Joint Blog Series: Sieges, Evacuations and Urban Warfare: Thoughts from the Transatlantic Workshop on International Law and Armed Conflict

Editor’s Note: This post is part of the joint series of posts hosted by the ICRC Humanitarian Law & Policy Blog, EJIL Talk! and Lawfare, and arising out of the 6th Transatlantic Workshop on International Law and Armed Conflict held at the European University Institute in Florence in July.

Conflict in urban or populated areas poses an enormous danger to civilians and to the civilian infrastructure that sustains the civilian population. The law of armed conflict (LOAC) requires that parties to a conflict take constant care ‘to spare the civilian population, civilians and civilian objects.’ During the conduct of hostilities, the principles of distinction, proportionality and precautions mandate rules for the identification of targets, minimization of incidental harm to civilians, and precautionary measures to avoid or minimize harm to civilians.

How to implement these principles and obligations is, of course, a significant challenge—but the questions of how to mitigate civilian risk and harm while carrying out the military mission go beyond the conduct of hostilities and are equally challenging and important. Our discussions at the Transatlantic Workshop highlighted and probed one particular set of concerns—how to balance different legal norms and operational considerations regarding evacuation and the corresponding prohibition of forced displacement, and sieges and the corresponding prohibition of starvation as a method of warfare.

Is evacuation of civilians an option? In what circumstances?

Imagine a military wants to evacuate civilians from an area—either within its own territory or in another State’s territory where the hostilities are expected or ongoing. The motivation is both humanitarian and based on military considerations: removing civilians from the area of hostilities protects them from harm and also frustrates the enemy’s attempts to use them as human shields or otherwise endanger them for tactical or strategic purposes. At the same time, any evacuation raises serious concerns about displacement, the ability of those civilians to received adequate food, shelter, medical care and other necessities, and the legal parameters for any such evacuation.

LOAC does provide space for evacuation of civilians. And, indeed, when utilized to protect civilians from the hazards of military operations, evacuation can help achieve LOAC’s core purpose of minimizing harm to civilians. At the same time, LOAC prohibits forcible displacement or transfer of the civilian population. During international armed conflict, Article 49 of the Fourth Geneva Convention prohibits forcible transfer or deportation of civilians from occupied territory. During non-international armed conflict, ordering the displacement of the civilian population is forbidden ‘unless the security of the civilians involved or imperative military reasons so demand.’ Legitimate evacuation is thus allowed for the security of the civilian population or for imperative military reasons. The Commentary to the Fourth Geneva Convention explains that security interests can justify displacement of civilians, noting that ‘if . . . an area is in danger as a result of military operations or is liable to be subjected to intense bombing,’ it may or must be evacuated ‘partially or wholly, by placing the inhabitants in places of refuge.’ Similarly, the Commentary to Additional Protocol II emphasizes that it is ‘self-evident that a displacement designed to prevent the population from being exposed to grave danger cannot be expressly prohibited.’ 

A few key questions arise from this basic framework. For example, given that nearly any situation of urban warfare or hostilities in populated areas poses severe danger to the civilian population, is evacuation always allowed in such situations? Could it be required? Where is the line between the legitimate humanitarian action of evacuation and the LOAC violation of forcible displacement? And finally, if the party seeking to evacuate civilians must first wrest control of that area from the enemy—such as if the adversary is forcing civilians to stay within certain areas—how should the party weigh the harm to civilians caused in the course of seeking to evacuate them to an area of safety?

How should evacuation be carried out? By whom and to where? And for how long?

LOAC treaty law provides scant guidance for the implementation of evacuation. It notes primarily that an occupying power undertaking evacuations or transfers ‘shall ensure, to the greatest practicable extent, that proper accommodation is provided to receive the protected persons, that the removals are effected in satisfactory conditions of hygiene, health, safety and nutrition, and that members of the same family are not separated’ (Article 49 of the Fourth Geneva Convention). In essence, LOAC’s fundamental purposes mandate that any evacuation, including those outside of occupation, must ensure appropriate measures for the welfare of the evacuated civilians. But pragmatically, a host of questions remain that must be addressed, or at least posed, in the planning of any such operation.

One set of questions concerns the provision, timing and amount of information about the evacuation that must be provided, as well as information that can be sought about the population being evacuated. When the military is the party undertaking the evacuation, it will have added security concerns regarding the provision of information and will need to find an appropriate balance between providing information that is useful for civilians and affords appropriate assurances and comfort regarding their safety—while still maintaining necessary controls on information for security purposes. Humanitarian organizations will naturally have fewer security imperatives with regard to provision and sharing of information. Both actors, however, have legitimate interests in screening potential evacuees to ensure that only civilians are evacuated and to prevent any fighters, weapons or ammunition from passing through.

Additional considerations include the appropriate parameters for humanitarian organizations and militaries to coordinate with one another in evacuations and the measures that a State or other organization conducting an evacuation must take to ensure the welfare of the civilian population. Finally, the location of place of refuge and the duration and end of evacuation also present questions regarding the continuing obligations to the civilian population. For example, can civilians be evacuated to a location that is safer than the area of hostilities but not entirely safe from the effects of conflict? A military or humanitarian organization must also evaluate how to determine when an area can be deemed safe for civilians to return and how to address potentially conflicting desires to return sooner or remain in the area of evacuation longer. LOAC is also silent on a final question of whether the military or entity needs to secure the civilians’ return home or whether simply notifying them that the area is safe for return is sufficient.

Sieges: If evacuation is not feasible, is isolation permissible if a decision is made not to attack?

Encirclement or siege is permitted under LOAC and has been used for centuries to pressure the defended locality to surrender by maintaining constant pressure on the opposing force. Article 27 of the Hague Regulations of 1907 and Article 17 of the Fourth Geneva Convention recognize the use of siege as a military tactic. The articles mandate that religious and cultural property in besieged areas be spared as far as possible and that parties seek to reach agreements to evacuate certain classes of vulnerable civilians from and find ways to allow free passage of religious and medical personnel and equipment into the besieged areas. Neither treaty law nor customary law prohibit the use of siege, but our discussions at the Transatlantic Workshop raised interesting questions about whether and how sieges lawfully can be used given other essential provisions of LOAC—in particular, those that prohibit starvation as a method of warfare and mandate free passage of relief supplies.

A siege seeks to completely isolate the defending party and all persons within the besieged area—physically, psychologically and electronically. Indeed, it can be an essential tactic to force the enemy to surrender or otherwise submit. At the same time, however, Article 54 of Additional Protocol I prohibits the use of starvation as a method of warfare and further prohibits attacking, destroying, removing or otherwise rendering useless objects indispensable to the survival of the civilian population. Article 14 of Additional Protocol II similarly prohibits starvation as a method of warfare in non-international armed conflict. According to ICRC Customary IHL Rule 53, ‘the prohibition of starvation as a method of warfare does not prohibit siege warfare as long as the purpose is to achieve a military objective and not to starve a civilian population.’ The Oxford Guidance on the Law Relating to Humanitarian Relief Operations in Situations of Armed Conflict, the Manual on International Law Relating to Air and Missile Warfare, and the San Remo Manual on International Law Applicable to Armed Conflicts at Sea all advise that a siege or blockade ‘whose sole or primary purpose is to starve the civilian population or to deny it other objects essential for its survival’ (Oxford Guidance p37 n 80) is prohibited. In addition, if the civilian population is inadequately provided with food and other essential supplies, the besieging or blockading party must allow the free passage of such supplies in accordance with technical arrangements—including search—that it may establish.

How should parties to a conflict—and the interpretation and application of LOAC—balance these rules and values—the lawfulness of sieges on the one hand and the prohibition on starvation as a method of warfare on the other? One could interpret the prohibition on starvation as a method of warfare as creating a presumption that sieges are prima facie unlawful. LOAC does not mandate that result—as the treaty provisions above confirm. However, differentiating between sieges undertaken for lawful military purposes and those imposed with the intent to starve the civilian population can be extraordinarily difficult. At the same time, in some circumstances, a siege will offer greater opportunity to spare civilian life in comparison to an all-out assault on an urban area, an operation that is certain to produce massive civilian suffering and destruction of the fabric of civilian society. In the end, parties to conflict and external observers must seek a balance between the humanitarian imperative of preventing starvation and the effectiveness of siege as a military tactic. 

A few final thoughts

In addition to the specific legal and operational challenges of evacuations and sieges, two remaining issues of interest are the role of the defending party and the role of technology. 

Beyond its obligations to take feasible precautions under Article 58 of Additional Protocol I, the defending party can take further steps to lessen civilian suffering, such as releasing civilians, for example. Is forcing civilians to remain in an area where hostilities are ongoing or imminent a violation of LOAC? What about diverting relief supplies intended for civilians? If the defending party agrees to release civilians, does it have an obligation to screen persons leaving to ensure that only civilians are leaving? And if the defending party refuses to release civilians, does it bear responsibility for civilian suffering and deaths resulting from a siege or encirclement?

Modern technology also introduces new capabilities and questions. Technology can offer States or other parties to conflict more diverse and more effective tools for warning the civilian population; capabilities for tracking relief shipments to identify and prevent diversion to enemy forces; and options for screening and identifying enemy fighters or military equipment. Some future questions will therefore include whether a party has obligations to use such technology to mitigate civilian harm or to expand possible courses of action so as to maximize protection of civilians and minimize civilian suffering.