An important story in yesterday’s New York Times explains how the U.S. and coalition forces have dramatically increased their targeting of ISIL’s oil facilities (including oil trucks and oil wells) and ISIL’s “bulk cash stockpiles,” which are found in places such as “bank vaults [and] private residences.” “The destruction in recent months of these targets, deep behind enemy lines, which commanders previously avoided for fear of causing civilian casualties — has seriously damaged the Islamic State’s ability to pay its fighters, govern and attract new recruits, military officials say,” reports the Times.

Because ISIL, unlike al Qaeda, is not exclusively a military organization — it holds territory and engages in at least some “civilian” governance functions — the targeting of these facilities and stockpiles raises important issues under the laws of armed conflict. (In an earlier post I discussed how this distinction between AQ and ISIL might complicate the potential targeting and detention of individuals who are “part of” ISIL.) The customary law of armed conflict, reflected in Article 52 of Additional Protocol I to the Geneva Conventions, provides that attacks on objects must be “limited strictly to military objectives,” i.e, “to those objects which by their nature, location, purpose or use make an effective contribution to military action and whose total or partial destruction, capture or neutralization, in the circumstances ruling at the time, offers a definite military advantage.” But what if the anticipated military advantage, although “definite,” is also indirect? In particular, what if that advantage consists of the anticipated diminution of the enemy’s revenues, some (but not all) of which are used to support military operations? Does the fact that some of the proceeds of a particular industry or operation contribute to the war effort mean that the operation itself is a “military objective” and thus a legitimate target? And, if so, does this mean that virtually all economic enterprises are legitimate targets, simply because of the indirect advantages they offer to the military arm of the state?

An important new article by Just Security’s Co-Editor-in-Chief (on leave) Ryan Goodman is the best treatment I’ve seen of this difficult and increasingly significant topic. (Goodman is currently Special Counsel to the General Counsel of the Department of Defense, but the article is written in his personal capacity.)

Ryan demonstrates that, contrary to the views of many commentators, the laws of war are not best read to categorically prohibit the targeting of operations that are war-sustaining by virtue of the revenue that they raise. Or, at the very least, such a categorical prohibition would be difficult to reconcile with important historical precedents that states have not clearly repudiated, including the Union’s destruction of raw cotton in the Civil War because its sale provided funds for the Confederacy’s purchase of arms and ammunition. (More specifically, Ryan shows that many states, as well as the highly authoritative Bothe, Partsch and Solf Commentaries on the First Geneva Protocol, have adopted the view that the example of cotton in the Civil War demonstrates that at least some revenue-generating objects can make an “effective contribution” to military action, at least under some circumstances.)

Ryan also identifies, however, at least four difficult LOAC issues that such targeting implicates. One of them relates to the principle of distinction itself (is the target a military objective that can be attacked?). The other three involve how to assess whether a strike upon a war-sustaining, revenue-enhancing operation satisfies the principle of proportionality, even if it is a permissible military objective.

First, what’s the limiting principle? After all, within a state engaged in armed conflict, virtually all industries and operations enhance the war effort: some of the taxes or other revenues that they generate eventually end up supporting, or sustaining, military activities, somewhere down the line. More broadly still: Destroying or degrading a state’s economic infrastructure can obviously degrade that state’s ability to fight, at least in the long run. Is that entire infrastructure thus subject to lawful targeting? (It might seem intuitive that the U.S. should be able to target ISIL’s oil tankers and “cash stockpiles.” But what would you think if a state at war with the U.S. were to target Apple headquarters, or various other tax-generating operations in Silicon Valley?)  In a 2014 post here, Ken Watkin emphasized such “slippery slope” concerns as the principal reason why many commentators and others have been increasingly anxious about the legality of targeting what we might call “indirectly-war-sustaining” revenue-generating operations.

Ryan offers two possible limiting principles. First, focusing on the fact that such targeting is permissible only if it will foreseeably result in a “definite military advantage,” he suggests that “potential substitution effects” must be taken into account: “If a source of economic support to a military can be easily substituted by another source, the military advantage gained from the destruction or neutralization of the former is presumably more speculative. . . . [T]oward one end of the spectrum was the Confederacy’s direct trade of cotton for armaments. Toward the other end of the spectrum would be a private economic activity which helps generate economic growth or expand the tax base.” Relatedly, Ryan (citing Sandesh Sivakumaran) flags the possibility of a “great staple” limitation, which might be all that the historical precedents would support:

The three clearest and primary historic examples in support of targeting war-sustaining revenue streams — cotton (Confederacy), narcotics (Taliban), and petroleum (ISIL) — are very limited cases and involved an export that was a crucial component of an enemy armed forces’ capabilities. Professor Sandesh Sivakumaran suggests that the cotton example involves a potential limiting principle: “The specifics of the Confederates-cotton nexus makes it difficult to extrapolate to other situations, as is evident from the following passage [from the 1870 arbitral tribunal]: ‘That cotton in the insurrectionary States was peculiarly and eminently a legitimate subject for such destruction, from its relation to the enemy’s government, as the great staple from which were derived the principal means of that government for the carrying on the war.’ . . . As such the Cotton Claims position cannot support the suggestion that any war-sustaining objects can be attacked.” Similarly, references to the Civil War example in the Air Force and Navy military manuals note that cotton provided for importing “almost all Confederate arms and ammunition.” In other words, a limiting principle might be that the economic product constitutes an indispensable and principal source for directly maintaining military action.

The three other vexing questions relate to the principle of proportionality, which prohibits the attack on a legitimate military objective if it “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” (quoting Article 51(5)(b) of the First Geneva Protocol).

i. In most or all of the cases in question, the targeted operation or entity is what we might call “dual-use”: Although it sustains military capabilities by virtue of the revenue generated, its predominant uses, and value, are civilian in character. Does the “civilian value” of the targeted entity count in the proportionality analysis and, if so, if that value is greater to the civilian population than the military value, is it unlawful to target the entity? “For example,” Ryan writes, “if a revenue-generating industry is considered a military objective due to funds distributed to armed forces, should a proportionality analysis include the percentage of funds distributed to nonmilitary purposes (such as running civilian hospitals, schools, etc.)?”

Ryan does not attempt to answer this question. One recent, thus-far-obscure ICTY decision, however, is (to my knowledge) the first judicial precedent that might offer some potential (albeit controversial) guidance on it. In Prosecutor v. Prlic, IT-04-74 (May 2013), three of the four judges of an ICTY Trial Chamber held that the 1993 attack by Croatia on the Stari Most — the Old Bridge that connects the two parts of the City of Mostar in Bosnia and Herzegovina — violated the principle of proportionality, even though the bridge was a legitimate military objective of the Croats, and even though the attack “may have been justified by military necessity,” because the anticipated harms to the civilian uses of the bridge outweighed the anticipated military advantage to be gained by destroying it. The court majority reasoned as follows (paras. 1582-84):

1582. The Chamber established that the Old Bridge, real property normally used by civilians, was used by both the ABiH [the Army of the Republic of Bosnia and Herzegovina] and the inhabitants of the right and left banks of the Neretva between May and November 1993 as a means of communication and supply. In this respect, it considers that the Old Bridge was essential to the ABiH for combat activities of its units on the front line, for evacuations, for the sending of troops, food and material, and that it was indeed utilised to this end. Furthermore, the ABiH was holding positions in the immediate vicinity of the Old Bridge. For this reason, the armed forces of the HVO had a military interest in destroying this structure since its destruction cut off practically all possibilities for the ABiH to continue its supply operations. Consequently, at the time of the attack, the Old Bridge was a military target.

1583. The Chamber, however, also noted that the destruction of the Old Bridge put the residents of Donja Mahala, the Muslim enclave on the right bank of the Neretva, in virtually total isolation, making it impossible for them to get food and medical supplies resulting in a serious deterioration of the humanitarian situation for the population living there. The Chamber determined that there were very few supply routes available to the inhabitants, other than the Old Bridge; that between May and November 1993, in addition to the Old Bridge, they could only use the Kamenica bridge, a makeshift bridge constructed by the ABiH in March 1993 and used until November 1993, or a path over the mountain from the neighbourhood of Donja Mahala to Jablanica, which was considered very dangerous; and that, as such, the destruction of the Kamenica bridge by the armed forces of [Croatia] on 10, 11 or 17 November 1993, that is, only a few days after the destruction of the Old Bridge, cut off all access across the Neretva River in Mostar definitively. The Chamber also determined that the destruction of the Old Bridge had a very significant psychological impact on the Muslim population of Mostar.

1584. The Chamber therefore holds that although the destruction of the Old Bridge by the [Croatian armed forces] may have been justified by military necessity, the damage to the civilian population was indisputable and substantial. It therefore holds by a majority, with Judge Antonetti dissenting, that the impact on the Muslim civilian population of Mostar was disproportionate to the concrete and direct military advantage expected by the destruction of the Old Bridge.

Judge Antonetti, dissenting, objected to the majority’s basic assumption that the civilian value of a legitimate military objective must be weighed against the anticipated military advantage under the principle of proportionality: “[T]o my mind,” he wrote, “the Old Bridge was a legitimate military objective whose destruction gave the [Croatian armed forces] a definite military advantage by cutting off communications and the supply of food. I fail to see how the principle of proportionality could be applicable in this case. If the Old Bridge was a military objective, it quite simply had to be destroyed. In any event, there is no such thing as proportionate destruction.”

The Prlic decision is currently on appeal to the appeals chamber of the ICTY. The question of whether the destruction of the Old Bridge violated the principle of proportionality because of the overwhelming value of its civilian uses is only one of numerous issues on the appeal; it’s barely mentioned in the briefs. (There’s a fairly desultory reference to the issue in Slobodan Praljak’s brief: “The [Trial Chamber] committed an error when it applied the principle of proportionality to the destruction of the military target itself. The principle of proportionality does not protect military objectives, even when they primary purpose was not military.”) Whether or not the ICTY offers additional guidance on the question, Ryan Goodman is certainly right to flag it as an important issue in the years to come.

ii. What about the persons who work in the targeted facilities or, for example, the individuals who drive the oil rigs? For purposes of the proportionality analysis, should they be treated as civilians in cases where the target’s war-sustaining contribution is indirect? This question has been previously flagged here and on Lawfare by Beth van Schaack, Ryan Santicola, and Aurel Sari, among others. At least two provisions in the new DoD Law of War Manual (see notes 77 & 78 of Ryan’s article) suggest that at least some such persons need not be considered on the “civilian” side of the proportionality analysis, either because they are said to be directly participating in hostilities, or because they have assumed a risk by taking a job within a legitimate military objective. Ryan suggests (without quite saying so expressly) that the Manual is wrong, and that the anticipated harm to such persons must be taken into account in the proportionality analysis.

iii. Finally, Ryan emphasizes an important aspect of the proportionality principle that could have a significant impact on the lawfulness of targeting “war-sustaining” objectives: In the cases in question, the object is targeted because its destruction will foreseeably result in concrete but indirect military advantage to the attacking party (in that some of the money generated will find its way into the military’s coffers). The proportionality principle, however, permits only “concrete and direct military advantage” to be counted on the “military advantage” side of the comparative “excessiveness” balance. As Ryan points out, this might mean that, in many of the cases in question, such as attacks on oil facilities and cash reserves, there is no “concrete and direct military advantage” to be weighed — the advantage is concrete, and might not be hypothetical or speculative, but it’s not direct. Moreover, as he points out, “[t]he Commentaries to Protocol I suggest that these textual choices were deliberate: ‘the words “concrete and direct” impose stricter conditions on the attacker than those implied by the criteria defining military objectives in Article 52 (General protection of civilian objects), paragraph 2’” (quoting the API Commentaries). This might lead to very counterintuitive results, Ryan suggests. For example, “would an attacker have to refrain from striking indirect contributions to military action—such as communications or transportation lines—if one civilian’s life were at risk and the definite military advantage was enormous? Have States ever applied such an understanding in practice?” Ryan does not (yet) attempt to answer such questions (apart from cryptically offering that “[a] more defensible and productive line of analysis may be found in the standard application of proportionality discussed above”). He’s absolutely right, however, to flag this as a very important puzzle going forward.