A Squarable Circle?: The Revised DoD Law of War Manual and the Challenge of Human Shields

In a new update to its Law of War (LoW) Manual, the U.S. Department of Defense kept its promise to be, as the DoD General Counsel (GC) Hon. Jennifer O’Connor reiterated at NYU in late November, “a living document.”  Commendably, from its very issuance DoD has affirmatively sought comment and input, and its first revision last May reflects its willingness to act on those inputs.

This latest change continues that effort by attempting to address a number of issues.  It especially focuses on protecting civilians, and provides more detail on how the proportionality rule and the obligation to take “feasible precautions” should operate to protect them.  According to the GC, this version of the Manual “provides greater clarity [as to protecting civilians] and also reflects important developments, such as the president’s recent executive order on civilian casualties.”

(Parenthetically, it’s unfortunate that the GC chose to reference President Obama on the eve of a new administration committed to a more aggressive use of force than the policy-restrained rules currently in place.  I believe her words alone could cause the revised Manual to draw more scrutiny than might otherwise have been the case.)

I was particularly interested in how the revision addresses an issue that has generated a lot of thoughtful discussion on Just Security: human shields.  In particular, the revision grapples with clarifying one of the most controversial aspects of the Manual, that is, its interpretation of proportionality vis-à-vis certain persons, including human shields.

Why the controversy?  The original version of the Manual concluded that harm to human shields would “not prohibit attacks under the proportionality rule” by reasoning that:

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. [§5.12.3.3].

I believe that this is a correct, if unsettling, view of the law – even though current U.S. use of force policies affecting human shields appears to be far more restrictive than international law would require.

The Islamic State in Iraq and Syria is keenly aware of those restrictions and exploits them to a degree rarely seen in history.  As journalist Judit Neurink puts it, ISIS is bringing “the use of human shields to a completely new level,” And we’ve seen that they are enjoying a significant military advantage as a result.  USA Today reported last August that “hundreds of [ISIS] fighters escaped Manbij in northern Syria by placing civilians in a convoy of 500 vehicles” adding that the “Pentagon said it didn’t fire on the convoy for fear of hitting civilians.”

My concern is that these more-than-what-the-law requires policies are getting civilians killed.  Even though not firing on the convoy “for fear of hitting civilians” will ensure no civilians are killed at that point in time, the cruel truth is that “hundreds” of ISIS fighters who would likely have been eliminated survived to wreak all kinds of horrors on other  civilians whom the LoW aims to protect.  This kind of self-imposed restriction produces a phenomena I call the “moral hazard of inaction.”

While the Manual’s revision does not reference that phraseology, it does seem to pay some homage to the idea by including in its discussion of the evaluation of expected incidental harm consideration of “civilians at risk if the attack is not taken” (§5.12.3).  In my opinion this is the kind of strategic thinking some critics will not like, but to me it is a vitally important step (and a rather courageous one) if we are to seriously try to eliminate the most dangerous threats to civilians.

Is it enough?  The grim truth is that ISIS (and, we should expect, other such groups) is exploiting not just popular interpretations of the law of human shields, but also the policies which further restrict the use of force.  When a warfighting entity is able to use (or exploit) policies or political directives that exceed what the law would dictate as effective substitutes for traditional military means, they can be said to be waging what I would call polifare.

Militarily speaking, ISIS’s proliferate use of human shields has been extraordinarily successful in actually defending their forces from attack.  I’m convinced that human shields protect ISIS much more effectively than any anti-aircraft systems they might have been able to muster, so we can only expect the tactic to metastasize further.  And because ISIS is indifferent to the LoW, the international community has been utterly unable to limit, let alone stop, the practice from expanding rapidly.

Nevertheless, the Manual revision deletes the controversial paragraph noted above (§5.12.3.3) about proportionality and human shields.  What does the elimination mean?  A Pentagon official speaking on background earlier this week insisted DoD is not abandoning the concept formerly expressed, which, on my view, is sourced in the international law maxim of ex injuria non jus oritur or “legal rights should not be understood to result from the commission of wrongful acts.”

However, the GC’s speech at NYU in November said that the then forthcoming revision was intended to:

…provide greater clarity on the DoD legal view of human shields, noting that whether voluntary or involuntary, civilian ‘human shields’ would not be considered to be directly participating in hostilities and would not thereby lose their protections from attack. (emphasis added.)

This seems to say that in her view even voluntary human shields are not directly participating in hostilities and would not lose their protection.  It’s hard to figure how such a characterization would not “perversely encourage the use of human shields.”  Indeed, if voluntary human shields retain their protected status, why wouldn’t unscrupulous adversaries like ISIS employ them in even greater numbers since human shields have proved to be exceptionally effective in deterring the U.S. and its allies from strikes against bona fide military targets?

Yesterday Hays Parks reinforced to me essentially what he has said in the past:

If the U.S. rewards an enemy’s use of civilians as human shields to protect lawful targets of attack by increasing constraints on their attack by U.S. military forces well beyond what is required by the law of war, it rewards (rather than condemning) the enemy for its illegal hostage taking. Ignoring history, President Obama’s restrictions have placed innocent civilians at greater risk…In placing full responsibility on U.S. Military forces for civilian casualty avoidance, President Obama has assured enemy use of human shields, increased risk to U.S. Military forces, and risk of mission failure that otherwise complies with the law of war.

Obviously, his concerns are exacerbated if voluntary human shields are given civilian protections.  Let’s be frank, how can it be said that someone knowingly, actively, and – especially – voluntarily attempting to shield an otherwise legitimate military target from attack is doing anything other than directly participating in hostilities (DPH)?

Isn’t protecting a target something that, as the ICRC might put it, is “directly harming [the attacker’s] military operation”?  Isn’t deliberately shielding lawful targets an action that should, as a matter of law, cause those civilians doing so to lose their protection from attack?  (To be sure, there may be a myriad of political and even strategic reasons not to attack them in a specific – or even most – cases, but the question here is their status under law.)

Fortunately, in what may be an illustration of DoD’s continuing effort to refine its views, the interpretation represented in the revision does not go as far as the GC’s statement in November, at least insofar as voluntary human shields are concerned.  Specifically, it says in part (§5.12.3.4):

[T]he enemy use of voluntary human shields may be considered as a factor in assessing the legality of an attack. Based on the facts and circumstances of a particular case, the commander may determine that persons characterized as voluntary human shields are taking a direct part in hostilities. (Emphasis added.)

This is entirely sensible; in fact, I think it would undermine the legitimacy of LoW if warfighters were told that legally-protected civilians were nevertheless allowed to volunteer to be human shields and thus effectively defend legitimate military targets.

I believe, however, we need to go beyond the relatively easy case of voluntary human shields and ask this question: can’t people be DPH, even if they don’t want to be?  Actually, this whole discussion of voluntary versus involuntary DPH raises an interesting question: since when does the mental state of an actor really determine DPH?  For example, Article 51(3) of Additional Protocol I presents the law as being activity based, not intent based, by indicating that civilians lose their protection from direct attack “for such time as they take a direct part in hostilities.” Nothing about whether or not the “taking part” was voluntary or involuntary.  Rather, civilians are judged by their actions.

As Ryan has pointed out to me, the ICRC’s Interpretive Guidance on the Notion of Direct Participation in Hostilities Under International Humanitarian Law does admit that:

As an objective criterion linked to the act alone, belligerent nexus is generally not influenced by factors such as personal distress or preferences, or by the mental ability or willingness of persons to assume responsibility for their conduct.  Accordingly, even civilians forced to directly participate in hostilities or children below the lawful recruitment age may lose protection against direct attack. (pp. 59-60, citations omitted)

However, as Ryan also noted, the Interpretive Guidance goes on to claim that in “exceptional situations…the mental state of civilians [could] call into question the belligerent nexus of their conduct.”  According to the ICRC, an example would be when civilians “are completely deprived of their physical freedom of action (e.g. when they are involuntary human shields physically coerced into providing cover in close combat).”  In such circumstances, the ICRC asserts, they “remain protected against direct attack despite the belligerent nexus” and “would have to be taken into account in the proportionality assessment during any military operation likely to inflict incidental harm on them.”

Putting aside the fact that there could be many strategic or humanitarian reasons to follow the Interpretive Guidance in a particular situation as a matter of prudent policy, it is still useful to remind ourselves that by its own terms, the Interpretive Guidance represents only the ICRC’s institutional position and, again by its own terms, “does not necessarily reflect a unanimous view or majority opinion of the experts” it consulted during its drafting process.

In this instance I must part company with the Interpretive Guidance for a number of reasons, starting with the sheer impracticality if not literal impossibility of a battlespace determination as to which persons actively shielding a legitimate target might be doing so only because they have been coerced, and that coercion has reached the level of depriving them of “physical freedom of action.”  Wouldn’t that be subjective as to whether or not they’ve been deprived of physical freedom of action?  Some people, for example, might be cowed rather more easily than others.  How can we see into their minds?

The ICRC seems to contend that there is some kind of obligation not only to read the mind of a shield, but also to make a psychological distinction as to whether the civilian is merely “forced” into DPH (in which case the ICRC seems to say the person is targetable and need not be accounted for in the proportionality analysis) or whether that coercion rises to the level in the mind of that particular person to a complete deprivation of “physical freedom of action.”

Put another way, how can we possibly expect 19 year olds in the midst of the chaos of combat to ever perform a complex psychological analysis (that could stymie a highly-trained psychiatrist) of an individual plainly engaged in DPH?  Let’s not forget the reality of combat that the Supreme Court pointed out in Chappell v. Wallace (1983): “The inescapable demands of military discipline and obedience to orders cannot be taught on battlefields; the habit of immediate compliance with military procedures and orders must be virtually reflex, with no time for debate or reflection.” (emphasis added)

As I have said elsewhere, law, and especially the LoW, depends upon its being rational and workable in the context in which it is sought to be applied.  No good can come from warfighters concluding that the law is out of touch with the harsh realities they see on today’s battlefields.  In this application the Interpretive Guidance’s effort to require attackers to psychologically analyze persons doing acts it concedes could be DPH simply cannot – and should not – be permitted to effectively facilitate the use of human shields by unprincipled actors.

More broadly, how much should voluntariness of participants even matter in armed conflicts?  My sense is that it’s well accepted that involuntary conscripts in the armies of nation-states are not, for example, given any sort of benefit under the LoW merely for the fact that they are compelled – sometimes under threat of capital punishment – to become members of the armed forces and thus targetable combatants.

If the ICRC basically wants to excuse civilians from responsibility for their actions on the theory that those civilians are thought to be “completely deprived of their physical freedom of action,” how does that situation differ from the compulsion put upon a conscript who, under U.S. law for example, can face death for disobedience, desertion, and a whole range of “close combat” type offenses, including running away?  Or are we to suppose that if ISIS “conscripted” unwilling civilians into being human shields, and perhaps clothed them with a uniform of sorts and declared them as their fighters, the ICRC would then consider them targetable persons?

Interestingly, Michael Gross in a fascinating 2014 essay on the Carnegie Council’s Ethics & International Law blog, argues that in “a just guerrilla war…[human] shielding has its legitimate uses.”  Analogizing involuntary human shields to nation-state conscription, he argues that they provide “guerrillas with a low cost and effective tactic to counter the military might of their adversaries.”  According to Gross, human shielding can be a “permissible strategy of war.”

Despite the wording of the revision, a DoD official insisted that the revision was “not trying to create an ‘intent’ test’ for DPH,” and maintained it was “not trying to draw a legal distinction between voluntariness and involuntariness.”   Maybe so, but that position doesn’t seem to square very neatly with the GC’s pronouncement or the explicit language in the revision reference to “voluntary” human shields.

Is it possible to square the circle of the challenge of human shields without getting bogged down in a DPH debate?  In a new essay (mostly written before the Manual revision) entitled “No Good Options Against ISIS Barbarism? Addressing the Conundrums of Human Shields in 21st Century Conflicts” and soon to appear (early January 2017) online as part of a symposium issue of the American Journal of International Law “UnBound” feature, I offer what might be something of a compromise position.

Specifically, even assuming all human shields are protected “civilians” for the purpose of determining if anticipated losses are “excessive in relation to the concrete and direct military advantage anticipated,” it may be possible to produce an approach that would discourage the use of the tactic, while reasonably protecting civilians in the near term and, especially, in the longer term.

How?  Given the military effectiveness of today’s epidemic of human shields, in determining the “concrete and direct military advantage anticipated” in conducting a specific attack, a commander might conclude that discouraging the use of human shields is one of those advantages anticipated.  That is, the U.S. could interpret as a proper military objective the neutralization of the effectiveness of the enemy’s use of human shields where it creates an illicit sanctuary for otherwise legitimate military targets.

Accordingly, if the aim is to demonstrate to the enemy the futility of using human shields as a way of inducing him to abandon the tactic, accounting for human shields in a proportionality analysis might produce a different conclusion as to what constitutes “excessive” civilian deaths than what might otherwise be the case were human shields not part of the equation.  Of course, it would need to be reasonably anticipated that an enemy’s use of human shields could be diminished in this way, and that there is a reasonable belief that doing so would save more civilian lives during the conflict than might otherwise be lost.

Candidly, this could mean accepting considerable tolerance for civilian casualties in the short term in the effort to get the enemy to stop defending their fighters and assets through the use of human shields.  In my opinion, it’s imperative to remove from adversaries like ISIS the incentive that exists today to use human shields.  We just can’t continue to allow convoys of “hundreds” of ISIS fighters to escape to inflict more human misery on the powerless, particularly since the current efforts simply aren’t working.

Would the Manual revision permit tackling the human shield problem in the way I propose?  Maybe.  After all, the drafters cracked the door open by including in §5.12.3.4 that the “enemy use of voluntary human shields may be considered as a factor in assessing the legality of an attack.”  In addition, the discussion of consideration of “civilians at risk if the attack is not taken” in §5.12.3 suggests a recognition that the short term benefits to some civilians by refraining from an attack where human shields are present may, in fact, cause more civilian losses overall.

Kluging together these snippets is evidence that what the Pentagon official indicated to me, that is, that there was no fundamental change in legal position, is accurate.  Obviously, these parts of the text show DoD is aggressively looking for a way of lawfully dealing with the uniquely perplexing issue of human shields.  My suggested approach doesn’t require finding human shields as DPH, but rather focuses on allowing an assessment of their effectiveness in particular situations.

As discussed, that assessment could weigh in the determination of whether expected human-shield losses would be “excessive” in relation to the anticipated military advantage of halting a tactic that is today successfully defending ISIS targets.  I think there might be room in the revision as written to square, as best as possible, the “circle” of the human shield quandary.

In other words, with some innovative application of the proportionality analysis (as opposed to revising the rule itself) as well as some re-evaluation of the nature of “military advantage” in light of current circumstances, it could be possible to avoid “perversely encourag[ing] the use of human shields” which seems to have been the Manual’s aim all along.

So why the deletion of what had been a rather clear position set out in §5.12.3.3 of the original Manual?  My guess is that in the prior version of the Manual it was simply too blunt, and left some readers with the unwarranted impression that the U.S. was not sensitive enough to civilian losses.  The elimination of the easily-misconstrued paragraph, the re-wording elsewhere, and the (almost too heavy-handed) emphasis in the revision on “feasible precautions” ought to go a long way toward correcting that misimpression.

To me, the revision reflects a continuing awareness by DoD that in the real world the Manual speaks to audiences larger than just American warfighters.  That is certainly a laudable – and wise – effort.  At the same time, caution is indicated in that efforts to satiate critics can become unproductively quixotic. In a 1,200+ page document, there will always be opportunities for criticism, to include disparagement by those who have no responsibility for actually winning wars in a lawful fashion, and who may not have fully internalized an understanding of the technologies and mechanics of modern war.

As really important as the effort to answer critics’ charges – and to assure publics around the world – is, the Manual should not be allowed to morph into an opaque document that tries to be all things to all people. If that happens,  in the end it could fail to provide the level of clarity and lucidity honorable warfighters operating under great stress in complex battlespaces really need.

That said, I applaud the new revision, even if imperfect, as a reflection of a sincere “let’s-make-it-better” attitude on the part of DoD – something too rare in governments anywhere.  It is crucial that when we think about trying to solve issues (like trying to square the circle with respect to dilemmas like human shields) that we be open to new ideas and approaches, even radical ones.  It’s vital, it seems to me, for all of us to keep in mind the admonition that the International Military Tribunal at Nuremberg gave us, that is, that the “law is not static, but by continual adaption follows the needs of a changing world.

  

About the Author(s)

Charles J. Dunlap, Jr.

Professor of the Practice of Law and Executive Director of the Center on Law, Ethics and National Security at Duke Law School He retired from the Air Force in 2010 as a Major General.