IRRC No. 919

“Safe zones”: A protective alternative to flight or a tool of refugee containment? Clarifying the international legal framework governing access to refugee protection against the backdrop of “safe zones” in conflict-affected contexts

Reading time 38 min read
Download PDF

So-called “safe zones” pose an increasingly pressing threat to genuine and robust international legal protection for persons fleeing conflict. This paper aims to address the key challenges and risks of safe zones under international law and to provide some clarifications on the legal framework which must be respected by refugee-receiving States. Through assessing the intentions of preventing migration flows which underlie their creation, this paper will demonstrate that the existence of safe zones cannot be used to circumvent the obligations of refugee-receiving States under international law, specifically the right to leave and seek asylum and the prohibition of non-refoulement. This paper concludes that safe zones should only be created as an urgent response to humanitarian crises in order to ensure the immediate safety of civilians in conflict zones, and only under very strict conditions. In this respect, this paper will demonstrate that even if safe zones comply with certain minimum protective standards, because of the volatility and complexities of the conflict environment they should not and cannot act as a substitute for genuine refugee protection under international law. * The author was the recipient of the 2021 Henry Dunant Prize for a previous version of this paper that was submitted as part of the LLM in International Humanitarian Law and Human Rights at the Geneva Academy of International Humanitarian Law and Human Rights.


Safe zones are not a novel feature of the conflict landscape and numerous terms, including “safe zone”, “safe area”, “safe haven”, “demilitarized zone” and “protected zone”, have been coined in multiple prior conflict-affected contexts.1 These broadly refer to specifically designated areas that aim to afford a form of heightened physical and humanitarian protection to the displaced civilian population in an ongoing armed conflict.2 In theory, safe zones have the potential to provide additional protection from attack, facilitate humanitarian and medical assistance and even enable education, employment and other opportunities in the midst of an armed conflict.3 However, as will be shown in this paper, safe zones are interconnected with responses to conflict-induced displacement and risk evolving into a tool utilized by refugee-receiving States to either knowingly or unknowingly avoid complying with their obligations under international law.

Most prior examples of safe zones, while outwardly aiming to protect the civilian population, have been closely associated with underlying refugee-containment strategies as their proclaimed safety is relied upon to both prevent the flight of individuals and promote their return.4 As the increasingly protracted nature of armed conflict triggers further mass displacement and large-scale cross-border movements, there is a growing risk that refugee-receiving States will more frequently consider safe zones as a means of containing persons in need of protection within their country of origin.5 In turn, the present author posits that safe zones have the potential to become an increasingly common feature of contemporary conflicts as a means of controlling migration flows, requiring clarification of the legal framework governing their existence. While they raise a number of jus ad bellum and jus in bello issues that will be alluded to, due to the more limited attention in existing literature this paper will focus on the risks that safe zones pose to comprehensive and effective refugee protection.6

Through an assessment of the state of international law on this matter deriving from the complementary application of international refugee law (IRL), international humanitarian law (IHL) and international human rights law (IHRL), this paper will emphasize that the illusion of safety that undercuts most safe zones entails that they cannot be treated as a permissible alternative to robust refugee protection under international law. It will begin by outlining the core typologies of safe zones that have emerged in international practice and briefly discuss their legal basis and impact on the civilian population. Following this initial overview, in light of the potential association of containment strategies with safe zones, this paper will undertake a comprehensive analysis of the refugee protection issues which could arise from their establishment. This will address the right to leave and seek asylum, the Internal Protection Alternative (IPA) as an impermissible basis for the rejection of asylum claims, the importance of the principle of non-refoulement as well as broader matters of return in order to demonstrate the incompatibility of safe zones with the obligations of refugee-receiving States.

The paradox of safety: Ensuring the robust protection of persons in safe zones

This section will compare the legal basis and core features of two broad forms of safe zones, namely “conventional” and “imposed” safe zones.7 Accordingly, it will reflect on the limitations of these zones in practice and the consequent risk that future safe zones will evolve beyond these two distinct typologies, with a particular emphasis on the threat that this would pose to robust refugee protection.

“Conventional” safe zones

The only explicit legal basis for the creation of safe zones can be found in IHL, which provides for the possibility of establishing numerous forms of so-called “protected zones”.8 The premise of these “conventional” safe zones is to provide enhanced protection from the effects of hostilities for the civilian population, as well as the wounded and sick, as under the relevant IHL provisions these spaces should not contain any military objectives and therefore cannot be deliberately attacked.9 The most protective and comprehensively defined forms of safe zones in the IHL framework are demilitarized zones, which can be broadly understood as delineated areas in which belligerents agree not to conduct any hostile activities or military operations under specified conditions.10 Although the Geneva Conventions and their Additional Protocols only explicitly foresee their establishment in international armed conflicts (IACs), the International Committee of the Red Cross (ICRC) has recognized that the prohibition of directing an attack against safe zones is a customary rule equally applicable in non-international armed conflicts (NIACs),11 in which zones could be established by special agreements under Article 3 common to the four Geneva Conventions.12

The crucial protective benefit of “conventional” safe zones is that they require the express consent of all parties to the conflict, which increases the likelihood that their neutral and exclusively demilitarized character will be respected.13 However, this is also their biggest challenge, as obtaining the consent of belligerents has proven extremely difficult in practice, and is also likely to be withdrawn in the changing conflict environment, creating further risks for the civilian population.14 Notably, safe zones are typically established in response to repeated attacks against the civilian population and it has been argued that parties who do not wish to respect IHL rules on the protection of civilians are unlikely to consent to a safe zone that is premised on their enhanced protection.15 Thus, despite the conditions enshrined in IHL for the establishment of safe zones, which have the potential to provide a degree of safety, this framework is rarely applicable.16

“Imposed” safe zones

The next typology of “imposed” safe zones refers to those established under the auspices of the United Nations (UN). Particularly evident in the 1990s, the UN Security Council has previously authorized the creation of safe zones in order to maintain or restore “international peace and security” under Chapter VII of the UN Charter, the impacts of which have been analysed extensively in existing literature.17 For the purposes of this paper, it is sufficient to note that this form of safe zone has very rarely retained its promises of safety. Although UN Security Council authorization means that its establishment is not in violation of the prohibition on the use of force,18 it compromises its civilian character as it has been argued that parties are unlikely to refrain from hostilities in a safe zone that is imposed non-consensually by foreign military powers. In turn, all prior safe zones established with UN involvement are considered as having failed to ensure the long-term and comprehensive protection of the civilian population as they suffered from continued, and sometimes heightened, large-scale attacks against civilians.19

The crucial difference between UN-sanctioned safe zones and the jus in bello regime is that these safe zones do not require the consent of belligerents, so can overcome the challenges addressed in the previous section.20 However, they still require the consent or non-veto of the Security Council's permanent five members.21 By consequence, the UN Security Council has not lent its authorization to the creation of safe zones since the 1990s and it seems unlikely to in the near future.22 The possible reasons for this are twofold. First, there is presumably a natural reluctance to authorize the creation of safe zones when, as aforementioned, they have ultimately failed in ensuring the long-term protection of the civilian population in almost all prior instances.23 A further challenge is that the interests and divergences of the permanent five members can sometimes lead to the exercise of the veto to block action in response to humanitarian crises where properly established safe zones could arguably have some temporary benefit.24

While UN-imposed safe zones are by no means a preferred response, inactivity at the UN Security Council does not necessarily mean that “imposed” safe zones will no longer be established in conflict-affected contexts. Rather, this paper anticipates that practice could shift towards the unilateral establishment of safe zones by one or more States in foreign territory without consent nor UN Security Council authorization, as will now be discussed.

Evolving practices and future risks

Given the stringent conditions for the specific IHL provisions on safe zones to be applicable and the unlikelihood that the UN Security Council will consent to their establishment in the near future, there is a prominent risk that safe zones could increasingly be unilaterally established by a foreign State both in violation of the UN Charter and in a manner that is not specifically foreseen by IHL. In particular, it seems somewhat unlikely that a State would expressly consent to foreign intervention to establish a safe zone on its territory in any ongoing or future conflict scenarios, especially if it would prevent the flight and facilitate the return of that State's own nationals to situations of insecurity and possibly result in a loss of complete control over its territory. Based on prior practice, it is equally unlikely that a State engaged in an armed conflict would, on its own initiative, establish a safe zone for the protection of its nationals on its own territory, or that it would be respected by enemy belligerents if it did so.

Furthermore, even if the establishment of a safe zone was consented to in some manner by the territorial State, as aforementioned, it is still unlikely that there would be clear and express agreement from all belligerents specifically conferring it protected status as required under the “conventional” framework. In that instance, the safe zone would not benefit from the enhanced protection foreseen by IHL and there would be no explicit obligation on the parties to the conflict, including those administering the safe zone, to respect any commitment towards complete demilitarization and refrain from attack. Alternatively, if consent were not at all forthcoming, then there would be a risk of the intensification of hostilities between the intervening and non-consenting forces, either within or in the vicinity of the safe zone, further compromising any possibility of durable protection. It is also suggested that without UN support, the capacity of the intervening forces and their willingness to ensure the genuine protection of the safe zone's inhabitants would be limited. Therefore, in all of these scenarios, the evident challenges faced in ensuring the long-term safety of the civilian population in prior safe zones would only worsen if practice evolved in this direction.

In order to provide the most comprehensive overview, three alternative scenarios – a safe zone established by the home State, a safe zone expressly consented to by the home State and an imposed safe zone not consented to by the home State – will be considered in the subsequent discussion which aims to demonstrate the risks for refugee protection of creating safe zones which have, at best, a fragile basis in international law.

Safe zones and refugee protection

This paper will take a holistic approach to refugee protection in armed conflict, and views IRL, IHL and IHRL as sources of mutually reinforcing and complementary, rather than conflicting, protection.25 IRL will be considered due to the focus on refugee protection. Additionally, both IHL and IHRL are crucial reinforcing sources of protection that are applicable in the context of armed conflict in which safe zones would typically be established.26 With this in mind, this section will consider the varying safe zone typologies that have been presented in order to challenge three prominent arguments that risk being associated with their existence: first, that they can justify the denial of access to asylum; second, that their posited “safety” can ground the rejection of asylum applications; third, that persons can be returned to safe zones.

Safe zones and access to asylum

The first danger is that safe zones are relied on by States to justify the closure of borders and denial of access to asylum, forcing persons to settle in spaces which are misrepresented as safe. Although safe zones are an arguably protective alternative to the dangers of irregular border crossings,27 relying on them to prevent persons from seeking protection would be clearly incompatible with international law, particularly the right to leave and to seek asylum and the principle of non-refoulement.

The right to leave

The right to leave in the Refugee Convention obliges States Parties to permit “refugees lawfully staying in their territory”, namely the State of asylum, to travel outside the State.28 This has relatively limited relevance for the present discussion as safe zones are premised on being established in the State of origin rather than the State of asylum, and this provision does not offer any protection to those who remain in their home State. Therefore, the right to leave under IHL and IHRL must be considered as they provide more comprehensive protection to persons wishing to leave a territory where a safe zone is located.

The right to leave under IHL stipulates an entitlement to voluntarily leave a territory at the start of, and during, an IAC.29 However, this is subject to several caveats. First, the right to leave does not cover all instances in which safe zones could be established as it is predominantly foreseen for individuals on a State's own territory.30 Second, even on own territory, the right to leave will only benefit protected persons, as defined by Article 4 of GC IV, which broadly requires that an individual is in the hands of a party of which they are not nationals. Therefore, individuals in the power of their State of nationality, such as those residing in a safe zone controlled by the territorial State or attempting to cross that State's border to seek asylum elsewhere, would not benefit from the right to leave. While Article 73 of AP I recognizes that refugees can be protected persons regardless of their nationality, this is equally limited in its protective effect because they must have been recognized as a refugee prior to the outbreak of hostilities and the parties to the conflict must have ratified AP I.31 Additionally, individuals in the power of their country of nationality, and therefore without protected status, would not be recognized as refugees by their own State so would not benefit from this additional protection. Third, individuals can be prevented from leaving if their departure would be contrary to “national interests”, which has been interpreted as broader than the State's security interests and can also encompass economic considerations.32

Therefore, in light of the caveats above, there are very limited instances in which individuals in need of protection could benefit from the right to leave their country of origin and seek asylum elsewhere under IHL alone. Specifically, the right to leave under IHL would only substantively benefit persons in a safe zone that is located outside their country of nationality who wish to return to their home State or another country. However, this is a relatively unforeseeable scenario as a State that wishes to prevent the entry of foreign nationals is unlikely to establish a safe zone on its own territory. Even if this were the case, individuals present in the safe zone would be seeking protection in that receiving State rather than wishing to return to their State of persecution and would therefore not substantively benefit from the right to leave in this scenario.

As a result of these limitations, recourse must be had to the right to leave under IHRL, which applies to all individuals regardless of their nationality and is considered as applicable in armed conflict.33 In this respect, as well as the State's obligation to secure the rights of all persons on its territory including those arriving at its border, the extraterritorial application of IHRL has been recognized where a State has established effective control over territory or its agents exercise physical control and authority over persons.34 This entails that regardless of the conditions of its establishment, those administering a safe zone would be responsible for respecting the rights of persons therein, including the right to leave. In particular, if a safe zone is established or administered by the territorial State, then it is uncontroversial that it must secure the right to leave of all persons present in the safe zone as they would be clearly within its jurisdiction. Additionally, if a safe zone is directly administered by another State's forces, it is considered that those administering the safe zone could have extra-territorial IHRL obligations based either on the degree of control and authority they would exercise over its inhabitants or their effective control over the territory.35

Despite its broad applicability, the right to leave is not absolute. In particular, in times of “public emergency which threatens the life of the nation”, under certain treaties States can derogate from the right to leave, provided measures are strictly required by the situation, not discriminatory, nor inconsistent with other international obligations.36 This has been restrictively interpreted as requiring more than the existence of an armed conflict.37 Specifically, when a State is involved in an extraterritorial armed conflict in which hostilities take place outside its territory, the extent to which there is a legitimate “threat to the life of a nation” permitting derogation is debated.38 Under a similar approach, it has been argued that if a refugee-receiving State imposes a safe zone in another State where there is an ongoing conflict to which they are not party, then it could not derogate from the right to leave because of the lack of a “threat” on its own territory.39 This reasoning could also be extended to a situation where the imposing State is party to the conflict and is engaged in hostilities both within and beyond the safe zone provided hostilities remain confined to the foreign territory.

Furthermore, Chetail has importantly observed that the ability to derogate is limited to certain instruments as the right to leave is enshrined in a number of relevant universal and regional IHRL treaties without any possibility of derogation.40 This scholar has also noted that even if permitted by a certain instrument, States will not always derogate in situations of emergency, especially if they do not want to recognize the existence of an armed conflict on their territory.41 Under this reasoning, it can be argued that there are a number of scenarios where the right to leave could be fully applicable without derogation and would therefore continue to provide crucial protection to those wishing to flee a State where a safe zone is located.

The right to leave can also be restricted on the basis of the legitimate aims of national security and public order, among others. However, this is only as long as measures are provided by law and necessary, which entails that they must be the “least intrusive” measure and “proportionate to the interest to be protected”.42 These conditions therefore significantly limit a State's ability to prevent flight, and were notably drafted as the “exception” rather than the “rule”.43 In particular, a State could not ground a limitation of the right to leave on the mere existence of a safe zone as any restriction must be based on the individual circumstances of the case rather than the general conditions in the country of origin.44 In this respect, the individual seeking protection would probably be fleeing persecution or serious harm in an ongoing conflict, and only in the exceptional instance of serious criminality could they be considered as posing any kind of “threat” to the receiving State.45 Therefore, in virtually all instances, the exposure to continued danger through restricting an individual's right to leave is neither necessary nor proportionate to any legitimate aim.

The right to seek asylum and non-refoulement

The right to leave is reinforced by the right to “seek and enjoy asylum” as enshrined in the Universal Declaration of Human Rights,46 which has been argued as implicit in the Refugee Convention and an emerging customary norm.47 However, on its own, this is limited in effect as it does not oblige States to actually grant asylum and is therefore only made operable through the principle of non-refoulement.48

Non-refoulement is a crucially important norm enshrined in IRL, IHRL and IHL. Under IRL, this principle applies to both asylum seekers and formally recognized refugees and prohibits return “in any manner whatsoever” to a State where they face persecution on five limitative grounds – nationality, political opinion, race, religion and membership of a particular social group.49 Similarly, under IHL, parties to an IAC are prohibited from transferring protected persons from their own territory to another State where they fear persecution, but on more limited grounds of political opinion or religious belief.50

Under IHRL, the principle of non-refoulement proscribes return where there are “substantial grounds” to consider that an individual faces a “real risk of irreparable harm” owing to a serious human rights violation.51 This traditionally encompasses harm amounting to torture and cruel, inhuman or degrading treatment but has also been interpreted to include other core rights, including the right to life and right to a fair trial, namely in situations of armed conflict.52 Moreover, as argued by Chetail, it is not necessarily limited to specific rights and will be engaged when there is a serious violation amounting to degrading treatment.53 Therefore, this prohibition is broader than under IRL and IHL, especially in its application to all individuals including those who are not protected persons under IHL, who fall under Article 33(2) of the Refugee Convention or who do not otherwise meet the refugee definition under Article 1A(2) of the Refugee Convention.54

The application of non-refoulement to safe zones will be considered in the section on returns below. At this stage, it is important to emphasize that non-refoulement is at the core of international refugee protection and is an essential guarantee when challenging possible responses by refugee-receiving States towards individuals fleeing a State where a safe zone is located. Taken together with the right to leave and to seek asylum, this provides a crucial layer of protection that precludes States from preventing admission to their territory for individuals in need of protection.

Push backs and border closures

Past practice has indicated that safe zones risk being associated with border closures and other push-back practices, as refugee-receiving States could argue that they mitigate the need for individuals to seek refuge elsewhere due to their apparent safety.55 This would be clearly incompatible with the international legal framework protecting the right to leave, as well as the principle of non-refoulement which is settled as applying to rejection at the frontier.56 The expansion of this principle to include interception on the high seas is also crucial in protecting individuals who make the crossing across international waters to seek protection in Europe.57

Beyond these instances, a pertinent consideration is whether non-refoulement protects persons who remain in their home State. This would be particularly relevant if push-back practices are employed at borders against those fleeing States where safe zones are located.58 In this scenario, the State of asylum would be constructively refouling persons to a place of persecution or serious harm by indirectly forcing them to seek protection in a safe zone rather than travel to or across the border where they know their asylum claim would either not be considered or be arbitrarily rejected.59 In this respect, it has been argued that non-refoulement under IRL is exclusively territorial and will not protect those who remain in their country of origin.60 However, as affirmed by the UN High Commissioner for Refugees (UNHCR) and supported in scholarship, there is “growing consensus” that Article 33(1) of the Refugee Convention applies to all persons who fall under the jurisdiction of a State, even if on another State's territory.61 The core argument supporting this conclusion is that given their similar object and purpose, there should not be a discrepancy between the geographical scope of application of non-refoulement under IRL and IHRL.62 While this is still somewhat debatable, it is clear that based on the extraterritorial application of IHRL outlined in the context of the right to leave above, the principle of non-refoulement under IHRL must be respected when the individual falls under the jurisdiction of the State, and it is argued that this would include a safe zone controlled and administered by foreign or peacekeeping forces.63

Given that a central concern of refugee-receiving States is the scale of individuals who arrive at their borders in need of protection, it is also possible that safe zones would be associated with arguments of mass influx. While there are remaining controversies, Chetail has convincingly argued that based on the “inclusive” wording of the Refugee Convention, mass influx cannot be a permissible exception to non-refoulement under IRL.64 Taken together with the absolute nature of non-refoulement under IHRL, persons cannot be refouled to a place where they face persecution or serious irreparable harm regardless of the alleged burden on the State of asylum.65 Additionally, the prohibition of collective expulsion requires that all individuals benefit from an individual assessment in their asylum claim, including in situations of mass influx.66 Therefore, any collective decision to refuse entry to a State of all persons fleeing a conflict-affected context in which there is a safe zone would be prohibited.

In sum, under the reinforcing protections of IHL, IRL and IHRL, particularly the right to leave, the principle of non-refoulement and the prohibition of collective expulsion, regardless of the existence or apparent safety of a safe zone, individuals must still be able to leave their country of origin and access asylum procedures in another State.

Safe zones and determination of asylum claims

The next barrier for refugee protection posed by safe zones is that their existence in the applicant's country of origin could justify the refusal of an asylum claim on the grounds of the IPA. This section will address this notion in the alternative instances that a safe zone is administered by the military powers of the home State, a foreign State or a UN peacekeeping operation.

The internal protection alternative

The IPA is subject to controversy as it was not initially envisaged by the system of refugee protection and is not mentioned in the Refugee Convention.67 It is settled that the IPA cannot be invoked by African Union Member States who have ratified the Organisation of African Unity (OAU) Convention, as its definition of a refugee explicitly includes persons compelled to leave their country of origin owing to a number of specified scenarios taking place in part of the country.68 However, beyond this region, the matter is not so clear cut and States frequently rely on the IPA as an implicit part of the assessment of a well-founded fear of persecution and whether the applicant is “able or willing” to avail themselves of protection under Article 1A(2) of the Refugee Convention.69

The UNHCR has acknowledged the possibility of the IPA but has specified a two-fold test of “safety” and “reasonableness” to limit its scope.70 This entails that in the applicant's particular circumstances, there is an area of the country of origin where they do not have a well-founded fear of persecution or can receive protection from it, and which they can safely access and “lead a relatively normal life without … undue hardship”.71 These conditions are explicitly endorsed by the European Union Qualification Directive 2011/95/EU (EUQD) which enshrines a clearer legal basis for the IPA and crucially specifies the minimum standards required for there to be adequate protection from persecution.72

Although the EUQD is not applicable to persons seeking asylum outside of the European Union (EU), its provisions can shed light on the evolving content and substance of the IPA, including beyond the EU system, and provide more specificity on a relatively vague notion. While this is yet to occur, it is also not unforeseeable that in the future EU Member States could rely on the IPA to deny asylum on the basis of the existence of a safe zone in the individual's country of origin that promoted some form of safety.

Under normal circumstances, the IPA is reserved for cases of persecution by non-State actors, as a State is presumed to be able to exercise its power everywhere on its territory.73 Therefore, if the territorial State consents to the establishment of a safe zone, the scope of the IPA will be considerably limited as the safe zone would be administered either directly by or with the acquiescence of the State, which would be able to continue to persecute its inhabitants. However, it is considered that safe zones will not always be explicitly consented to and, instead, could be controlled and administered by external actors, specifically a UN peacekeeping operation or a foreign power. In this respect, it has been recognized that a State's ability to persecute can be refuted in the exceptional circumstance that it does not have control over the whole territory.74 This creates a risk of the expansion of the IPA to reject asylum claims based on false narratives of safety as refugee-receiving States could argue that an individual is able to receive protection from persecution at the hands of both State and non-State actors if there is a safe zone administered by an external actor in their country of origin. However, this argument can be dismantled, as mere safety in a safe zone, let alone comprehensive protection from persecutory harm, is considered as mostly unrealistic.

Actors of protection: UN peacekeeping forces

This paper contends that future safe zones could increasingly be non-consensually established by foreign States without UN Security Council authorization. However, as has been endorsed by some authors,75 a UN peacekeeping force could still be deployed to an existing safe zone to ensure its safety, thus requiring consideration of how this would affect an IPA assessment. This is even more pertinent in light of the role of peacekeeping forces in protection of civilian (PoC) sites, a form of safe zone coined following the spontaneous large-scale arrival of civilians at a UN peacekeeping base in South Sudan.76 PoC sites differ from the more common understanding of safe zones presented at the outset of this paper due to their unplanned character and raise several issues, related to jurisdiction and supervision, beyond the scope of this paper.77

The EUQD recognizes at Article 7(1)(b) that “parties or organisations, including international organisations, controlling the State or a substantial part of the territory of the State”, such as UN peacekeeping operations, are potential actors of protection against persecution. Article 7(2) further provides that protection must be “effective” and “non-temporary”, and the actor must take steps to prevent persecution, such as through operating an “effective legal system” for the punishment of persecutory acts. The UNHCR has similarly accepted the possibility of non-State actors of protection in exceptional circumstances where they have a high degree of control and the capacity to provide comprehensive protection.78 Thus, both within and outside of the EU system, there is a relatively high threshold which is beyond the typically limited mandate of UN peacekeepers to protect against imminent physical harm.

Although it is possible that they could provide physical protection from hostilities in a safe zone, the shortcomings of PoC sites suggest that UN peacekeeping operations would not have the capacity to provide sufficient protection for the IPA.79 Indeed, in South Sudan, despite the prominent protection needs, only basic medical assistance and water were provided and there was no functioning judicial or administrative system.80 This was notably due to the limitations in their mandate, which did not grant powers of law enforcement or judicial authority, preventing the ability to detain or try individuals.81 In this respect, in order for a peacekeeping operation to be a competent actor of protection in a safe zone, it must have a mandate that allows for the detention and prosecution of persons in a manner that respects fair trial guarantees. However, this seems relatively unlikely given the already evident resistance of the UN Security Council to lend support to safe zones, as discussed above.

Actors of protection: Foreign military powers

In the possible scenario that a safe zone is established by or with the support of a foreign power and administered by its forces, it is similarly considered that the administering powers would not have the capacity to be considered as a competent actor of protection for the purposes of the IPA. Although these forces could probably provide a form of physical protection to individuals through military enforcement activities, their capacity to provide comprehensive and effective protection from persecution in the manner required by the IPA is more limited. The principal issue is that the control and administration of a safe zone by foreign military actors would likely take place in the context of an armed conflict with a constant risk of the resurgence of active hostilities and limited safety. This cannot be considered as an environment where individuals could reasonably be expected to settle in a long-term or durable manner, the final component of the IPA that will now be considered.

“Reasonably expected to settle”

In addition to the challenges already presented, even if the applicant could receive immediate protection from persecution in a safe zone, they must also be able to safely and legally travel there and be “reasonably expected to settle”.82 These are the strongest arguments in favour of this section's conclusion that, in almost all instances, safe zones cannot be invoked as an IPA.

With regard to the condition of safe and legal travel, the traditional premise of safe zones under the IHL framework foresaw their establishment in the midst of intense fighting to protect civilians unable to flee hostilities, rather than as a space for persons in distant areas to travel to. Expecting persons to travel through an active war zone is clearly problematic and precludes the possibility to invoke the IPA for those not in proximity to the safe zone.83 Moreover, the fundamental condition of settlement entails that a safe zone must guarantee more than protection from hostilities and be a “habitable” and “safe” environment, in which the individual can comprehensively and freely enjoy civil and political as well as economic, social and cultural rights.84 This must also take into account the risk of indirect refoulement, namely the return of persons to a safe zone where the socio-economic conditions are insufficient for them to remain indefinitely, driving return to the original place of persecution or other areas where they face harm.85

Beyond the specific context of the IPA, a number of standards have been suggested by Gilbert and Rüsch in order for a safe zone to provide genuine and robust protection to its inhabitants, which can also be relied on to establish some possible benchmarks for when a safe zone could be sufficiently safe for the IPA. In particular, at the absolute minimum, there must be respect for the right to life, freedom from torture and cruel, inhuman or degrading treatment, freedom from sexual and gender-based violence, and unimpeded humanitarian access. In addition, those administering the safe zone must secure an adequate standard of living for all inhabitants, through providing, at least, medical care, food, water and shelter.86 If these conditions were met, then it could be argued that the IPA would be engaged. However, it is relatively unrealistic given that the vast majority of safe zones would be established in a conflict-affected context where the presence of unwanted forces and ongoing attacks either in the safe zone itself or in surrounding areas would foreseeably result in direct and indirect threats to the livelihood and wellbeing of civilians.87

Nevertheless, even if a safe zone could meet the conditions of the IPA, there is still the risk that refugee-receiving States would misinterpret it as safe for every individual and reject all cases without any consideration of their merit.88 While it might provide some immediate protection, the existence of a safe zone could not be considered as eliminating all forms of persecution, particularly those which are not conflict related. Furthermore, as enshrined in the EUQD, affirmed by the UNHCR and reinforced by the prohibition of collective expulsion, there cannot be automaticity in an asylum determination assessment.89 Therefore, any rejection based on the IPA must be following an individualized assessment that takes into account the applicant's personal circumstances, beyond the mere existence of the safe zone.

Safe zones and returns

Both IHL and IHRL recognize the right of return.90 However, as highlighted throughout this paper, safe zones risk being associated with illusory notions of safety and being employed to justify the return of persons to their country of origin where a safe zone is located.91 As will be demonstrated in this section, genuine and comprehensive safety is often unachievable and mostly unrealistic in the context of armed conflict, precluding the return of persons to safe zones in almost all instances.

The application of non-refoulement to safe zones

The principle of non-refoulement under IRL and IHL would prohibit return to a safe zone where an individual faces an immediate and apparent danger of persecution in varying instances. Firstly, under IHL, in certain instances, individuals cannot be forcibly returned by a foreign power to a place where they fear persecution on the grounds of their political or religious beliefs.92 While this could include the return of individuals present in another State to a safe zone in their country of nationality or origin, this must be caveated as it would only apply to returns from States who are party to the IAC and therefore bound by these provisions, and would only benefit protected persons. Therefore, despite their important protective benefits, the relevant IHL provisions would not cover all scenarios, also due to the more limited grounds of persecution and the arguable exception of deportation.93

Nevertheless, persons falling outside the scope of protection of IHL would likely be protected by non-refoulement under IRL, which covers more instances of persecution and prohibits all forms of return. Notably, in light of the conflict environment in which the safe zone would probably be established, it is more than foreseeable that an individual would face persecution on their return. In particular, if the safe zone were administered by or with the acquiescence of the State, then persecution faced by an individual at the hands of the government prior to their departure would not be alleviated by the existence of a safe zone. It is also considered that persecution could arise in the aftermath of conflict and be triggered by return to the safe zone, such as against those who were perceived as supporting the enemy forces of those administering the safe zone. Additionally, as introduced in the context of the IPA, the administering forces are also likely to be ill equipped to provide adequate protection against persecution from other actors.

Notwithstanding the likelihood of persecution in a safe zone, given the challenges of achieving genuine and long-term safety identified in this paper, it is difficult to conceive of a situation where a safe zone would be free from the threat of harm to justify return in accordance with the broader definition of non-refoulement under IHRL. More generally, as documented by the ICRC, persons who are displaced by conflict, in particular women and minority groups, often do not wish to return to their country of origin out of fear for their safety and security stemming from the continued risk of instability, even if the immediate threat of attack is no longer present.94 This is even more prominent in the case of a safe zone that could, at best, only temporarily provide safety to the civilian population from attack. In particular, it has been noted that if belligerents do not agree to the safe zone, a credible military presence would probably be needed in order to deter activities which compromise its safety.95 In turn, it would probably be immediately established in a space prone to hostilities between those administering the zone and non-consenting belligerents, giving rise to the risk of serious harm against the civilian population, and engaging the IHRL principle of non-refoulement to preclude return to safe zones in almost all other instances.

Return as voluntary, safe, dignified and durable

Although not explicit in the Refugee Convention, a correlative of non-refoulement is that return must always be voluntary, meaning that, regardless of the conditions of safety, individuals could in no instances be forcibly or coercively compelled to return to a safe zone.96 Additionally, in recent years, emphasis has increasingly been placed on the “objective conditions” in the country of origin grounded in the language of safety and dignity on return.97 Consequently, this would require more than just freedom from hostilities in the safe zone but comprehensive “physical, legal and material safety”, in which returning individuals could fully enjoy their rights, including economic, social and cultural rights, and access services without discrimination.98 Authors have also increasingly argued that return must be durable, a somewhat undefined term associated with indirect refoulement, which requires that return is sustainable and does not lead to further displacement.99

Building upon this paper's prior analysis in the context of the IPA, active steps must therefore be taken by those who are administering a safe zone to secure the livelihood of persons and enable the exercise of their rights in a genuine and non-discriminatory manner, including the provision of education, comprehensive medical care and employment opportunities.100 There should also be identifiable longevity in safety, which could be demonstrated by the comprehensive (re)construction of infrastructure, including hospitals, schools, homes and sites of worship that were likely to have been destroyed by conflict.101 In addition, when coupled with the principle of non-refoulement, it would be essential that the safe zone had an effective police force, and ideally a functioning judicial system, that would effectively protect all civilians against the risk of persecution and serious harm in a manner that fully upholds IHRL guarantees, including freedom from arbitrary arrest and detention.102 If a future safe zone were to uphold these comprehensive guarantees, then it is possible that such a space could be considered as potentially suitable for return. However, it is argued that this would only be attainable if the safe zone were located in an area where there is no risk of the resumption of hostilities. As it stands, if safe zones continue to be established in situations of ongoing armed conflict, then the volatility of the conflict environment entails that they can never be considered as durably safe to ensure the long-term protection of the civilian population and justify their return.

Safe zones as an immediate, ad hoc and short-term humanitarian response

By taking into account emerging practices related to their creation, this paper has provided some clarity on the legal framework governing safe zones and demonstrated their illegality as alternatives to refugee protection under international law. In this conclusion, the present author wishes to make some final qualifications. Primarily, it is not necessarily the concept of safe zones in abstracto that is problematic but their potential evolution into a tool used to conceal a State's anti-migration interests. Indeed, given the increasingly protracted nature of armed conflict and the serious risks this entails for the civilian population, the potential of a safe zone to ensure the provision of humanitarian assistance and enhanced physical protection is not something to be overlooked.

While it may be somewhat idealistic in light of the complexities of safe zones that have been discussed throughout this paper, properly constituted and respected safe zones, overseen by an impartial humanitarian agency such as the UNHCR or ICRC, can provide vital protection for those trapped in conflict and do have the potential to lessen excessive loss of civilian life.103 They can also be particularly beneficial for the protection of internally displaced persons who, for either voluntary or coercive reasons, do not leave their home country.104 However, the establishment of safe zones must form part of a genuinely humanitarian strategy focused on the immediate protection of civilians to complement, rather than substitute, robust refugee protection. Crucially, any safe zone must be accompanied by complete respect of the international legal framework governing refugee protection that has been elucidated throughout this paper. In this respect, to have any chance of success, the essential shift that must occur in the discourse and practice surrounding safe zones is their disentanglement from the problematic refugee-containment strategies of States and the migration context altogether.

  • 1For an overview of prior examples, see, e.g., Mélanie Jacques, Armed Conflict and Displacement: The Protection of Refugees and Displaced Persons under International Humanitarian Law, Cambridge University Press, Cambridge, 202, pp. 232–44.
  • 2This follows the common definition in scholarship. This paper will refer to such spaces as “safe zones” throughout. See, e.g., Wilson Chun Hei Chau, “Creating Refuge in Hell: The Coming of Age of Safe Areas for the Protection of Civilians in Armed Conflict”, Auckland University Law Review, Vol. 18, 01, p. 19; Phil Orchard, “Revisiting Humanitarian Safe Areas for Civilian Protection”, Global Governance, Vol. 0, No. 1, 014, p. 55; Geoff Gilbert and Anna Magdalena Rüsch, Creating Safe Zones and Safe Corridors in Conflict Situations: Providing Protection at Home or Preventing the Search for Asylum?, Policy Brief 5, The Andrew & Renata Kaldor Centre for International Refugee Law, June 017, p. 3, available at:… (all internet references were accessed in January 0).
  • 3G. Gilbert and A. M. Rüsch, ibid., pp. 1–.
  • 4Cécile Dubernet, The International Containment of Displaced Persons: Humanitarian Spaces without Exit, Ashgate Publishing Limited, Aldershot, 2001, p. 122; Hikaru Yamashita, Humanitarian Space and International Politics: The Creation of Safe Areas, Ashgate Publishing Limited, Aldershot, 200, p. 3; M. Jacques, above note 1, p. 235; Bríd Ní Ghráinne, “Safe Zones and the Internal Protection Alternative”, British Institute of International and Comparative Law, Vol. 69, No. 2, 2020, p. 336.
  • 5Katy Long, “In Search of Sanctuary: Border Closures, ‘Safe’ Zones and Refugee Protection”, Journal of Refugee Studies, Vol. 26, No. 3, 2013, p. 462. On protracted conflicts and displacement, see, e.g., Ellen Policinski and Jovana Kuzmanovic, “Protracted Conflicts: The Enduring Legacy of Endless War”, International Review of the Red Cross, Vol. 101, 2019, pp. 971–2.
  • 6Although the State where a safe zone is located has obligations towards internally displaced persons (IDPs), this paper will focus on the obligations of refugee-receiving States towards those who are able to, or desire to, cross international borders in search of protection. For the IDP framework, see, e.g., Catherine Phuong, The International Protection of Internally Displaced Persons, Cambridge University Press, Cambridge, UK, 2005.
  • 7Based on an assessment of literature, these are the most commonly used terms. For “conventional” safe zones, see H. Yamashita, above note 4; P. Orchard, above note 2, p. 60. For “imposed” safe zones, see W. C. H. Chau, above note 2, p. 198; Emanuela-Chiara Gillard, “‘Safe Areas’: The International Legal Framework”, International Review of the Red Cross, Vol. 99, 201, pp. 1088–93.
  • 8IHL foresees the possibility of establishing safe zones to protect the wounded and sick in Geneva Convention (I) for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field of 12 August 1949, 75 UNTS 31 (entered into force 21 October 1950) (GC I), Art. 23, and for the civilian population in Geneva Convention (IV) Relative to the Protection of Civilian Persons in Time of War of 12 August 1949, 75 UNTS 27 (entered into force 21 October 1950) (GC IV), Arts 14 and 15; Protocol Additional (I) to the Geneva Conventions of 12 August 1949, and Relating to the Protection of Victims of International Armed Conflicts, 1125 UNTS 3, June 1977 (entered into force 7 December 197) (AP I), Arts 59 and 60. See more on the IHL framework in E.-C. Gillard, ibid., pp. 1077–7.
  • 9Regardless of this additional protection, civilians cannot be targeted at any time except if and for such time as they directly participate in hostilities and must be factored into the proportionality assessment for attacks against legitimate targets. Marco Sassòli, International Humanitarian Law: Rules, Controversies, and Solutions to Problems Arising in Warfare, Elgar, Cheltenham, 201, pp. 241–2 and 376–8.
  • 10AP I, Art. 60.
  • 11Jean-Marie Henckaerts and Louise Doswald-Beck (eds), Customary International Humanitarian Law, Vol. 1: Rules, Cambridge University Press, Cambridge, 2005 (ICRC Customary Law Study), rule 35, available at:
  • 12M. Sassòli, above note 9, p. 242.
  • 13M. Jacques, above note 1, p. 235; G. Gilbert and A. M. Rüsch, above note 2, p. ; Rutger Birnie and Jennifer Welsh, “Displacement, Protection and Responsibility: A Case for Safe Areas”, Global Responsibility to Protect, Vol. 10, No. 3, 2018, p. 337.
  • 14W. C. H. Chau, above note 2, p. 194.
  • 15M. Jacques, above note 1, p. 241; E.-C. Gillard, above note 7, p. 1088.
  • 16The Open Relief Centres in Sri Lanka have been considered as most similar to protected zones under IHL, and consequently the most successful. These spaces were a “temporary” place for displaced persons to reside and receive “relief assistance”, which were consented to by both parties to the conflict and formally recognized in a memorandum of understanding. M. Jacques, above note 1, pp. 238–40; P. Orchard, above note 2, p. 60.
  • 17See, e.g., M. Jacques, above note 1, pp. 235–44; David Keen, “Anything but Safe: Problems with the Protection of Civilians in So-Called ‘Safe Zones’”, Working Paper Series No. –187, London School of Economics and Political Science, November 20, p. 36.
  • 18Charter of the United Nations, 1 UNTS XVI, 26 June 1945 (entered into force 24 October 1945), Art. 2(4).
  • 19The first internationally sanctioned safe zone in Northern Iraq (91) did arguably offer some immediate protection but its fragile legal basis has been heavily criticized. In 93, the UN Security Council explicitly authorized a safe zone in Srebrenica. The concentration of civilians without sufficient protection ultimately led to the July 95 massacre by Bosnian Serb forces. Other UN-sanctioned zones in Somalia (92) and Rwanda (94) also failed to prevent armed attack because of the lack of consent from all belligerents and continued militarization. M. Jacques, above note 1, pp. 240–1; P. Orchard, above note 2, p. 60; R. Birnie and J. Welsh, above note 13, p. 337.
  • 20W. C. H. Chau, above note 2, pp. 198–2.
  • 21UN Charter, above note 18, Art. 27.
  • 22H. Yamashita, above note 4, p. 193.
  • 23See, e.g., M. Jacques, above note 1, pp. 240–41.
  • 24For an overview of the challenges posed by the exercise of the veto, see, e.g., Frédéric Mégret, “The Security Council”, in Frédéric Mégret and Philip Alston (eds), The United Nations and Human Rights: A Critical Appraisal, Oxford University Press, London, 2020, pp. 39–98; UN Security Council, The Veto, Security Council Report, Research Report No. 3, 19 October 2015, available at:….
  • 25This is in line with the Human Rights Committee's complementarity approach and the position of Professor Vincent Chetail. Vincent Chetail, “Armed Conflict and Forced Migration: A Systemic Approach to International Humanitarian Law, Refugee Law and Human Rights Law”, in Andrew Clapham and Paola Gaeta (eds), The Oxford Handbook of International Law in Armed Conflict, Oxford University Press, Oxford, 2014, pp. 700–34; Vincent Chetail, “Moving towards an Integrated Approach of Refugee Law and Human Rights Law”, in Cathryn Costello, Michelle Foster and Jane McAdam (eds), The Oxford Handbook of International Refugee Law, Vol. 1, Oxford University Press, Oxford, 2021, pp. 210–12.
  • 26For the applicability of IHRL in armed conflict, see International Court of Justice, Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory, Advisory Opinion, ICJ Reports 136, 2004, para. 106; General Comment No. 31, Nature of the General Legal Obligation Imposed on States Parties to the Covenant, UN Doc. CCPR/C/21/Rev.1/Add.13, May 2004, para. 11. For the material scope of application of IHL, see, e.g., M. Sassòli, above note 9, pp. 168–85.
  • 27G. Gilbert and A. M. Rüsch, above note 2, p. 3; B. Ní Ghráinne, above note 4, p. 336.
  • 28Convention relating to the Status of Refugees, 189 UNTS 137, July 1951 (entered into force 22 April 1954) (Refugee Convention), Art. .
  • 29GC IV, Art. 35.
  • 30M. Sassòli, above note 9, p. 336.
  • 31V. Chetail, “Armed Conflict and Forced Migration”, above note 25, pp. 706–10.
  • 32M. Sassòli, above note 9, p. 297.
  • 33See, e.g., International Covenant on Civil and Political Rights, 999 UNTS 17, 16 December 1966 (entered into force 23 March 1976) (ICCPR), Art. 12(2). The right to leave has been reaffirmed in numerous treaties and argued as customary in nature; see Vincent Chetail, International Migration Law, Oxford University Press, Oxford, 2019, pp. 82–92.
  • 34E.-C. Gillard, above note 7, p. 1097. See also General Comment No. 31, above note 26, para. 10: “anyone within the power or effective control” of a State party.
  • 35E.-C. Gillard, above note 7, pp. 1098–9; W. C. H. Chau, above note 2, pp. 210–14.
  • 36See, e.g., ICCPR, Art. 4.
  • 37General Comment No. 29: Article 4: Derogations During a State of Emergency, UN Doc. CCPR/C/21/Rev.1/Add.11, 31 August 2001, para. 3.
  • 38For detailed analysis, see Marko Milanovic, “Extraterritorial Derogations from Human Rights Treaties in Armed Conflict”, in Nehal Bhuta (ed.), The Frontiers of Human Rights, Vol. 1, Oxford University Press, Oxford, 2016.
  • 39G. Gilbert and A. M. Rüsch, above note 2, p. 5.
  • 40V. Chetail, “Armed Conflict and Forced Migration”, above note 25, p. 716. See, e.g., African Charter on Human and Peoples’ Rights, Art. 12(2); International Convention on the Rights of the Child, Art. 10(2) and others.
  • 41Ibid. See, e.g., European Court of Human Rights, Isayeva v. Russia, Case No. 57950/00, 24 February 2005.
  • 42ICCPR, Art. 12(3); General Comment No. 27: Article 12 (Freedom of Movement), UN Doc. CCPR/C/21/Rev.1/Add.13, 2 November 1999, paras 14–16.
  • 43V. Chetail, above note 33, p. 80.
  • 44Ibid., pp. 84–5.
  • 45This analysis focuses on the obligations of refugee-receiving States to respect the right to leave as practice has shown that safe zones would be established with the intention to prevent individuals from entering the State of asylum, rather than by the home State to prevent the flight of its nationals.
  • 46Universal Declaration of Human Rights, UN General Assembly Resolution 217 A(III), 10 December 1948, Art. 14.
  • 47Alice Edwards, “Human Rights, Refugees, and The Right ‘To Enjoy’ Asylum”, International Journal of Refugee Law, Vol. 17, No. 2, 2005, p. 301.
  • 48Vincent Chetail, “Are Refugee Rights Human Rights? An Unorthodox Questioning of the Relations between Refugee Law and Human Rights Law”, in Ruth Rubio-Marín (ed.), Human Rights and Immigration, Vol. 1, Oxford University Press, Oxford, 2014, p. 31.
  • 49Refugee Convention, Art. 33. See also Rebecca M. M. Wallace, “The Principle of Non-refoulement in International Refugee Law”, in Vincent Chetail and Céline Bauloz (eds), Research Handbook on International Law and Migration, Elgar, Cheltenham, 2014, p. 418.
  • 50GC IV, Art. 45. According to this provision, transfer is also prohibited if the destination state is not party to GC IV or is not willing or able to respect it. For analysis of non-refoulement under IHL see Vincent Chetail, “The Transfer and Deportation of Civilians”, in Andrew Clapham, Paola Gaeta and Marco Sassòli (eds), The 1949 Geneva Conventions: A Commentary, Oxford University Press, Oxford, 2015, pp. 1187–9 and 1198–209.
  • 51General Comment No. 31, above note 26, para. 12. Non-refoulement has been explicitly endorsed in the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, 1465 UNTS 85, 10 December 1984 (entered into force 26 June 1987), Art. 3. It is also recognized as implicit in the European Convention on Human Rights; see European Court of Human Rights, Soering v UK, Application No. 14038/88, 7 July 1989 and a customary norm, arguably amounting to jus cogens. See V. Chetail, above note 48, pp. 29–39.
  • 52See, e.g., European Court of Human Rights, Sufi and Elmi v. United Kingdom, Application Nos. 8319/07 and 1149/07, 28 June 2011 in V. Chetail, above note 50, p. 1204. For analysis of the scope of non-refoulement, see Emanuela-Chiara Gillard, “There's No Place like Home: States’ Obligations in Relation to Transfers of Persons”, International Review of the Red Cross, Vol. 90, 2008, pp. 716–23; V. Chetail, above note 48, pp. 34–9.
  • 53V. Chetail, above note 48, p. 35.
  • 54For an assessment of Art. 33(2), see Aoife Duffy, “Expulsion to Face Torture – Non-Refoulement in International Law”, International Journal of Refugee Law, Vol. 20, No. 3, 2008, pp. 374–7. For the application of the refugee definition to situations of armed conflict, see V. Chetail, “Armed Conflict and Forced Migration”, above note 25, pp. 722–7.
  • 55K. Long, above note 5, pp. 462–7.
  • 56UN High Commissioner for Refugees (UNHCR), “Advisory Opinion on the Extraterritorial Application of Non-Refoulement Obligations under the 1951 Convention relating to the Status of Refugees and its 1967 Protocol”, 26 January 2007, para. 7, available at:
  • 57European Court of Human Rights, Hirsi Jamaa and Others v. Italy, Application No. 27765/09, 23 February 2012, paras 180–1.
  • 58B. Ní Ghráinne, above note 4, p. 344.
  • 59Bríd Ní Ghráinne, “Internally Displaced Persons and International Law”, in Satvinder Singh Juss (ed.), Research Handbook on International Refugee Law, Elgar, Cheltenham, 2019, pp. 37–8.
  • 60This derives from the wording of the definition of a refugee as “outside the country of his nationality” and was affirmed in R v. Immigration Officer at Prague Airport and Another, Ex parte European Roma Rights Centre and Others [2004] UKHL 55. See V. Chetail, above note 48, p. 36.
  • 61This is based on the ordinary meaning of the Refugee Convention, in light of its object and purpose, taking into account subsequent practice under Article 31 of the Vienna Convention on the Law of Treaties. Thomas Gammeltoft-Hansen, “Extraterritorial Migration Control and the Reach of Human Rights”, in V. Chetail and C. Bauloz (eds), Research Handbook on International Law and Migration, Elgar, Cheltenham, 2014, p. 116; UNHCR, above note 56, paras 23–4.
  • 62UNHCR, above note 56, paras 42–3.
  • 63V. Chetail, above note 48, pp. 36–7.
  • 64As affirmed by the UNHCR; V. Chetail, “Armed Conflict and Forced Migration”, above note 25, pp. 719–20.
  • 65See, e.g., European Court of Human Rights, MSS v. Belgium and Greece, Application No. 30696/09, 21 January 2011, para. 223.
  • 66The prohibition of collective expulsion is explicit in some regional treaties, acknowledged by the Human Rights Committee as implicit in Article 13 of the ICCPR and is arguably a customary norm. V. Chetail, “Armed Conflict and Forced Migration”, above note 25, pp. 720–1; V. Chetail, above note 33, pp. 139–42.
  • 67Jessica Schultz, The Internal Protection Alternative in Refugee Law, Brill Nijhoff, Boston, MA, 2019, p. 2.
  • 68Convention Governing the Specific Aspects of Refugee Problems in Africa, 1001 UNTS 45, 10 September 1969 (entered into force 20 June 1974) (OAU Convention), Art. I(2).
  • 69Jessica Schultz, “The Internal Protection Alternative and its Relation to Refugee Status”, in S. S. Juss (ed.), Research Handbook on International Refugee Law, Elgar, Cheltenham, 2019, p. 129.
  • 70Ibid.
  • 71UNHCR, Guidelines on International Protection: “Internal Flight or Relocation Alternative” within the Context of Article 1A (2) of the 1951 Refugee Convention and/or 1967 Protocol relating to the Status of Refugees, UN Doc. HCR/GIP/03/04, 23 July 2003, paras 6–7.
  • 72Article 8 of the EUQD defines the IPA and explicitly refers to Article 7 which defines actors of protection. See Council Directive 2011/95/EU of 13 December 2011 [2011] OJ L 337/9-337/26 (EUQD).
  • 73UNHCR, above note 71, para. 13.
  • 74Ibid.
  • 75P. Orchard, above note 2, p. 68.
  • 76E.-C. Gillard, above note 7, p. 1093.
  • 77See, e.g., Caelin Briggs, “Protection of Civilians Sites: Lessons from South Sudan for Future Operations”, Norwegian Refugee Council, 31 May 2017, available at:….
  • 78UNHCR, above note 71, para. 17. “Protection must be effective and of a durable nature … provided by an organised and stable authority exercising full control over the territory and population in question.”
  • 79The limited capacity of international organizations to offer protection has been recognised by the UNHCR. Ibid., para. 16.
  • 80Damian Lilly, “Protection of Civilians Sites: A New Type of Displacement Settlement?” Humanitarian Exchange, September 2014, pp. 31–3, available at:….
  • 81E.-C. Gillard, above note 7, p. 1095.
  • 82EUQD, Art. 8(1). Similar notions of reasonableness and settlement have been endorsed by the UNHCR. See UNHCR, above note 71, pp. 24–30.
  • 83B. Ní Ghráinne, above note 4, p. 363.
  • 84UNHCR, Interpreting Article 1 of the 1951 Geneva Convention Relating to the Status of Refugees, Geneva, April 2001, para. 13.
  • 85B. Ní Ghráinne, above note 4, p. 350.
  • 86G. Gilbert and A. M. Rüsch, above note 2, p. 2.
  • 87B. Ní Ghráinne, above note 4, p. 351.
  • 88Ibid., p. 344.
  • 89EUQD, Art. 8(2); UNHCR, above note 71, para. 4.
  • 90See, e.g., Universal Declaration of Human Rights, Art. 13(2); ICCPR, Art. 12(4). See also ICRC Customary Law Study, above note 11, rule 132, which recognizes a right to return in IAC and NIAC. Articles 35, 45 and 49(1) of GC IV implicitly permit the voluntary return of civilians. For analysis, see V. Chetail, “Armed Conflict and Forced Migration”, above note 25, pp. 728–9.
  • 91K. Long, above note 5, p. 471.
  • 92GC IV, Art. 45(4).
  • 93V. Chetail, above note 50, pp. 1190–202.
  • 94Cédric Cotter, Displacement in Times of Armed Conflict: How International Humanitarian Law Protects in War and Why it Matters, ICRC's IHL Impact Series, Geneva, April 2019, pp. 55–6.
  • 95R. Birnie and J. Welsh, above note 13, p. 339.
  • 96Vincent Chetail, “Voluntary Repatriation in Public International Law: Concepts and Contents”, Refugee Survey Quarterly, Vol. 23, No. 1, 2004, p. 19; UNHCR, Handbook on Voluntary Repatriation: International Protection, January 19, Chapter 2, available at: The “voluntary character of repatriation” is also endorsed in Article 5 of the OAU Convention.
  • 97V. Chetail, ibid., pp. 17–18. This is affirmed by Objective 21 of UN General Assembly, Global Compact for Safe, Orderly and Regular Migration, 19 December 2019, 73rd Session, UN Doc. A/RES/73/195, para. 37.
  • 98UNHCR, Global Consultations on International Protection/Third Track: Voluntary Repatriation, UN Doc. EC/GC/02/5, 25 April 2002, para. 15; UNHCR, above note 96, p. 11.
  • 99Katy Long, The Point of No Return: Refugees, Rights, and Repatriation, Oxford University Press, Oxford, 2013, p. 174.
  • 100K. Long, above note 5, p. 459.
  • 101As recognized by the ICRC, a key factor that impedes the return of people displaced by armed conflict is the ongoing impact of the destruction of their homes and other infrastructure that are essential to meeting their basic needs on return, including electricity, drinking water and health-care services. Reconstruction processes in a safe zone could overcome this potential hurdle to return. C. Cotter, above note 94, p. 56.
  • 102ICCPR, Art. 9.
  • 103G. Gilbert and A. M. Rüsch, above note 2, p. 7.
  • 104R. Birnie and J. Welsh, above note 13, p. 332.

Continue reading #IRRC No. 919

More about International Humanitarian Law (IHL), IHL, Migrants, refugees, asylum seekers, armed conflict