Oona Hathaway is absolutely right to be alarmed, and deeply concerned, about the “civilians’ assumption of risk” provisions in the new DoD Law of War Manual. If you haven’t done so already, you really should read her excellent post. In this complementary post, I’ll elaborate a bit on how the problem is even more endemic within the Manual than Oona describes; why it’s wrong as a matter of law; its origins; and what’s most (and least) worrisome about it.

Back in February, I was honored to participate in a Georgetown Law Panel on the new Manual, along with Georgetown Visiting Professor Jamie Baker, DoD Deputy General Counsel Chuck Allen, Associate General Counsels Karl Chang and Matt McCormack (both of whom worked on the Manual), and Professor Laura Dickinson. You can view the event here. In my remarks, I lauded the authors for their extraordinary dedication, thoroughness and care, and for what Chuck Allen rightly described as the major improvements that had been made since the previous iteration of the Manual was circulated interagency in 2010, when I worked at DOJ. The Manual is an extraordinary document; it has the potential to be a resource of extraordinary value, notwithstanding its somewhat unwieldy nature (it’s 1,180 single-spaced pages and contains 6,916 footnotes).

Nevertheless, I warned that there was a serious risk that much of the Department’s hard work, and good will, would be for naught — that much of the rest of the world, and perhaps even U.S. military officials themselves, would question the Manual’s authoritativeness — for two principal reasons:

First, although the Manual purports to be “a description of the law [of war, the jus in bello] as of the date of the manual’s promulgation” (section 1.1.2), it relies almost exclusively on authorities promulgated by the U.S. military itself, and — despite the 7,000 or so footnotes — it rarely cites, or even contends with, the practice and opinio juris of the other nations of the world, even (or especially) when those international sources contradict or complicate the U.S. views. A purported description of customary international law will not be taken seriously if it focuses exclusively on the views and practices of the military department a single nation — the very department that is purporting to describe the law.

Second, even if the vast majority of the Manual does correctly describe the customary jus in bello — and at the Georgetown panel I hazarded a (perhaps conservative) guess that well upward of 95 percent of the Manual does so — the rest of the world will likely view the whole thing with deep skepticism (or even disdain) because of a handful of very important sections involving questions relating to the protection of civilians.

For the most part, those sections of the Manual describe the law accurately, and appear to reflect very robust, protective perspectives on the fundamental laws of war that are designed to protect civilians and civilian objects from the ravages of war (e.g., the rules of distinction, precaution, and proportionality). Even so, many of them contain apparent qualifications that, upon scrutiny, appear to establish serious gaps in protection, thereby undermining the laudable humanitarian impulses in the pages that surround them. Moreover, those qualifications — which many may come to see as loopholes, or the fly in the ointment—tend to be outlier positions, to put it mildly: they are propositions that few, if any, other states endorse, and for which the Manual cites almost no persuasive authority. The credibility of the Manual, and more broadly of the United States’ stated commitment to compliance with the laws of war, is undermined by this handful of controversial and ill-defended propositions, which collectively will suggest to many readers that although the U.S. Department of Defense talks a good game about protecting civilians, it has quietly preserved the authority to disregard or de-emphasize harm to civilians in its targeting practices.

Of these potentially controversial provisions, I raised two that I think are of the most concern: The first is the statement in Section that “[u]nder customary international law, no legal presumption of civilian status exists for persons or objects.” That section also expressly denies that customary law includes the requirement in AP I Article 50(1) that “[i]n case of doubt whether a person is a civilian, that person shall be considered to be a civilian” (as well as the corresponding provision of Article 52(3), establishing a similar presumption in a case of doubt about whether or not an object is dedicated to civilian purposes).

Even if one concedes, as almost all serious authorities and commentators do, that customary law does not require a presumption of civilian status whenever there is any doubt whatsoever (e.g., even when the commander is 95 percent certain that the person or object is a legitimate military target), the notion that there is no legal presumption of civilian status is, to say the least, startling. Nor does the Manual cite any authority in support of it, let alone authority sufficient to support a proposition about what customary law requires in cases of doubt. It is statements such as this one that give fuel to the unfortunate and inaccurate (see the footnote), but oft-stated, belief that the United States treats all military-aged males in the vicinity of al Qaeda operatives as “combatants.”

The Department should amend the Manual to more accurately reflect customary law concerning how to treat individuals in cases of doubt about their status. Even if it may be difficult to precisely articulate how much doubt is too much doubt to permit the use of force (see Adil Haque’s article here), the Manual should, at a minimum, acknowledge that surely there is a “legal presumption of civilian status” when the relevant officer has little or no reason to think that a person is part of enemy forces (or a civilian directly participating in hostilities), where the preponderance of evidence points to civilian status, or where the officer is not fairly confident that the person is a lawful target.*

The second of the two most troubling principles in the Manual that I emphasized at the Georgetown event in February is the one that Oona has flagged: the Manual’s repeated insistence that when civilians — or other “protected” persons who cannot themselves be targeted — are in or near lawful military targets, they have in some sense “accepted the risk” of death or further injury due to their proximity to military operations, and that therefore any foreseeable harm to those civilians and other protected persons need not be taken into account in assessing whether an attack would comply with the principle of proportionality.

*          *          *

Oona identified three places where that proposition appears in the Manual; but there are more — many more. The principle is expressly stated or strongly implied in a dozen or more provisions, all of which cross-reference, or eventually circle back to, the section of the Manual where it is most comprehensively discussed, Section According to the Manual, officers applying a proportionality analysis do not need to account for the foreseeable harm to all of the following categories of persons — none of whom could themselves be targeted — whenever they are in proximity to lawful targets (or, at a minimum, when they know that they are in such proximity):

— civilian workers “who place themselves in or on a military objective, knowing that it is susceptible to attack, such as workers in munitions factories” (see Section;

— civilians authorized to accompany the armed forces (see Sections,, 14.4.2);

— medical and religious personnel ( see Sections (“Because medical and religious personnel are deemed to have accepted the risk of death or further injury due to proximity to military operations, they need not be considered as incidental harm in assessing proportionality in conducting attacks.”), (“Certain medical and religious personnel [in a noninternational armed conflict]. . . are deemed to have accepted the risk of harm due to their deliberate proximity to military objectives; thus, expected incidental harm to such persons would be understood not to prohibit attacks under the proportionality rule.”));

— medical units and facilities (see Sections (“Because medical units and facilities that are positioned near military objectives are deemed to have accepted the risk of death or further injury due to proximity to military operations, they need not be considered as incidental harm in assessing proportionality in conducting attacks.”), (similar, as to noninternational armed conflicts));

— so-called “involuntary human shields,” i.e., civilians whom the enemy has coerced to be present in or near lawful targets (see Section (“If the proportionality rule were interpreted to permit the use of human shields to prohibit attacks, such an interpretation would perversely encourage the use of human shields and allow violations by the defending force to increase the legal obligations on the attacking force.”));

— combatants who are wounded, sick or shipwrecked (see Sections (“Because combatants who are wounded, sick, or shipwrecked on the battlefield are deemed to have accepted the risk of death or further injury due to proximity to military operations, they need not be considered as incidental harm in assessing proportionality in conducting attacks.”), (similar, as to noninternational armed conflicts));

— prisoners of war and other “protected persons” under the Geneva Conventions, including persons protected by a flag of truce, such as parlementaires (persons sent to conduct negotiations under the display of the white flag of truce) (see Sections, 5.16.1 (“Incidental harm to those individuals would be understood not to prohibit attacks under the proportionality rule.”), 12.5.3).

In addition to these many provisions, all of which insist that customary law excludes such persons from the required proportionality analysis, Section 5.3.1 even goes so far as to suggest that the laws of war impose an affirmative obligation upon such persons to avoid becoming collateral damage: “Civilians also may share in the responsibility to take precautions for their own protection.” (The only authority cited in support of this remarkable proposition is a stray U.S. comment in 1991 to an ICRC memorandum about the Gulf War, as well as a 1907 U.S. “working memorandum” that does not even include any such suggestion of an affirmative responsibility.)

*          *          *

As Oona quite rightly points out, the Manual cites very few authorities as ostensible support for its mistaken proposition about the effect of an “assumption of risk” on the application of the rule of proportionality in these many Sections. Greenspan’s 1959 Modern Law of Land Warfare does not support it. Nor do the Bothe, Partsch & Solf commentaries on Additional Protocol I, nor the ICRC’s “Draft Rules” that led to the first Protocol.

Indeed, the latter two authorities actually reject the Manual’s proposition about proportionality. See especially page 57 of the ICRC’s 1973 Draft Rules, which states expressly that although “civilians who are within or in the immediate vicinity of military objectives run the risk of ‘incidental’ effects as a result of attacks launched against those objectives” — and therefore such civilian presence does not automatically preclude attacks on such military objectives — nevertheless, “other provisions of the draft would be applicable,” including, specifically “paragraph 3 (b) of the present article” (which reflected the rule of proportionality as applied to losses to the civilian population) and “Art. 50 (1) (a) and (b) concerning proportionality.” See also, e.g., BP&S at pages 128 (asserting that the proportionality requirement applies as a “general principle of armed conflict” to potentially limit the permissible “collateral damage” to medical units, just as it dos to the “civilian population” generally) and 346 (suggesting agreement with the ICRC’s view that the permissibility of collateral damage to civilians “in or near a military objective” depends on application of the “rule of proportionality”).

All that remains, then, of the nominal authorities cited by the Manual, is one stray (and undefended) comment in the U.S. Navy’s own 2007 Commander’s Handbook on the Law of Naval Operations, to the effect that not only does “[t]he presence of civilian workers, such as technical representatives aboard a warship or employees in a munitions factory, in or on a military objective, . . . not alter the status of the military objective” — which is correct — but also that “[t]hese civilians may be excluded from the proportionality analysis.”

Suffice it to say that this single, unadorned and undefended comment in a 2007 U.S. military manual does not, in and of itself, establish or reflect a customary rule of international law.

*          *          *

So, where, then, did the Manual derive such a notion? It is not difficult to identify the source: The idea was, for many years, strongly endorsed by W. Hays Parks, a Senior Associate Deputy General Counsel at the Pentagon from 2003 to 2010, who was principally responsible for drafting earlier iterations of the Manual.

The most elaborate expression of Parks’s views on the question of proportionality in targeting is contained in his comprehensive 1990 Air Force Law Review article entitled “Air War and the Law of War,” written when he was a Special Assistant for Law of War Matters in the Army’s Office of the Judge Advocate General. (That’s the same article in which Parks announced that the new Law of War Manual would be “completed in 1990” (by the Army), 25 years before the Pentagon actually issued it.)

One animating principle of Parks’s 1990 article was, in short, that “[a]ttack of a military objective, wherever located, is lawful,” without regard to the anticipated collateral harm to civilians and civilian objects. This meant, in effect — and as Parks freely acknowledged — that, on his view, customary law does not include the “proportionality” requirement articulated in AP I Article 57(2)(a)(iii), namely, that those who plan or decide upon an attack must “[r]efrain from deciding to launch any attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated.”

Parks wrote that “[i]n the course of the American military review of Protocol I, it was concluded that the concept of proportionality is not a rule of customary law.” Perhaps that’s so — i.e., that some internal “military review” rejected the proposition that AP I did not reflect customary law. Perhaps Parks was referring to page 15 of this 1982 “informal” and “preliminary” review of AP I by the Joint Chiefs, which likewise failed to offer any support for the proposition. If so, however, that conclusion of the “military review” would have been inconsistent with Article 41 of the Army Field Manual in effect at the time (and still in effect when Parks wrote in 1990). With a specific citation to the Protocol, Article 41 of Field Manual 27-10 provided that “loss of life and damage to property incidental to attacks must not be excessive in relation to the concrete and direct military advantage expected to be gained.” Moreover, the United States did not adopt the reported position of Parks’s “American military review”; to the contrary, the final and more extensive review by the Joint Chiefs of Staff in 1985 accepted the rule of proportionality, and even acknowledged (see page 48) that “[m]any legal experts believe that this rule is already binding on the United States as part of customary international law.” Then, in 1987, three years before Parks’s article, the United States publicly announced (see page 426 here) that it accepted the AP I proportionality rule as a “fundamental principle.”

In any event, according to Parks’s own, decidedly idiosyncratic, view reflected in his 1990 article, the so-called “concept” of proportionality — he refused to acknowledge it as a customary rule or a requirement — “does not establish a separate standard,” but instead merely “serves as a means for determining whether a nation or military commander responsible for planning, deciding upon, or executing a military operation has engaged in the intentional attack of civilians not engaged in the hostilities.” Or, at the very most, as Parks wrote in an internal 1983 document entitled “Proportionality in a Nutshell,” he would have recognized a violation of “the concept of proportionality” only with respect to acts “that are tantamount to the direct attack of the civilian population, . . . or involve wanton negligence that is tantamount to an intentional attack of the civilian population.”

Not surprisingly, then, Parks was unwilling to “count” persons who in the vicinity of a lawful target in any proportionality calculus. “In the course of numerous fora on the Protocols,” he wrote, he had “asked an equally numerous number of proponents what an attacker is expected to do if the enemy civilian population fails to heed warning of an impending attack, or if the defender forces the civilian population to remain in or near an obvious military objective.” His interlocutors replied that, in some such cases, the rule described in Article 57 might then preclude, or require postponement of, the attack. This answer, wrote Parks, was “altogether impractical.” (On that view, Parks reasoned, the U.S. bombing of Dresden in 1945 might have been unlawful, because the deaths of tens of thousands of civilians in that city were foreseeable, and arguably excessive “in relation to the concrete and direct military advantage anticipated.” Parks appeared to be unwilling to entertain the notion that just because the United States took certain acts in World War II does not necessarily mean that such acts would be lawful in light of the customary law of 1990.)

Parks wrote in his article that he “disagrees with the statement contained in the Bothe, Partsch and Solf treatise that suggests that individuals working within a legitimate military target must be considered in weighing the concept or proportionality.” And thus, unsurprisingly, his “Nutshell” would have excluded the following persons, among others, from any “determination of collateral civilian casualties”: “civilians injured or killed while working in or immediately adjacent to a lawful target,” and “civilians injured or killed as a result of the enemy placing them around a lawful target in an effort to shield it from attack.”

Fortunately, the new DoD Law of War Manual, like its predecessor, soundly rejects Parks’s general view that the rule of proportionality is not a customary requirement that might preclude what would otherwise be lawful attacks on military targets. Section 5.12 provides, unambiguously, that “[c]ombatants must refrain from attacks in which the expected loss of life or injury to civilians, and damage to civilian objects incidental to the attack, would be excessive in relation to the concrete and direct military advantage expected to be gained. This is commonly called the proportionality rule.” This proportionality requirement is repeated throughout the Manual. See, for example, Sections 4.8.2; 5.3.2, (applying it to attacks on passenger vessels that are military objectives), (to blockades), (to attacks on passenger aircraft that are military objectives), (to cyber operations), and 17.7 (in NIACs).

Unfortunately, however, the Manual retains — and, as shown above, invokes repeatedly — Parks’s notion that “the expected loss of life or injury to civilians” or other protected persons should not count in the proportionality assessment if the persons in question are knowingly within or nearby a lawful target. Moreover, the provocation in Section 5.1.3 — the suggestion that “[c]ivilians also may share in the responsibility to take precautions for their own protection” — emerges directly from Parks’s 1990 article, in which he wrote “that the individual civilian has an obligation to remove himself from the vicinity of a military objective.”

*          *          *

Oona is right: The Manual’s “assumption of risk” proposition as to the requirement of proportionality is indefensible, and threatens to unravel the proportionality rule through the back door. If an “assumption of risk” were sufficient to disqualify a civilian from consideration in a proportionality analysis, that would render the proportionality requirement virtually meaningless or inapposite, especially in light of the United States’s expansive notion of permissible military objectives in noninternational conflicts, to include many “war-sustaining” facilities and resources. After all, most of the civilians, and other protected persons, who find themselves in or near a terrorist safe house in a residential neighborhood, or in the vicinity of, say, ISIL oil production facilities, oil tankers, and “banks,” are now on notice that those are lawful military targets. Therefore, if the risk they “assume” by staying near such potential targets were, in effect, to disqualify them from the proportionality calculus, then it’s difficult to see how the military would be required to assess virtually any anticipated harm to civilians in our current armed conflicts, in order to determine whether it would be excessive in relation to the anticipated concrete and direct military advantage.

The better view — indeed, the only defensible account of the current customary law — is that although the presence of such civilians and other protected parties in or near a lawful military target does not rule out the use of force against that target per se, the commander in charge (and her JAG(s)) assuredly must consider the anticipated harm to those civilians in the required proportionality analysis. It is, after all, the very purpose of the proportionality rule to protect civilians who are in the vicinity of lawful targets.

*          *          *

I don’t think there’s any realistic likelihood that the Manual’s embrace of Hays Parks’s “assumption of risk” understanding of the proportionality rule will find any traction in the development of customary law: No other nations are likely to adopt it. Indeed, allies such as Israel and Great Britain have already rejected it. (For Israel, see paragraph 42 of this decision of the Israel Supreme Court, and paragraphs 46, 306, 307, 324 of this Report of the Israel Defense Forces on the 2014 Operation Protective Edge in Gaza. For the UK, see paragraph 2.5.2 of its Manual of the Law of Armed Conflict.)

Moreover, that understanding is not even the official view of the United States. As the preface to the Manual explains, the Manual is simply “an institutional publication and reflects the views of the Department of Defense.” Although it “has benefited significantly from the participation of experts from the Department of State, Office of the Legal Adviser, and the Department of Justice, Office of Legal Counsel, . . . the views in this manual do not necessarily reflect the views of those Departments or the U.S. Government as a whole.” (From the outside one can only hazard a guess, but it would not be surprising if the rest of the government declined to endorse the Manual, notwithstanding its many virtues, precisely because of provisions such as those I’ve described here.) And if the United States government itself does not vouch for these provisions of the Manual, it stands to reason that no other state will do so, either.

My principal concern is thus not that the appearance of the “Parks rule of assumed risk” in the Manual will affect the development of custom, but instead that it might have a serious, damaging affect on how officers in the U.S. military themselves understand their obligations under the rule of proportionality. If they entirely discount any anticipated harm to persons who are found near military targets based upon an “acceptance of risk” rationale, they will not adequately take into account those risks, even if they do presume that such persons are civilians rather than combatants. That way lies the effective demise of the proportionality rule, which the Manual otherwise takes such pains to honor. Moreover, even if military officers are not, in fact, applying the “assumption of risk” reasoning in the Manual, its prominence in that document will, at a minimum, exacerbate the recent doubts of some observers about the extent to which the United States complies with its humanitarian law obligations. It is therefore important for DoD to disclaim the rationale, if only to demonstrate plainly what the very first sentence of the Manual promises: that “[t]he law of war is of fundamental importance to the Armed Forces of the United States.”

Fortunately, the Manual itself invites “comments and suggestions from users of the DoD Law of War Manual,” which should be addressed by email to:


As Chuck Allen emphasized at our Georgetown event, the Department is taking such comments and suggestions very seriously, and intends to promulgate updated editions of the Manual that may reflect the Department’s responses to those submissions. So, if you agree with Oona and me — or if you notice other problems or ambiguities in the Manual — keep those cards and letters coming . . .


* Fortunately, as Adil Haque has recently blogged here, it is possible that the Department of Defense has not taken to heart this particular statement in the Manual: The recent DoD report on the 2015 airstrike of the Médicins Sans Frontières (MSF) hospital in Kunduz, Afghanistan proceeds on exactly the opposite view—that until the MSF Trauma Center was confirmed as a lawful target “it should have been presumed to be a civilian compound.”