NATO Interoperability When Targeting Indispensable Objects
A recent Lieber Paper strikes an optimistic note concerning legal interoperability, arguing that the dissimilarities stemming from uneven treaty membership between NATO allies are less significant than they may seem. However, one contentious subject has the potential to rattle the legal balance between members of the alliance. The U.S. Department of Defense Law of War Manual (DoD Manual) provides a more permissive (yet ambiguous) position than (most) NATO allies concerning the obligation to refrain from targeting objects indispensable to the survival of the civilian population.
This post outlines the U.S. position in comparison with the obligations contained in Article 54 paragraphs 2 and 3 of Additional Protocol I to the 1949 Geneva Conventions (AP I) and proposes a route toward an interoperability compromise for NATO members. It does not seek to exhaust all the layers of complexity originating from the various interpretations of the AP I obligation between member States. Instead, the analysis addresses the overarching positions concerning the obligation to spare objects indispensable to the survival of the civilian population. This leads to the difficult question at the heart of the matter: in a joint operation, whose law applies?
The Prohibition in AP I Article 54
Targeting the natural environment and indispensable supplies to weaken enemy forces has long been a widespread practice in war. Historically, even extreme measures aimed at starving the adversary (including the civilian population) into submission were considered permissible as a matter of customary international law (e.g., Lieber Code, arts. 17, 18). Only in 1919 did the report of the Commission on the Responsibility of the Authors of the War and on Enforcement of Penalties inaugurate a dramatic change, bringing civilian protection into focus and listing the deliberate starvation of civilians amongst the violations of international law committed by the authors of the First World War.
After the Second World War, the 1949 Fourth Geneva Convention (GC IV) introduced novel restrictions on the scale of harm civilians may suffer in times of armed conflict. The 1977 Additional Protocols then increased belligerent States’ humanitarian obligations once more. Article 54 of AP I introduced a new constraint of particular relevance to siege warfare and a significant addition to the legal regime protecting civilians: the prohibition on starving civilians as a method of warfare.
Though there is no universal consensus on how the prohibition should be interpreted, its wording can generally be divided into two parts. The first paragraph of the article contains a general prohibition, while the following paragraphs enumerate specific examples of prohibited starvation practices. Article 54 provides in relevant part as follows,
1. Starvation of civilians as a method of warfare is prohibited.
2. It is prohibited to attack, destroy, remove or render useless objects indispensable to the survival of the civilian population, such as food-stuffs, agricultural areas for the production of food-stuffs, crops, livestock, drinking water installations and supplies and irrigation works, for the specific purpose of denying them for their sustenance value to the civilian population or to the adverse Party, whatever the motive, whether in order to starve out civilians, to cause them to move away, or for any other motive.
3. The prohibitions in paragraph 2 shall not apply to such of the objects covered by it as are used by an adverse Party:
(a) as sustenance solely for the members of its armed forces; or
(b) if not as sustenance, then in direct support of military action, provided, however, that in no event shall actions against these objects be taken which may be expected to leave the civilian population with such inadequate food or water as to cause its starvation or force its movement.
… .
This framework prohibiting starvation as a method of warfare and attacks on objects indispensable to the survival of the civilian population is generally considered customary in character (see International Committee of the Red Cross, (ICRC), Study of Customary IHL, rules 53, 54). However, the exact reach of its prohibitive effect is subject to much debate, and there is no agreement on how the respective paragraphs of the article should be read.
In short, two overarching views persist. The first considers the wording of paragraph 1 to be purposive in nature, leading proponents of this view to argue that any tactic not aimed at starving civilians would fail to do so as a method of warfare and therefore not violate the prohibition (e.g., Drew, p. 313-14; Watts, p. 19; Akande & Gillard, p. 762-65).
The second view draws on the wording contained in the following paragraphs and submits that the prohibition should be read as a whole, banning all practices with the potential to cause civilian starvation, irrespective of the motive behind the military action in question (e.g., Dannenbaum, p. 364-74; Gaggioli). On this view, purposefully depriving civilians of objects indispensable to their survival would be prohibited whatever the motive behind the tactic in question may be. This means incidental starvation, too, would be curtailed (Dannenbaum, p. 370).
These two views lead to significantly different legal outcomes. Under the first view, a tactic aiming to starve combatants while incidentally starving civilians would be permissible, subject to proportionality considerations. The second view would prohibit any military action foreseeably causing civilians to starve alongside combatants, irrespective of the underlying pursuit of a military goal or proportionality.
It should be noted that even parties to the Additional Protocols disagree on the exact meaning of the prohibition(s) contained in Article 54 of AP I. For instance, the UK Joint Service Manual on the Law Applicable to Armed Conflict submits that “[t]he rule on starvation does not apply where these objects [indispensable to the survival of the civilian population] are for the sole use of members of the enemy armed forces or in direct support of military action” (para. 5.27.1, emphasis added).
It appears the United Kingdom would not support the view that attacks on indispensable objects also serving a military purpose would be prohibited, even if they caused civilians to starve or forced their movement, contradicting the wording contained in paragraph 3(b) of Article 54. However, the UK Manual is somewhat self-contradicting, as its paragraph 5.19 contains an exact quote of the prohibition contained in Article 54, including paragraph 3(b), complicating how the State’s overall position should be interpreted. In any case, there are layers of complexity attaching to the rule’s interpretation. While originating from the prohibition contained in AP I, these uncertainties bleed into its customary counterpart.
Consequently, if a commander is tasked with devising a joint NATO operation on an urban battlefield involving the risk of starving civilians, attempting to find a lawful tactical approach is complicated, even among States party to AP I. This becomes even more complex when States that are not party to the Additional Protocols are involved (Turkey is the only other NATO member aside the United States that is not party to the 1977 Additional Protocols).
The U.S. approach to attacking objects indispensable to the survival of the civilian population does not reflect the rule contained in AP I. Instead, the DoD Manual maintains the following position.
5.20.1 Starvation – Distinction. It is a legitimate method of war to starve enemy forces. For example, it is permitted to destroy food intended as sustenance for enemy forces with a view towards weakening them and diverting their resources. Enemy forces, for the purpose of this rule, means those persons constituting military objectives …
5.20.4 AP I Provision on Objects Indispensable to the Survival of the Civilian Population …
[Article 54(2) of AP I] would not apply to attacks that are carried out for specific purposes other than to deny sustenance. For example, this rule would not prohibit destroying a field of crops to prevent it from being used as concealment by the enemy or destroying a supply route that is used to move military supplies but is also used to supply the civilian population with food.
Similarly, this AP I prohibition does not apply to objects that would otherwise be covered by it if those objects are used by an adverse party “as sustenance solely for the members of its armed forces” or “if not as sustenance, then in direct support of military action.” Actions against this latter category of objects forfeiting protection, however, may not be taken if they “may be expected to leave the civilian population with such inadequate food or water as to cause its starvation or force its movement.”
When adopted, this AP I prohibition was novel and the product of extensive diplomatic negotiation. Given the intricacy of this provision of AP I, it would be difficult to conclude that all of its particulars reflect customary international law. Nonetheless, the United States has supported the underlying principle that starvation of civilians may not be used as a method of warfare, as discussed in the sub-sections above (emphasis added).
Though it is not obvious at first glance, the DoD Manual is at odds with the wording contained in Article 54 paragraphs 2 and 3 of AP I. Recall that paragraph 2 prohibits attacks on indispensable objects “for the specific purpose of denying them for their sustenance value … whatever the motive, whether in order to starve out civilians, to cause them to move away, or for any other motive” (emphasis added), therefore arguably barring even the pursuit of a military advantage. Complementarily, Article 54 paragraphs 3(a) and (b) only allow for such attacks in exceptional, limited cases governed by military necessity, drawing a firm line where civilian starvation may be the result of such attacks. Those provisions state, “in no event shall actions against these objects be taken which may be expected to leave the civilian population with such inadequate food or water as to cause its starvation or force its movement” (emphasis added). In this sense, Article 54 removes attacks on indispensable objects from the legal framework governing ordinary military objectives under Article 52(2) of AP I, rendering Article 54 paragraphs 2 and 3 lex specialis (Dannenbaum, p. 368-69).
By contrast, the DoD Manual account of customary international law does not replicate this protective qualification concerning objects indispensable to civilian survival. Although the DoD Manual remains somewhat ambiguous on this point, it leaves room to argue that an act not pursuing the denial of sustenance to civilians would be permissible if military necessity dictated as much. In this sense, subparagraph 5.20.4 of the DoD Manual suggests that on the U.S. view, only the deprivation of objects indispensable to the survival of civilians for the specific purpose of denying them for their sustenance value is prohibited.
The DoD Manual merely considers the wording contained in Article 54 paragraphs 2 and 3 AP I in its rules governing starvation at its intersection with the principle of distinction (see the corresponding footnote 758). Yet, overall, the DoD Manual does not confirm that objects indispensable to the survival of the civilian population may not be attacked if the civilian population would starve as a result. Generally leaning towards a purposive view of the starvation ban, the U.S. position would leave significantly more room to argue military necessity when attacking indispensable objects. It therefore appears possible that the U.S. would deem civilian starvation, either by attacks on indispensable objects or by encirclement alike, permissible as long as the expected civilian harm was not excessive in relation to anticipated military advantage. To that end, the U.S. Manual provides, “Military action intended to starve enemy forces, however, must not be taken where it is expected to result in incidental harm to the civilian population that is excessive in relation to the military advantage anticipated to be gained” (5.20.2).
Between States parties to AP I and States bound by customary IHL, all the above leads to two complicated questions. First, is belligerent intent to starve civilians a requirement to violate the prohibition? If so, uncertainties concerning the delineation of incidental starvation from intentional starvation would follow, whether in connection with attacks on objects indispensable to the survival of civilians or encirclement. At this point, States must not only clarify their positions on the starvation of civilians as a method of warfare (an issue previously addressed by IHL scholars, advocates, and the ICRC). More generally, the difficulty in differentiating thresholds of intent is an overarching problem of IHL, leading to one central question: could oblique intent, or the knowledge of causing a harmful consequence, be enough to confirm that a violation has occurred (Dannenbaum & Dill, p. 10-12)?
Toward Interoperability: a “Minimalist” or “Maximalist” Approach?
In a joint operation by the NATO alliance, these positions must be aligned under a unified framework. As Peter Olsen suggests, coming to a common understanding,
has a great effect on NATO from a practical operational perspective, because NATO’s focus is on planning and conducting operations under operation plans (OPLANs) and rules of engagement (ROE) that are consistent with the legal rules that each individual Ally and participating partner considers applicable within its legal framework (p. 655).
In light of the above, it could prove particularly challenging to come to a shared legal structure governing the boundaries of siege warfare and attacks on objects indispensable to the survival of the civilian population. This leads back to the opening question of whose view would or should be decisive in a conflict involving the NATO alliance.
Legal interoperability could follow either a “minimalist” or a “maximalist” approach. As Professor Michael Schmitt et al. summarize,
[a]ccording to a maximalist approach, the common parameters for an operation would be set according to the most restrictive rules drawn from across the coalition. … By contrast, the minimalist approach is to set the common parameters for the operation according to the least restrictive rules from across the coalition, with individual States then further restricting the actions of their respective armed forces through additional national instructions (p. 322).
Accordingly, a NATO alliance framework regarding the prohibition on starving civilians as a method of warfare, i.e., on attacking objects indispensable to the survival of the civilian population, could follow one of two routes.
First, NATO could generally agree upon a minimal threshold in line with the U.S. view, at its core merely agreeing to refrain only from purposive starvation. This, however, would not relieve NATO States party to AP I of their duties regarding attacks on objects indispensable to the survival of the civilian population contained in paragraphs 2 and 3 of Article 54. Members of the coalition bound by AP I could not attack objects indispensable to the survival of the civilian population if this “may be expected to leave the civilian population with such inadequate food or water as to cause its starvation or force its movement,” even if the object in question directly supported military action.
Although the States concerned could indicate that their forces would adhere to a higher standard by entering what NATO defines as a “caveat,” there would have to be a common understanding of the AP I rules governing starvation to avoid confusion and a patchwork of conflicting legal frameworks (Zwanenburg, p. 702-03). Overall, this option could bring NATO members into significant legal turmoil, setting aside the additional need to resolve matters of belligerent intent.
A second option, explicitly prohibiting attacks on objects indispensable to the survival of the civilian population along the lines of Article 54 paragraphs 2 and 3 would seem more practicable. Concerning the types of operation in question, the alliance would have to apply stricter obligations than the United States usually recognizes. Hence the United States would likely reject a “maximalist” approach like this, seeing as this may cause “fear that such a policy could be interpreted as acceptance of the rules that are applied as a matter of policy as customary international law” (Zwanenburg, p. 701).
This concern might be alleviated by highlighting in a joint framework that the rule would only be accepted over the course of the joint operation in question to accommodate the obligations of States party to AP I and to spare confusion. It should be added that acceptance of the AP I rule concerning attacks on objects indispensable to the survival of the civilian population need not necessarily contradict the underlying U.S. position on starvation of civilians as a prohibited method of warfare. As the U.S. Joint Chiefs of Staff noted in their commentary to AP I, Article 54 “was a new rule but was militarily acceptable because ‘there is little military need for a modern armed force to retain the option of starving the enemy’s civilian population into submission’” (p. 111, emphasis added). Along these lines, it may even be consistent for the United States to reconsider its position and to gradually adopt a stricter view on attacks against indispensable objects.
Conclusion
Returning to the above-referenced Lieber Paper analysis, it holds true that “the United States is more cautious about foreclosing military options as a matter of law than many partners with whom it is likely to operate during armed conflicts.” The interoperability of the rules binding all States versus those applying only to States parties to certain treaties requires deliberation and compromise. Currently, it appears that on the U.S. view, the purpose of attacks on objects indispensable to the survival of civilians is the central indicator distinguishing lawful attacks from unlawful ones.
Against it, the rule contained in Article 54 paragraphs 2 and 3 of AP I sets a higher threshold for civilian protection. While the United States might make its case in favor of its own more permissive position on the customary standard concerning starvation, it is clear that, without a compromise, a joint NATO operation could unfold the potential for massive confusion. If the NATO alliance addressed this issue and aligned its position beforehand, this could be prevented. I argue in this post that a “maximalist” approach, compelling NATO States to refrain from attacks that may cause civilian starvation, would be the preferable route. After all, adapting to a legal regime that promises higher humanitarian protection would not only offer some legal certainty to a NATO commander in charge. It might also forestall criticism that IHL is bent beyond recognition over the course of a military campaign involving the alliance.
This is a slightly revised version of the original post based on input from one of the cited authors.
***
Rosa-Lena Lauterbach is a PhD Candidate at the University of Cologne and a former Visiting Researcher at the Lieber Institute and Columbia Law School.
Photo credit: U.S. Army, Spc. William Livingston
