States nor the ICRC appear to challenge the traditional criteria for identifying a rule of customary international law – a consistent and uniform practice of some duration that states follow out of a sense of legal obligation (a.k.a. our namesake, opinio juris). Where they disagree is in the application of those criteria, specifically what constitutes state practice in the first place. The United States complains, for example, about how that the ICRC study undervalues contrary or “negative” practice and that of specially affected states (here, the United States appears to be asserting a some-are-more-equal-than-others approach to dictating the content of customary laws of war on the grounds that the U.S. practice should matter more, given its experience and frequent involvement in armed conflicts).

The treaty lawyer in me found the discussion of how treaties affect custom fascinating. The U.S. letter questions how the ICRC study can provide a more general formulation of the protection due humanitarian relief personnel than in the qualified protections available under the relevant treaties, which incorporate a state consent element absent in the ICRC formulation:

Despite the fact that these treaties clearly qualify State obligations regarding [humanitarian relef personnel on state consent, the ICRC rule] 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.

The United States voices a similar objection about the ICRC inferring a customary right for national courts to assert universal jurisdiction over war crimes since states might infer that right from their treaty obligations rather than a more general customary rule.

The BIG difference between the United States and the ICRC, however, lies in the general question of what “counts” for state practice. Does the rhetoric employed by states and other actors (via military manuals, certain widely adopted U.N. General Assembly resolutions, NGO statements) qualify or—as the U.S. letter asserts—must practice turn more on an examination of what states actually do (or don’t do). In other words, does custom emerge from rhetoric or actual conduct? Of course, this is not a new debate, but one that’s plagued international lawyers for some time. It’s also a question courts frequently face. Indeed, one can see the judicial version of the US-ICRC debate in the contrasting views of custom expressed by the Supreme Court in Sosa v. Alvarez-Machain and Judge Kaufman’s views in Filartiga v. Pena-Irala.

Finally, on the question of opinio juris itself, the U.S. letter challenges the ICRC’s willingness to presume its existence when state practice is sufficiently dense: “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.” The U.S. view may actually find some support in the ICJ’s jurisprudence (see, e.g., Lotus, North Sea Continental Shelf and Nicaragua), which generally seems to require states seeking to establish a customary rule to provide affirmative evidence of opinio juris. The ICRC, in contrast, seems to follow the ILA’s lead and operate with a presumption of opinio juris in the face of state silence on whether conduct results from a sense of legal obligation or a non-legal basis (e.g., courtesy, political expedience, etc.).

Of course, as important as the U.S. views are on these issues, they are not definitive. As a result, it would be interesting to see if other states have articulated responses to the ICRC study, and, if so, whether they view it more favorably than the United States? Do readers know of any such reactions? Of course, I’d also be interested to see what readers think of the U.S. letter or the ICRC study that spawned it.

7 Responses

  1. Duncan: Thanks for the post. It is, as you say, fascinating to see this debate play out between the U.S. &and the ICRC. I just wanted to point out that those who don’t have the time (or inclination) to wade through the 3 volumes of the ICRC study might want to take a look at an article by the head of the project, Jean-Marie Henckaerts, in the IRRC (Vol. 87 No. 857 March 2005). It explains the methodology and organization of the study and summarizes some its major findings It’s available at the ICRC website at the following link

  2. Duncan,

    My thanks for this post. I’d first note that Ken Anderson has also posted about this letter a couple of weeks ago. I’d also note that most of the US government quite reasonable criticism of the ICRC Study has been foreshadowed by Prof. Yoram Dinstein’s initial review of the study in Vol. 36 of the Israel Yearbook of Human Rights.

    Anyway, when it comes to the substance, I tend to agree with the general statement that the ICRC was at times a bit too expansive in its assessment of state practice and opinio juris. But I also think that the US government’s objection to the claim 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 those that did not join the Protocols, or have become binding as a matter of customary international law in internal armed conflicts as well as international ones is a bit too general to be taken as opinio juris against the formation of any specific rule.

    Surely, many rules from the APs were a codification of preexisting customary law (such the rule of proportionality, for instance), while others have certainly evolved to that status since. What is always true, however, is that it can be extremely difficult to deduce custom from treaty practice.

    When it comes to the specific illustrative comments offered by the US government, I would have two additional remarks.

    First, the government letter makes absolutely no mention of Rules 87-105 of the ICRC Study, which enshine fundamental guarantees to which all persons in the power of the enemy are entitled to regardless of their status. In other words, the letter does not address the issue of the customary status of Article 75 of Protocol I, which was acknowledged by all US legal advisers until the present one, who stated that the administration is ‘looking into’ the issue. Needless to say, this might have implications on the treatment of those designated as unlawful combatants.

    Second, the government’s discussion of universal jurisdiction doesn’t seem to be all that persuasive. Moreover, it is rather obviously motivated by the Bush administration’s concern about some of its officials who might be subjected to prosecution elsewhere in the world. I would actually posit that we would need affirmative state practice and opinio juris which would show that states CANNOT exercise universal jurisdiction over serious international crimes, just as I would ask the proponents of universal jurisdiction to show that customary law has evolved to the point at which normal immunities of state officials for their official acts are suspended, particularly when it comes to universal jurisdiction in absentia.

  3. As Duncan notes the ICRC’s study raises some very interesting issues regarding the methodology to be used in determining customary international law. The US responses raises similar interesting questions.

    1) One can only agree with the US that an area where the ICRC study unacceptably glosses over matters is the use made of treaties and the practice of parties to treaty in assessing whether a new rule of custom has formed. The issue raised here is the well known Baxter paradox (named after the former Harvard Professor and ICJ Judge). Where there is a widely ratified treaty (like the Geneva Conventions or perhaps more relevant here, the Additional Protocols to the GC) which operates in an area, which practice does one look at in assessing what customary international law in that area is. Does one look at the practice of the parties to the treaty? The problem with this is that presumably the practice of the parties is not supported by opinio juris since that practice follows from the obligations the parties have undertaken under the treaty. Or does one use only the practice of non-parties? The problem is that only a few states would then be creating customary international law. So since there are only about 30 states not party to Add. Protocol I, is it only their practice that counts? This would be bizarre.

    The ICRC study tries to get round this problem by saying that one can use the practice of States parties to treaties in their relations with non-parties. I agree with this approach since in such a case the practice of the party cannot be said to be based on the treaty. However, the problem is that the study does not appear to confine use of the practice of treaty parties to these sorts of cases. On the contrary, as the US says, the ICRC seems to rely rather extensively on the practice of parties to treaty without in all (or perhaps most) cases determining that this practice is actually based on opinio juris (a belief that the practice is required by or relevant to customary international law).

    2) In my view, the US response goes too far when it says “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.” This appears to represent what may be called a common sense view that “action speaks louder than words”. (It’s a position taken by writers such as D’Amato and rejected by writers such as Akehurst} In this case, the view is misguided. State practice is what states do. And States can “act” in many ways. It is difficult to see why one should prefer operational practice to verbal practice. Both are reflections of State behaviour. They are indications of a State’s position on a particular point. One could even take the view that operational practice might in certain cases be less valuable than verbal practice because the former is carried out in the heat of the moment perhaps without much thought as to its implications while the latter is more likely to reflect the considered position of the State on the issue at hand. Indeed, when a State’s armed forces (or other officials) have engaged in certain actions it only makes sense that the State should be free to repudiate such action through its official policy, its national legislation and the statements of its senior officials. In such a context it would make little sense to continue to say that the State’s practice is only that or mainly that reflected in its “actions.” Likewise, if the actions come after the official policy and position are spelled out one ought not to prefer that action unless it is clear that there is an intent to override the considered position of the State.

    3). While the US is right that special weight ought to be given the practice of specially affected States (see the ICJ’s North Sea Continental Shelf cases), the US goes too far in how it defines specially affected States. The US takes the view that in matters pertaining to the use of nuclear weapons (eg the application of the rule protecting the natural environment) those States that possess nuclear weapons are specially affected. A similar position was taken by Judge Schwebel in the Nuclear Weapons Advisory Opinion. Where States are not equally affected by a rule then those particularly affected ought to have more of a say in determining the law. Nevertheless, one must be careful in judging who is specially affected by the law in an area. I don’t think the use of nuclear weapons is one of those areas where any State is any more “specially affected”. It is true that only some States have nuclear weapons and that they would bear the burden of any rule prohibiting or restricting the use of such weapons. However, all States could potentially be affected by (and subject to) the use of nuclear weapons. Therefore all States have an equal interest in the legality of the use of those weapons. There is no reason for the views of potentially users should carry more weight than the view of those potentially subject to the use of nuclear weapons.

    It would be great to see the US responses to the particular rules that the Study comes up with. In this initial response the US refers to only 3 of the rules. It makes some cogent points in relation to the rule in the ICRC study providing for universal jurisdiction over war crimes which I look forward to examining in greater detail.

  4. Marko,

    On what basis do you posit that one needs affirmative State practice and opinio juris that a State CANNOT exercise universal jurisdiction? Are you relying on the Lotus principle (ie the statement of the PCIJ in the Lotus case)? If so, would you not agree that international law as represented by State practice does not take the view that States generally have prescriptive jurisdiction unless one can show that such jurisdiction is prohibited. Practice and thinking have moved on since Lotus. No State asserts that it has jurisdiction just because that it does and that the burden is on the State arguing jurisdiction to show a prohibitive rule. Recent State practice unanimously takes the view that the burden is on the State seeking to assert jurisdiction to show a permissive rule.

    Is the US not right that there is a question as to what constitutes an international crime, even a war crime, for which universal jurisdiction is possible? The position with regard to Grave Breaches of the Geneva Conventions is clear. The position for Torture is clear. There are relevant treaties allowing – No, obliging the exercise of universal jurisdiction in such cases. But the position is much less clear when one moves to (i) other crimes committed in international armed conflicts, (ii) crimes committed in non-international armed conflicts and (iii)crimes against humanity. Even if one takes “verbal practice” as I do, then one does have to search for practice where States allow for prosecution without any link to the State. There is recent practice pointing towards universal jurisdiction in these cass but its not extensive.

  5. Could somebody explain why the U.S. just doesn’t clearly say it is a persistent objector? Who cares if it is custom or not The U.S. has never said that these norms apply to it; therefore, it seems like if anybody has a persistent objector claim it is the U.S.

  6. Dapo,

    Well, I’m not sure that I agree with you that the Lotus principle is totally dead in the water, especially when the issue is not whether universal jurisdiction can be exercised in principle, but whether it can be exercised for a specific crime. As you well know, in customary law the first of these crimes was piracy, with pirates being treated as hostes humani generis, enemies of the human race. If pirates and war criminals are enemies of the human race, why aren’t those who commit genocide and crimes against humanity?

    Furthermore, it’s not like the issue of universal jurisdiction is exactly new. Israel exercised universal jurisdiction over Eichmann some fifty years ago, and I heard of no state protesting (other than in relation to the circumstances of his capture). Moreover, many states have for decades incorporated the universality principle in their domestic criminal codes, and extended that principle beyond those crimes enumerated in treaty law. Again, no protests. The only real protests we have are instances in which a state tries to exercise universal jurisdiction in absentia, and disregards international immunities when doing so.


    Persistent objector to what, exactly? All of the rules in the Study? That would just be total nonsense, as the US agrees that many, probably most of these rules are part of customary law. Furthermore, persistent objection requires that a state indeed publicly objects to the formation of a rule. A state’s silence alone cannot be construed as an objection. When it comes to the specific rules that the US criticizes, the approach taken in the letter is much more honest than persistent objection alone, as it implies that a specific rule doesn’t even exist due to absence of state practice and opinio juris of states other than the US.

  7. Marko,

    I didn’t say the “whole study.” Nothing in your post answers my question, why doesn’t the U.S. just say we object to norms X, Y, and Z regardless of if they are custom for other states and, therefore, claim persistent objector status. That just seems like such a nice, legally coherent way to deal with the report. All this stuff on when state practice occurs would be superfluous.

Trackbacks and Pingbacks

  1. There are no trackbacks or pingbacks associated with this post at this time.