Introduction
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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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”.Footnote 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.Footnote 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.Footnote 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),Footnote 11 in which zones could be established by special agreements under Article 3 common to the four Geneva Conventions.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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,Footnote 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.Footnote 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.Footnote 20 However, they still require the consent or non-veto of the Security Council's permanent five members.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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,Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 36 This has been restrictively interpreted as requiring more than the existence of an armed conflict.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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”.Footnote 42 These conditions therefore significantly limit a State's ability to prevent flight, and were notably drafted as the “exception” rather than the “rule”.Footnote 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.Footnote 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.Footnote 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,Footnote 46 which has been argued as implicit in the Refugee Convention and an emerging customary norm.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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”.Footnote 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.Footnote 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.Footnote 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.Footnote 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,Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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”.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.Footnote 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.
