Externalizing Asylum

A compendium of scientific knowledge

“Externalisation, Access to Territorial Asylum, and International Law.” Analytical Paper

David Cantor, Nikolas Feith Tan, Mariana Gkliati, Elizabeth Mavropoulou, Kathryn Allinson, Sreetapa Chakrabarty, Maja Grundler, Emilie McDonnell, Lynn Hillary, Riona Moodley, Stephen Phillips, Annick Pijnenburg, Adel-Naim Reyhani, Sophia Soares and Natasha Yacoub[1]

 

‘Externalisation’ – the process of shifting functions normally undertaken by a State within its own territory so they take place, in part or in whole, outside its territory – has an important impact on the asylum field. This paper provides a primarily legal analysis of the key international law considerations on externalisation outlined in the Refugee Law Initiative 2022 Declaration on Externalisation and Asylum. It starts by offering a distinctive conceptualisation of ‘externalisation’, before turning to the overarching international law principles that govern the kinds of conduct that externalisation in this field tends to involve. An analysis of the international law parameters of externalisation is then presented. We end by exploring issues of accountability that arise in such contexts.

 

1. Introduction

‘Externalisation’ – the process of shifting functions normally undertaken by a State within its own territory so they take place, in part or in whole, outside its territory – has an important impact on the asylum field. Particularly (but not exclusively) in the Global North, the body of laws, policies and practices that externalise aspects of the migration and asylum functions of States appears to be expanding. This trend has not gone unnoticed or unchallenged by international organisations (IOs), non-governmental organisations (NGOs), news media and academia concerned about the serious negative consequences of this trend for refugees and asylum seekers. Certainly, the process of externalisation, at least as it impinges on access to territorial asylum, raises a host of complex legal, moral and policy concerns.

The Refugee Law Initiative (RLI) Declaration on Externalisation and Asylum, adopted at its 6th Annual Conference on 29 June 2022, sets out key international law considerations on externalisation that reflect the view of a range of independent experts and scholars at the RLI. This paper provides a primarily legal analysis for the positions outlined therein.[2] It thus acknowledges but does not engage in depth with the political narrative on externalisation that runs in parallel. The paper draws on collaborative work by the many RLI staff and RLI Research Affiliates who researched and authored this paper over the preceding nine months.[3] The invaluable input on drafts of the Declaration and of this paper by RLI Senior Research Associates is also gratefully acknowledged.[4]

The paper starts by offering a distinctive conceptualisation of ‘externalisation’ as a process that can impinge on access to asylum in the territory of States (section 2). It then seeks to identify the overarching international law principles that govern the kinds of conduct that externalisation in this field tends to involve (section 3). An analysis of the international law parameters of externalisation is then presented, firstly for externalised border controls – with extraterritorial pushbacks of refugees and others taken here as a prime example – (section 4), and secondly for externalised asylum systems – focusing on third country processing of asylum claims, as the greatest source of concern in this regard (section 5). We end by exploring issues of accountability that arise in both contexts (section 6).

 

2. Conceptualising externalisation

Externalised migration control practices have a long history. The US, for example, first introduced visa requirements in 1924.[5] Carrier sanctions have even longer use, stretching back at least as far as the 19th century in relation to passport controls.[6] However, it was in the 1980s that States in North America, Europe and elsewhere began adopting increasingly far-reaching measures to stem irregular movement by refugees and others towards their shores through the imposition of visa regimes and carrier sanctions and via interceptions on the high seas or in the territory of third States.[7] In 1999, a regional dimension appeared, as the Tampere conference called for cooperation with ‘external’ countries outside the European Union (EU) in implementing the common European asylum system.[8] These dynamics have also generated increasing interest on the part of researchers in understanding the implications for refugees and asylum seekers of such externalised forms of migration control, and the parallel efforts by some States to externalise asylum responsibilities through third country processing arrangements of the kind implemented by the US and Australia.[9]

However, ‘externalisation’ is not a legal term of art in the asylum field. Indeed, its basic meaning is contested by scholars and policy actors, who have defined the term in differing ways.[10] Moreover, measures closely related to, or overlapping with, the externalisation of asylum and/or migration controls have also been characterised in the literature under a range of alternative labels, including ‘remote control’, ‘non-entrée’, ‘deterrence’, ‘off-shoring’, ‘extra-territorialisation’ and ‘protection elsewhere’. Against this potentially confusing backdrop, it is important to clarify how we conceptualise ‘externalisation’ in this analysis.

Like most other studies in the refugee field, we see externalisation as an ‘umbrella’ concept.[11] However, in contrast to most studies, we do not see the concept as defined solely by reference to asylum or migration concerns. Rather, more generally, it refers to the process of shifting functions which are normally undertaken by a State within its own territory so that they take place, in part or in whole, outside its territory. This broader focus is justified, since State functions across a range of different areas of governance beyond just asylum or migration can clearly be externalised.[12] It emphasises that externalisation in the asylum space does not exist in isolation from broader trends in governance, but rather forms part of them.

This approach is useful for another reason. The comparatively narrower existing attempts to define externalisation in our field have struggled to reconcile apparently externalised measures that restrict access to the territory for migrants in general, including asylum seekers, with those that offshore the processing of asylum claims specifically.[13] We believe that this is because those elements actually speak to interconnected but conceptually and often practically distinct State functions, respectively the operating of border controls on the one hand and the administration of a national asylum system on the other. Clearly, each of these distinct (and intensely-politicised) functions has seen an increasing emphasis on externalisation in a range of countries in recent years.

As we are concerned here precisely with those State functions that impinge on access to asylum in the territory of a State, our analysis focuses on how externalisation processes play out in relation to these two specific State functions. Of course, other State functions may also impinge on access to territorial asylum, and may also be increasingly externalised in practice. Nonetheless, by focusing on the two State functions that seem to us to be among the most relevant to regulating access to territorial asylum, we aim to identify the legal parameters that are most pertinent in the context of the challenge to asylum posed by current externalisation proposals and practices.

Alongside the attempt to shift certain State functions outside its own territory, externalisation often involves attempts to shift responsibility for the externalised measures to other entities. Thus, whilst externalised State functions that impinge on access to territorial asylum may be implemented by a State unilaterally, they can also be implemented jointly with other States and/or entities (including IOs and private actors) to which the State may have partially or wholly outsourced those functions.[14] This attempt to shift or share responsibility is a frequent practical corollary of States shifting border control or asylum functions outside their territory but, unlike some scholars,[15] we do not see it as a necessary feature of externalisation.[16] Certainly, though, it introduces an additional level of legal and practical complexity to many externalisation practices.

 

3. Applicability of international law to externalisation processes

Externalisation presents challenges to establishing the applicable legal rules. This is because many of the pertinent policies and practices exploit, deliberately or otherwise, ‘grey areas’ in which the international law rules governing refugee protection can appear to be silent. This is largely due to the extraterritorial component of externalisation (3.1) but can also reflect the tendency of States to delegate or share authority, with not only other States but other entities too (3.2). Thus, a key task for the international lawyer is to accurately identify the pertinent legal rules and their interaction with one another in the factual context of externalisation.

 

3.1 Primary rules – extraterritorial applicability

Externalisation raises questions about whether relevant specialised regimes of primary rules of international law, such as refugee law and human rights law, remain applicable to practices undertaken outside the State’s own territory. This is because, even if externalisation laws and policies are adopted on the territory of a State, their impact is usually felt beyond its borders, by virtue of the subject matter. It is therefore necessary to evaluate the legal application of different bodies of primary rules of international law to the context of externalisation.

General international law prohibits States from interfering in matters essentially within the domestic jurisdiction of other States or acting in any manner inconsistent with the United Nations (UN) Charter or with rules of international custom governing the relations between States.[17] The implications of sovereignty for other States thus form an initial check on the legality of measures taken by one State that may have effects in the territory of another. These rules are complemented by those from specialised regimes of international law that are concerned directly with the protection of the human person.

International refugee law establishes key obligations towards refugees. The applicability of the Refugee Convention is not limited by a ‘jurisdiction’ clause, meaning that States parties remain bound to observe the treaty when engaged in conduct outside their own territory.[18] However, the range of Convention guarantees applicable to any putative refugee will vary depending on the degree of attachment of the refugee to the State party exercising jurisdiction over her.[19] Moreover, persons who have not yet left their own country are not protected by the treaty as they are not yet considered to be ‘refugees’ under its terms.[20] The prohibition on refouling refugees in Article 33(1) of the Convention may well also be a customary norm.[21] A good faith duty of cooperation and responsibility-sharing is implicit in international refugee law and expressed by instruments such as the Global Compact on Refugees.[22]

The international law of armed conflict applies to any State party to an armed conflict that meets the definitional legal criteria.[23] In non-international armed conflicts, non-State parties to the conflict are also bound directly by the applicable rules.[24] Many of the rules expressed in the 1949 Geneva Conventions and their Additional Protocols now appear to be also customary in nature.[25] The general rules on the protection of civilians will apply where States involved in externalisation measures are a party to the conflict.[26] The special guarantees for civilians in the hands of a party to an international armed conflict (i.e. as persons in territory under occupation or as aliens in the power of a State party to a conflict) will be particularly relevant.[27] The rules of this specialised regime may shape the content of complementary human rights norms where the two bodies of law overlap in application.[28]

The law of the sea as codified in the United Nations Convention on the Law of the Sea (UNCLOS) applies to all externalisation practices undertaken within the various maritime zones establishing States’ jurisdictional competencies, i.e. the internal waters, territorial waters, the contiguous zone, the exclusive economic zone and the high seas.[29] UNCLOS and other maritime treaties prescribe certain duties for States that are applicable to externalisation measures, such as the duty to rescue people in distress at sea and to disembark them at a place of safety (which is also widely accepted as customary)[30] and the duty of coastal States to maintain effective search and rescue services within their SAR designated zones.[31]

International human rights law, by contrast, applies to all human beings regardless of their situation.[32] A multitude of UN and regional human rights treaties provide for human rights standards and/or prohibit discrimination on the basis of protected characteristics. However, the application of these obligations ratione loci is often conditioned on the State party exercising power through ‘jurisdiction’.[33] ‘Jurisdiction’ in this sense is normally presumed to exist throughout a State’s own territory, and may persist even in parts of that territory not under its effective control.[34] Yet international courts and treaty bodies have consistently determined that such jurisdiction can also occur in some situations where a State acts beyond its own borders.[35] The following broad trends in the jurisprudence of these bodies on extraterritorial ‘jurisdiction’ affirm the continuance of human rights duties in many externalisation contexts.

Firstly, there is a strong consensus among treaty bodies that extraterritorial jurisdiction for the purposes of human rights law persists wherever a State acting outside its borders exercises ‘effective control’ over a territory or ‘authority and control’ over a person.[36] The former is analogous to a State’s jurisdiction over its own territory and requires the State to secure ‘the entire range of substantive rights’ in the treaty.[37] Some UN treaty body practice suggests that jurisdiction on the basis of control over a person may equally extend the full gamut of treaty rights,[38] a position that is similar in some regional bodies,[39] although the European Court of Human Rights (ECtHR) has held that only those specific rights ‘relevant to the situation of that individual’ apply (i.e. the treaty rights can be ‘divided and tailored’ by situation).[40]

Secondly, UN human rights treaty bodies increasingly find common ground also in a ‘direct and foreseeable effects’ test for extraterritorial jurisdiction.[41] This so-called ‘functional’ approach does not require any control over the person or territory. Rather, if a State party has knowledge at the time it acts that the risk of an extraterritorial violation of protected rights is a ‘foreseeable consequence’ of its actions, merely being ‘a link in a causal chain that makes possible’ the violation of rights is sufficient for jurisdiction to exist.[42] A similar approach can be discerned among regional treaty bodies. Thus, the IACtHR affirms that jurisdiction can exist where there is a ‘causal link’ between an act over which the State exercises ‘effective control’ and the negative impact on ACHR rights of persons outside its territory’.[43] The AfComHR has held that jurisdiction can arise as a result of conduct that ‘could reasonably be foreseen to result in’ the violation of ACHPR rights.[44] Likewise, certain strands of ECtHR case-law suggest that the ECHR applies to executive or judicial measures directed at persons abroad.[45] The ECtHR has held that the jurisdiction clause ‘cannot be interpreted so as to allow a State party to perpetrate violations of the Convention on the territory of another State, which it could not perpetrate on its own territory’.[46] Yet, for extraterritorial effects to trigger jurisdiction on this functional approach, a sufficiently close causal link between the State conduct and the ensuing violation of rights must exist.[47]

Finally, any externalisation measures that claim to derogate from treaty obligations will need to be carefully justified against that treaty’s derogation criteria.[48] Although there is no formal bar on derogation in the context of extraterritorial conduct, the factual scenarios of border control or asylum are not likely, in themselves, to meet treaty thresholds for derogation in any but the most extreme circumstances.[49] In general, there is no basis in international law for States to refuse to receive or register asylum applications due to emergencies or crises.[50]

 

3.2 Secondary rules – responsibility, actors and attribution

In international law, responsibility is a corollary of obligation. A State’s breach of an international legal obligation, insofar as it is attributable to that State, entails its international responsibility, including a duty to make reparation.[51] As such, the rules of responsibility represent secondary rules of international law that help to determine the content and enforcement of primary rules.[52] The residual general regime on State responsibility is codified in the 2001 ILC Draft Articles on Responsibility of States for Internationally Wrongful Acts (ARSIWA) and that for IOs in the 2011 ILC Draft Articles on Responsibility of International Organisations (ARIO).[53] Crucially, then, these rules explain when States and/or IOs incur international legal responsibility for externalisation measures, including those that are shared with, or delegated to, other States, IOs or private entities.

The first condition for responsibility under international law is that the State (or IO) has committed an ‘internationally wrongful act’.[54] This occurs when a State breaches an international obligation incumbent upon it,[55] and none of the ‘circumstances precluding wrongfulness’ apply.[56] An internationally wrongful act can result from ‘any violation by a State of any obligation, of whatever origin…’. [57] As a matter of international law, then, it is international law alone that governs whether an obligation has been breached or not, regardless of any purported lawfulness of the conduct under national law.[58] Moreover, an internationally wrongful act may comprise of actions or omissions, singly or jointly, or a combination of both.[59] Yet an obligation need not be breached by conduct of the same description – rather, ‘any action… that is incompatible with these obligations is unlawful, regardless of the means by which it is brought about’.[60] Indeed, ICJ case law suggests that ‘responsibility may be attributed whenever a State, within whose territory substantial transboundary harm is generated, has knowledge or means of knowledge of harm and the opportunity to act’.[61]

The second condition for responsibility under international law is that the internationally wrongful act is attributable to the State (or IO). Most obviously, a State is responsible for the official actions and omissions of its de jure organs, whether legislative, executive or judicial in character, and any de facto organs (i.e. those ‘completely dependent’ on the State).[62] This includes but is not limited to law enforcement organs such as police, the coast guard, immigration officials and its armed forces, whatever their function and at whatever level (national, provincial, local etc.) they operate.[63] It also includes conduct by the master of a naval vessel, which is a de jure organ of the State and thus attributable to the flag State.[64] Purely private acts by officials are not attributable to the State but conduct in excess of authority or in contravention of instructions by officials remains attributable to the State,[65] as do any other acts committed by them in an apparently official capacity or under ‘colour of authority’.[66] Unreasonable acts of private violence by officials and a failure to take appropriate steps to punish the culprits may also give rise to State responsibility.[67]

The conduct of non-State entities may sometimes be attributed to a State. The distinction between public and private capacities is not always self-evident, especially when governments delegate externalised functions to parastatal or private entities.[68] Nonetheless, where a person or entity is empowered by the law of a State to exercise elements of governmental authority, any actions in that capacity are attributable to the State.[69] This is highly pertinent in light of the tendency of governments to delegate externalised government functions to airlines and security companies. More generally, specific acts or omissions are attributable to a State if the person, or group of persons, carrying out the conduct is ‘acting on the instructions of’, or under the ‘effective control’ of, that State.[70]

Finally, scenarios exist where one State (or IO) aids another. If the organ of one State is placed at the disposal of another and acts exclusively on behalf of it, then its conduct is attributed to the latter State only.[71] By contrast, where a State (or IO) aids or assists another in committing an internationally wrongful act, the former State (or IO) also has responsibility to the extent that its own conduct contributed to the internationally wrongful act, so long as it had ‘knowledge of the circumstances’ of the wrongful act and the act would be internationally wrongful if committed by that State.[72] It is accepted that, in the case of human rights violations, a State (or IO) may incur international responsibility if it provides, for example, material aid to a State (or IO) that uses the aid to commit human rights violations, where the assisting State acted with knowledge of such an eventuality.[73] Indeed, in contexts where human rights are at stake, some authors argue that a lower threshold of knowledge on the part of complicit States, such as ‘constructive knowledge’, is appropriate to attribute responsibility to them.[74]

 

4. Externalised border controls

Border control is one of the key areas where externalisation is practised. Border control is viewed by most States as an essential function for regulating the flows of people and goods into their territory. It usually includes interception at a State’s borders or within its territory of persons who wish to enter or stay in the territory and the assessment of those claims. Indeed, many States now claim a sovereign prerogative to regulate entry and stay in their territory, albeit one that they have long recognised to be qualified by international obligations.[75]

In this context, one border control measure that has become increasingly widespread in recent years is the use of ‘pushbacks’, which involve intercepting and summarily forcing back persons arriving at the border without assessing their claims for entry or protection.[76] In line with the increasing externalisation of border controls globally, States now frequently carry out or assist pushbacks that are implemented outside their own territories, with the aim of preventing ‘upstream’ any onward irregular movement to their territories. Such measures thus ‘prevent migrants from approaching or crossing the border’ in the first place.[77] They raise serious questions of compatibility with international law.

The ever greater ubiquity of pushbacks, their increasing use as an externalised practice, and the real dangers they pose for people ‘pushed back’, make them a paradigmatic contemporary case study of the legality of externalised border control measures in general. However, it is important to emphasise that pushbacks are not the only externalised border controls that seek to limit entry to territory. Indeed, ‘remote’ visa regimes and carrier sanctions, as well as ‘juxtaposed’ border controls (i.e. pre-embarkation checkpoints located outside the country), all form part of the wider raft of externalised border controls pursued by States, especially in the Global North, that seek to limit not only access to their territories but also travel towards them by people who are travelling irregularly.[78]

Here, then, we describe the use of externalised pushbacks in practice (4.1), before delineating the international law applicable to them (4.2) and, by extension, to other kinds of externalised border controls implemented outside the territory of the State concerned. While at least some externalised border controls have the positive potential to be used as a protection tool – as with the humanitarian visas that facilitate access to refuge in Brazil for Syrians (and now Ukrainians) or the Protection Transfer Arrangement that helps certain risk profiles of Central Americans escape to the USA[79] – this is not the case for pushbacks, which have the potential only for negative impact on the people to whom they are applied.

 

4.1 Pushbacks in practice

The practice of pushbacks is now seen along migration routes in various parts of the world, where it forms an integral part of border management.[80]

Land border pushbacks have recently been reported, inter alia, in Europe (for example, along the Balkan route),[81] in the Americas (for example, at the US-Mexico border),[82] and in Africa (for example, between Algeria and Niger or Libya and Morocco).[83] Land border pushbacks include so-called ‘hot returns’ (involving ‘direct deportations without individual examination directly at the border’),[84] as well as the construction of border fences without de facto accessible border crossing points.[85] Where third States support land border pushbacks in other territories with a view to limiting arrivals at their own territory, this represents an externalised form of border control. US pressure for Mexican pushbacks at its border with Guatemala, and by Guatemala at its border with Honduras, are but a few recent examples.[86]

Pushbacks also occur at sea. Here, the practice usually involves interception of the vessel, deflecting it away from the State’s own territory. Some sea pushbacks occur within the State’s territorial sea, as with the pushbacks of vessels entering Greek territorial waters from Turkey by the Greek Coast Guard,[87] and the pushing back of Rohingya boats from territorial waters to the high seas by some Association of Southeast Asian Nations (ASEAN) countries in the Andaman Sea and the Gulf of Bengal.[88] However, where a third State is involved for border control reasons, or where the pushbacks occur on the high seas, then this represents an externalised form of border control. Relevant practices include the US interdiction programme for Haitians in the 1990s,[89] or the widespread and systematic practice of pushbacks at present in the Mediterranean Sea,[90] or the pushbacks by Australia from the high seas to Malaysia, Indonesia and Vietnam.[91] The impact of pushbacks on the people concerned can have a gender dimension, including exposing women to sexual violence and other gender-specific forms of abuse.[92]

Pushbacks conducted in cooperation with a proxy State that drags people back to its own territory before they reach the destination State are known as ‘pullbacks’. These measures are often an attempt by the destination State to use the proxy State to stop arrivals ‘upstream’, and thus avoid engaging its own legal obligations towards the people concerned,[93] although dictatorial regimes may also use it to prevent inhabitants from escaping.[94] Such pullback practices are implemented in the central Mediterranean Sea as a result of EU support to Libya[95] and informal bilateral agreements between Libya and Italy and Malta and Libya.[96] Similarly, the efforts by Morocco to stop people crossing into the Spanish enclaves of Melilla and Ceuta in the context of agreements with Spain and the EU offer another example of land and sea border pullbacks.[97]

Increasingly, actors other than States are involved in pushback operations, either unilaterally or in coordination with States or at their direction.[98] In Europe, South America and parts of Africa, these actors are relied on to carry out or assist with pushback operations, often in an ancillary manner. For example, the European Border and Coastguard Agency (Frontex) has assisted EU Member States in the conduct of pushbacks at the Greek/Turkish border, Croatian border and Mediterranean/North African sea borders.[99] The Maltese government has reportedly hired private vessels to carry out operations to prevent migrant boats from reaching the Maltese search and rescue region,[100] and even to hold the persons intercepted, usually for an indefinite period of time on the high seas just outside its territorial waters.[101]

 

4.2 Legal issues

Externalising border controls is not prima facie lawful or unlawful under international law. Rather, legality is determined by the form and effect of the particular measure. Externalised measures that involve the physical presence of a State’s agents in the territory of another State will require the latter’s consent if their conduct is not to impinge unlawfully on its sovereign domestic jurisdiction. Even when such measures are the result of cooperation between two States, there remains a need for the measures to comply with international legal standards governing the treatment of persons.

Some externalised border control measures, such as visa regimes and carrier sanctions, do not necessarily require physical presence on the territory of another State. Such measures appear to be generally permitted. However, they may have unlawful effects where, for example, the visa regime or other remote measure is implemented on a discriminatory basis,[102] or in individual cases where the decision to a deny a visa forms a sufficiently close part of a chain of conduct that exposes an individual to a breach of protected human rights.[103] A State may carry responsibility, even where the decision is outsourced to a private actor such as a carrier. Crucially, the study of pushbacks provides a set of basic legal parameters that appear to be generally applicable to other forms of externalised border controls.

Pushbacks that do not allow for access to adequate individual determination of claims for asylum or international protection raise serious questions of legality under international law. They are likely to breach the prohibition on collective expulsion, a well-established principle of international law,[104] which is also expressed in UN and regional human rights treaties.[105] In essence, the prohibition translates to a due process right for each individual to have the act of removal administratively and judicially assessed.[106] In principle, it applies to all aliens, including those who have just crossed or are about to cross an international border, and regardless of lawful residence or irregular immigration status.[107]

Moreover, there is a serious risk that pushbacks will also result in refoulement contrary to international refugee law standards,[108] as well as those in human rights treaties,[109] where their summary character does not allow for an individual assessment of asylum claims and international protection needs. By denying access to territory, pushbacks may also infringe the right to leave a country,[110] and the right to seek asylum.[111] Other fundamental human rights often directly at risk during pushbacks include the freedom from torture, inhuman and degrading treatment,[112] the right to an effective remedy,[113] the right to life,[114] and freedom from enforced disappearance.[115]

The fact that pushbacks may be carried out, or supported, by a State as a form of externalised border control that is implemented outside its own territory does not render these legal parameters inapplicable. Border control measures that are unlawful in character when carried out in the territory of a State or at its borders remain so when carried out extraterritorially.[116] Thus, pushbacks that would be contrary to the international law rules outlined above if they were committed in the territory of a State will be equally unlawful if carried out extraterritorially in the land or maritime territory of another State or on the high seas.

Pushbacks at sea present additional concerns about legality under the law of the sea. This permits a State to intercept vessels in its own territorial waters only when they are suspected of having compromised their right to innocent passage.[117] Even when intercepting vessels carrying no flag, the scope of permissible interference is limited, as the vessel and the people on board remain subject to human rights and refugee law safeguards.[118] Whether the interception of vessels occurs in a State’s territorial waters, where the coastal State has full enforcement jurisdiction,[119] or its contiguous zone, where the coastal State has limited constabulary powers, including over preventing the infringement of its immigration laws,[120] or on the high seas, where freedom of navigation limits interception rights,[121] extraterritorial jurisdiction for the purpose of human rights law exists as a result of effective control over persons or the close causal link between the pushback and any resulting violations. Pushbacks are thus equally forbidden in the contiguous zone and on the high seas.

Additionally, the practice of pushbacks at sea has the potential to violate established positive obligations of search and rescue, such as the duty to render assistance to people in distress at sea under the law of the sea and the right to life, where a vessel is in distress or may be put in distress by the act of pushback.[122] The duty to render assistance to people in distress at sea is only fulfilled when the rescued people are disembarked in a place of safety.[123] For asylum seekers, this means access to land territory and effective asylum procedures in order to assess international protection needs.[124] If a place for disembarkation cannot be swiftly agreed, the International Maritime Organization (IMO) recommends that the State in whose search and rescue (SAR) zone the rescue takes place should allow for disembarkation.[125]

For the most part, the factors for determining the legality of pushbacks apply equally to ‘pullbacks’, where a proxy State agrees to prevent people from leaving its territory and/or to intercept them (usually at sea) and forcibly return them to its territory. State responsibility rules mean that the destination State is likely to share liability for any unlawful refoulement or exposure to other human rights violations as a result of the pullbacks carried out on its behalf by the proxy State,[126] where the destination State has presumed knowledge of the risks.[127] Whether pullback measures engage the prohibition on collective expulsion is more complex. A proxy State pulling people back to its territory is not directly involved in expulsion, although the pullback may well breach their human right to leave any country (for which the destination State may also share responsibility). But the destination State does engage in collective expulsion where it exercises extraterritorial jurisdiction over an act of transfer that does not allow for individual examination of cases (and this liability may be shared by the proxy State if it has knowledge of this intention by the destination State).[128]

One legal challenge in externalised pushbacks is that secondary rules of State responsibility do not provide for responsibility of the State in cases where it aids or assist a non-State actor (including an IO) in conducting pushbacks. This potentially allows States to avoid responsibility so long as the assistance does not reach the level to make the conduct attributable to them.[129] IOs may incur indirect responsibility for pushbacks where they aid or assist a State by providing vital support in the conduct of such operations, and if the IO had presumed knowledge of the ensuing collective expulsion.[130] Frontex, for example, has provided strategic guidance, data and operational and technical support to the Greek and Italian coastguards and failed to prevent systemic violations committed in the context of those operations.[131] But direct responsibility of IOs for pushbacks is also possible, for example if Frontex carries out pushbacks with its own personnel, with any wrongful conduct attributable to the EU.[132]

States sometimes insist on the legality of pushbacks by reference to readmission agreements with, or diplomatic assurances from, the State to which individuals are pushed back.[133] However, such arrangements cannot vitiate the violation of existing multilateral obligations.[134] States may also point to the allegedly ‘illegal’ conduct of people using irregular migration routes. In 2020, Spain succeeded with the latter strategy before the ECtHR in the (much-criticised) case of ND & NT.[135] States may also seek to justify pushbacks as a response to ‘emergencies’ and/or an influx of asylum seekers. Yet, in all these contexts, international law rules apply unchanged, subject to normal derogation rules.[136] Indeed, any claim to derogation for externalised border controls such as pushbacks will need to be carefully scrutinised to see whether there is a genuine state of emergency or whether States are using the language of crisis to shirk international obligations. Attempts by States, particularly in the Global North, to legitimise these practices also risk undermining principles of cooperation, solidarity and responsibility-sharing in the refugee and migration fields.[137]

Finally, where a State conducts or assists a pushback operation outside its own territory, or in any other way exercises jurisdiction for the purpose of border control over putative refugees or asylum seekers who are no longer within their own State, it remains subject to the wider standards concerning treatment of persons imposed by international human rights law, as well as the Refugee Convention guarantees for putative refugees under the jurisdiction of a State.[138] In contexts of armed conflict, pertinent rules of the laws of war equally bind the States involved, including those regulating the transfer of civilian protected persons.[139]

 

5. Externalised asylum systems

States need an asylum system to give effect to their legal obligations to refugees. At a minimum, this usually comprises a legal and institutional framework to determine refugee status and/or international protection needs and to implement obligations to refugees. This system is usually established and run by the State primarily within its own territory, although UNHCR cooperates substantively with a number of States by operating or supporting their asylum systems in line with its international protection mandate.

States have long engaged with refugee situations outside their territories, both through resettling refugees from countries of first asylum and supporting refugee protection in other countries. More recently, though, States have started externalising key elements of their own asylum systems by transferring asylum seekers who have arrived in their jurisdiction to another State for the purpose of determining their refugee status and, in some cases, providing them with territorial asylum there. Third country processing is thus distinct from more protective forms of externalised procedures, including pre-arrival processing of an asylum claim by the authorities of a destination country in another State. These approaches, which include protected entry procedures[140] and ‘regional processing’,[141] are generally for the purposes of expanding access to international protection in the destination State.

 

5.1 Third country processing in practice

A key feature of third country processing is that it involves a State externalising its own asylum system obligations towards refugees and asylum seekers (after they have arrived in its territory or jurisdiction) to other States or entities outside its territory. Such third country processing schemes take place through a range of legal, policy and operational modalities, including the use of ‘offshore processing’, exclusive jurisdiction zones in other States and, in some cases, to the application of safe third country concepts. Such third country processing is a form of externalisation that involves the post-arrival transfer of asylum seekers from the territory or jurisdiction of one State to the territory of another State for the purposes of assessing their asylum claims.[142] Often, the procedure in the third country is intended to lead to asylum in that country or in another country,[143] thus entirely excluding refugees from access to the intended State of asylum.

A frequent basis for third country processing is use of ‘safe third country’ (STC) concepts.[144] In principle, the use of STC notions need not involve externalisation, since it is merely a device that States use to justify removing putative refugees from their territory. However, in practice, some STC procedures incorporate an agreement by the receiving State to process the asylum claims, effectively allowing the expelling State to externalise its asylum obligations.[145] Since 2004, the most prominent European example of externalisation through processing based on the STC concept is the EU’s Dublin system,[146] which allocates responsibility for assessing asylum applications among EU Member States based on a hierarchical set of criteria.[147] The STC notion has also been applied in the context of cooperation with non-EU countries.[148] For example, Greek practice based on the EU-Turkey Statement is geared towards returning asylum seekers to Turkey, in exchange for financial and other assistance from the EU.[149] Outside the EU, in April 2022, the UK and Rwanda signed an Asylum Partnership Arrangement for the processing and protection of asylum seekers reaching the UK irregularly in Rwanda.[150]

In North America, a STC agreement between Canada and the US has been in operation since 2004, under which Canada deflects refugees to its neighbour in the south and vice-versa.[151] Moreover, in 2019, the US signed Asylum Cooperation Agreements (ACAs) with Guatemala, Honduras and El Salvador, making asylum applications in the US inadmissible if a person passing through those Central American countries had not yet been denied protection there.[152] An earlier example of US externalisation practice involved the transfer of Haitian and Cuban asylum seekers intercepted on the high seas to a camp in Guantanamo Bay in Cuba.[153] The US exercises exclusive jurisdiction over Guantanamo Bay under a 1903 agreement between the US and Cuba.[154] A subsequent 1934 agreement provides for indefinite US presence at Guantanamo Bay.[155] The camp last held migrants in 2017.

Since 2001, Australia has entered into bilateral agreements with Malaysia,[156] Papua New Guinea,[157] Nauru[158] and Cambodia[159] to transfer people seeking protection by sea to these countries in order to process their claims for asylum and/or receive international protection there. It provides funding to the governments to undertake processing and local integration. However, when Australia transferred asylum seekers to Papua New Guinea and Nauru for ‘processing’ in 2012, neither country had a domestic legislative framework for refugee status determination, nor any prior direct experience of processing claims.[160] On arrival, asylum seekers were detained in highly securitised, closed detention centres operated by private companies (and, in some instances, IOs) overseen by the Australian government.[161] Despite the cruel, inhuman and degrading conditions of Australia’s ‘joint processing’ arrangements,[162] as evidenced by a series of Australian government inquiries[163] and reports by UN human rights bodies,[164] the Australian model endures today and is being implemented by certain other countries.[165]

 

5.2 Legal issues

The absence of a positive international law right to be granted asylum in any particular country means that the practice of externalising elements of the asylum system through measures such as third country processing is neither explicitly authorised nor prohibited in international law. However, international law and policy does impose several important restrictions on where and how such measures can be implemented in practice.

 

5.2.1 Form and content of cooperation agreements

Third country processing agreements should meet certain formal requirements. They should be published and amenable to scrutiny by democratic and judicial mechanisms.[166] Given the rights issues posed by forcible transfers for the purpose of asylum determination (see below), agreements must take a form appropriate to protecting transferees’ rights, rather than press releases or mere political statements.[167] An agreement establishing binding legal duties for each party provides firmer guarantees for the transferees and greater legal certainty for cooperating States. It is also likely to better withstand judicial scrutiny, in view of the fact that it obliges the parties to adhere to the standards contained in the instrument.[168] For reasons of equity between sovereign States, such agreements should usually be given binding effect under international law.

The content of such agreements must not contravene either party’s existing obligations under international law. At a minimum, third party processing agreements must not permit any transfers of asylum seekers or refugees to a receiving country that would run counter to the principle of non-refoulement in refugee and human rights law.[169] This would include where a real risk of refoulement arises directly as a result of persecution or serious harm in the receiving State or adverse conditions in its asylum determination or reception system (including the risk that those conditions will force the person to refoule, i.e. ‘constructive’ refoulement), as well as any indirect risk of ‘chain’ refoulement due to inadequate determination procedures in the receiving State or its application of narrower international protection norms than in the transferring State.[170] As a minimum, the agreement must thus preserve the scope of the refugee (or international protection) definition applicable in the transferring State, although wider definitions in the receiving State would determine who else the latter would be obliged to treat as refugees.

A third country processing arrangement does not necessarily require that all States involved are parties to the Refugee Convention, although this is clearly preferable. Nonetheless, where the transferring State is a party to the Convention, then such agreements must require the receiving State to operate an asylum determination and reception system that, in law and in practice, accords with the standards set out in the Refugee Convention. This implies that the transferring State must make a good faith assessment that the receiving State will respect the applicable rights of refugees under international refugee law.[171]

Moreover, in the third country processing agreement, the transferring State should ensure that the receiving State is willing and able to abide in practice by the same human rights obligations that bind the transferring State. This follows not only from the specific nature of these forcible transfers of vulnerable persons and its legal implications,[172] but also from the proposition that a State should not be able to carry out violations of human rights standards in another State that it could not perpetrate on its own territory,[173] especially when outsourcing asylum system functions to that other State.

A third country processing arrangement must include an effective supervision mechanism to ensure the legal rights of transferred asylum seekers are respected during implementation.[174] The supervision mechanism itself should have the power to suspend the transfer of asylum seekers where substantial evidence of breaches of international obligations exists. Where the mechanism identifies evidence of breaches of international obligations, the transferring State at least must take effective measures to correct the situation or demand the return of transferred asylum seekers into its hands.[175]

 

5.2.2 Procedural standards and third country processing

A transferring State intending to remove refugees from its jurisdiction must comply, at a minimum, with the principle of non-refoulement. As a result, the transferring State must conduct individual assessments of the risk of direct or indirect refoulement before transfer. In principle, such assessments could take place extraterritorially (for example on board a vessel or aircraft). However, previous practice has shown this approach routinely to have fallen short of international law standards.[176] As a result, it is highly preferable that pre-transfer assessment procedures should be carried out within the transferring State’s territory, even if that involves temporarily transferring persons intercepted outside its territory to that territory.

As a matter of international law, for each individual, this pre-transfer procedure must assess whether any of the following elements that would render transfer unlawful are present:

  • Any real risk of direct or indirect refoulement as a result of the transfer,[177] including as a result of persecution or serious harm in the receiving State, adverse conditions in the receiving State’s reception system, or the prospect of chain refoulement as a result of unfair or ineffective procedures or the application of narrower criteria for asylum in the receiving State’s asylum determination system;[178]
  • Any extant lawful basis for the person to enter or remain in the transferring State,[179] or other legal rules which would prevent their transfer out of the territory;[180]
  • Any real risk that reception and asylum determination arrangements in the receiving State would breach international human rights standards applicable in the transferring State; or, where the transferring State is a party to the Refugee Convention, that they would breach the pertinent standards of treatment accruing to the individual under the Convention and/or other applicable international refugee law.[181] This includes the prohibitions on arbitrary detention and arbitrary restrictions on movement.[182]

Moreover, such pre-transfer assessments should also consider the following practical elements as a means to evaluate whether the individual transfer is reasonable on the facts:

  • Any risk that the person will not be admitted to the receiving country;
  • Any lack of prior connection with the receiving country, especially where the arrangement envisages the provision of asylum in the receiving State;
  • Any special needs of the person, including on the basis of their gender or other protected characteristics, and the capacity of the receiving State to meet those needs;
  • The stability of wider conditions in the receiving country, including the existence of armed conflict or generalised violence, the exposure of the country (and the location) to serious disasters and any patterns of widespread violations of human rights, especially if on a discriminatory basis pertinent to the individual concerned.

Asylum seekers subject to third country processing arrangements must be afforded access to a fair and efficient procedure in line with international standards in both the pre-transfer evaluation in the transferring State and the asylum determination process in the receiving State. At a minimum, this encompasses the right to an interview; safeguards related to the asylum procedure, including access to an interpreter; and the right to appeal to an independent decision-making body for the decision to transfer (in the pre-transfer procedure in the transferring country) and if the claim for asylum or international protection is rejected (in the asylum determination procedure in the receiving country).[183] These standards are detailed in established international refugee policy, deriving from the legal principle of effectiveness in implementing the Refugee Convention[184] and the prohibition on direct or indirect refoulement.[185] UN and regional human rights law jurisprudence confirms that these elements are required also in order to comply with human rights treaty obligations.[186]

In the absence of a functioning asylum system in the receiving State, the transferring State retains international responsibility for providing an adequate asylum procedure, flowing from its non-refoulement obligations under the Refugee Convention and international custom, as well as any relevant protection obligations established by regional treaties or custom.[187]

 

5.2.3 Third country processing arrangements involving forcible transfers

Some third country processing arrangements involve one State simply using the territory of another State to carry out its own asylum determination procedures. However, they equally often involve the forcible transfer of an asylum seeker into the hands of a third State, which takes on responsibility for processing the claim and/or providing international protection. Forcible transfer is a specific kind of administrative measure. It remains generally subject to the legal constraints applicable to expulsion, including the prohibition on refoulement and relevant due process guarantees in international refugee and human rights law.[188] However, unlike expulsion, forcible transfer involves not solely the ejection of a person from a State’s territory but also their forcible transportation to a third State that has agreed to take over certain responsibilities from the transferring State in relation to the person concerned. In third country processing contexts, these are legal responsibilities relating to the determination of the asylum claim and, in some cases, the provision of international protection.

The nature of the act of forcible transfer involved in this kind of third country processing has important implications in law. Crucially, it is well-recognised that a transferring State cannot escape its international law obligations through such transfers of responsibility. This principle is long-established in the parallel body of treaty and customary rules that govern forcible transfers under the international law of armed conflict.[189] Similarly, in the refugee law context, it is recognised that transferring an asylum seeker between States for the purpose of reassigning responsibility for processing their asylum claim and/or providing international protection is permissible only where both transferring and receiving States observe similar basic legal standards on asylum.[190] In other words, forcible transfers of the kind involved in some third country processing attract not only basic legal guarantees relevant to expulsion but they may also place limits on the extent to which a transferring State can divest itself of relevant international obligations by transferring the person into the power of another State.

 

5.2.4 Externalised asylum processing facilities

A key component of externalising asylum functions involves the setting up of asylum processing facilities in the receiving State by the transferring State, the receiving State or both. In this context, the jurisdiction of the receiving State is presumed to exist throughout its territory, even where that State exercises no control over a processing centre established by another State.[191] As such, it has international law obligations to take reasonable steps to ensure that any processing facilities operating on its territory does not violate its international obligations, including those relating to human rights and refugees.

By contrast, the transferring State will exercise extraterritorial jurisdiction for the purposes of refugee and human rights law where it exercises ‘effective control’ over the processing facilities or when it has ‘authority and control’ over the transferred asylum seekers in this context,[192] whether directly or through entities whose activities can be attributed to it under rules of State responsibility.[193] In principle, the full gamut of relevant human rights treaty rights will be applicable where the transferring State exercises jurisdiction under either head.[194] Furthermore, where a State transfers asylum seekers to a receiving country for processing of claims with knowledge that the risk of a breach of rights is a foreseeable consequence of its actions, then being a link in a causal chain that makes possible that breach of rights may well be sufficient for jurisdiction under human rights law to be established.[195]

Moreover, States cooperating in a third country processing arrangement are likely to bear shared responsibility for any breaches of international human rights and refugee law that occur in the course of such a policy. This responsibility may be established directly through the exercise of jurisdiction over processing arrangements or through indirect forms of responsibility, such as aid and assistance.[196] In certain cases, the apparent conduct of one State can be attributed to another State in the context of a third country processing arrangement. Most notably where the organ of one State is placed at the disposal of another and acts exclusively on behalf of it, then its conduct is attributed to the latter State only.[197] Similar rules will apply for attribution where IOs are involved in the creation and operation of externalised asylum processing facilities.[198]

 

5.2.5 International protection in the receiving State

In some cases, arrangements for externalisation of the asylum system extend beyond simply reception for the purpose of status determination to the ongoing provision of asylum or international protection by the receiving State in its territory for refugees and/or other persons determined to have international protection needs. In these circumstances, those persons must receive ongoing protection against refoulement. Where the transferring State is a party to the Refugee Convention, there is good reason to believe that the full set of rights contained therein must also be respected in law and practice in the receiving State, even if the latter is not a party to the Convention. Allowing a transferring State to avoid or divest itself of accrued Convention obligations by transferring the refugee to another State risks defeating the raison d’être of the treaty regime as a whole.[199] International human rights law standards applicable in the receiving State must be observed in all circumstances.[200]

International law is silent on the question of whether States are required to respect a refugee’s choice of country of asylum. Certainly, expulsion is not permitted where there is a risk of refoulement.[201] Nor may States parties to the Refugee Convention expel refugees who are ‘lawfully’ in their territory save on ‘grounds of national security or public order’.[202] Only where third country processing involves transfer prior to a refugee being ‘lawfully’ in the territory will the latter rule not apply.[203] However, even in the case, UNHCR recommends that claims by asylum seekers should ordinarily be processed by the State into whose territory or jurisdiction they arrive.[204] More generally, international refugee policy also sets out soft law criteria for identifying the country responsible for examining an asylum claim which recommend that regard be had to any ‘connection or close links’ that the asylum seeker has with a State and that, in ‘the interest of family reunification and for humanitarian reasons’, close family connections in a State should offer a basis for admission there.[205] In general, then, it would not be reasonable to deny an asylum seeker ongoing international protection in the territory of a transferring State where they have pertinent connections there, even if that State has initially transferred them elsewhere for the purposes of determining their claim.

 

5.2.6 The good faith duty and third country processing

Any State seeking to divest itself of its core asylum responsibilities under international law by forcibly transferring asylum seekers to another State for the purpose of asylum processing or provision of international protection may also be acting contrary to the good faith duty of cooperation and responsibility-sharing that underpins international refugee law.[206] This concern is particularly acute where, merely to discourage the arrival of asylum seekers in their own territories or to divest themselves of protection responsibilities, powerful States conclude third country processing agreements with third States that appear open to economic incentives or the application of diplomatic pressure. Such attempts to divest responsibility make this form of third country processing quite distinct from arrangements where a State assumes new responsibilities for asylum seekers and refugees located in a third country for good faith reasons. For example, resettlement or humanitarian evacuation schemes generally help to promote fair responsibility-sharing for refugees globally with the aim of reducing pressure on frontline States or serve to facilitate access to international protection for persons who face serious protection risks in countries of first asylum (or origin).

 

6. Accountability for externalisation measures

Externalisation measures raise the prospect that normal legal accountability mechanisms may not be applicable or accessible. Legal accountability mechanisms, in this sense, refer to the powers that courts (or quasi-judicial bodies, to the extent that they these influence the law and its development) exercise in reviewing the legality of an act and, in case of unlawfulness, annulling that act and awarding reparations.[207] It is thus necessary to review potential legal remedies (6.1) and the accountability challenges that arise in the externalisation context (6.2).

 

6.1 Remedies and accountability

In international law, the rules of responsibility speak directly to accountability concerns by telling us when a State is responsible for specific activities and the consequences that follow. Where the pertinent primary rules are silent, the ARSIWA (and ARIO) describe the content of State (and IO) responsibility, including the obligations of cessation of the internationally wrongful act, compliance with the legal obligations and non-repetition, and the duty to make reparations,[208] and they elaborate the rules governing how responsibility can be invoked and by which States (and/or IOs).[209] These inter-State rules for attributing and invoking responsibility are complemented now by a wide range of treaty-based courts and decision-making bodies dedicated to upholding specialised regimes of international law. Such forums are highly pertinent to ensuring legal accountability in the context of externalisation.

Within the UN system, the International Court of Justice provides a forum to resolve disputes (where parties to its Statute give consent to its jurisdiction) and some Charter-based mechanisms of the Human Rights Council (e.g. Universal Period Review) fulfil an accountability function, although the power to sanction is limited.[210] UN human rights treaty bodies can hear complaints of rights violations by States parties and express ‘views’ on the merits, although they are generally not considered formally binding. Regional human rights treaty bodies often exercise similar powers, and the ACtHPR, ECtHR and IACtHR are able to promulgate decisions binding on parties to the case.[211] The right of complaint to these UN and regional human rights treaty bodies is usually exercisable by consenting States parties[212] and, sometimes, by individuals who claim rights violations by States parties, although they must show that they have exhausted domestic remedies against the impugned State.[213] International courts created by regional arrangements, such as the Court of Justice of the EU, may also be relevant forums for complaints about the breach of rights protected by their constitutive treaties and fundamental rights instruments.[214]

Finally, in domestic law, national or subnational courts may provide legal accountability in relation to rights established under national law or integrated from treaties to which the State is party.[215] For example, the highest court in Papua New Guinea found the forced transfer and detention of asylum seekers transferred from Australia’s jurisdiction to Manus Island unconstitutional.[216] Some States also provide remedies based on a national legislation, especially with respect to fundamental rights included in constitutional documents, which may be relevant to externalisation contexts.

 

6.2 Challenges to accountability in the externalisation context

Depending on the geographical context and applicable legal instruments, a number of potential accountability fora may, theoretically, be competent to hear a complaint by individuals negatively affected by externalised border control or asylum measures. As a matter of fact, States have been held to some degree of accountability for externalised pushbacks[217] and for human rights breaches in the course of third country processing in a range of judicial and quasi-judicial fora.[218] Nevertheless, limited individual access to the relevant legal accountability fora, the length and cost of proceedings, and the sometimes ineffective enforcement of the judgements discourage the pursuit of accountability actions. This is especially true where survivors (i.e. foreign nationals often with an irregular status) are vulnerable, lack the means to pursue litigation, and are denied physical access to the territory where the fora of jurisdiction is located.

The multiplicity of actors that is often involved creates additional obstacles to accountability. This enhances the level of complexity not only regarding attribution of responsibility,[219] but also in identifying the relevant jurisdictions. For instance, in the case of pushbacks at sea, the various maritime zones and jurisdictions, as well as the overlap between human rights and law of the sea regimes, illustrate the challenge. In addition, not all States are bound by the same international instruments. This further complicates efforts to hold any single State accountable for its conduct. For example, in the ASEAN context, a regional human rights accountability framework and forum is lacking, and most States are not parties to the Refugee Convention and Protocol. In Europe, meanwhile, Greece, one of the States most involved in pushbacks, has not yet ratified ECHR Protocol 4, such that Article 4 – prohibiting collective expulsions – cannot be invoked before the ECtHR in this context.[220]

The involvement of non-State actors can also frustrate attempts for accountability, as States attempt to shield themselves from responsibility by acting through them. A relevant example is Malta’s practice of contracting commercial vessels to conduct pushbacks to Libya on its behalf, or to use them as floating detention centres on the high seas.[221] Moreover, the immunity of States and IOs before national courts presents real accountability challenges.[222] Thus, in Europe, the CJEU has exclusive jurisdiction over conduct of EU institutions and organs,[223] while Status Agreements concluded with third countries give immunity from civil and criminal prosecution to Frontex staff exercising their executive powers outside the EU.[224]

Finally, a lack of transparency in the modus operandi of externalisation measures further frustrates the oversight of legal accountability mechanisms, as gathering evidence, including on the identity of the actors involved, becomes particularly difficult. [225] Yet accountability must be understood as wider than just judicial scrutiny. Non-legal accountability should be promoted within democratic fora and with regional administrative authorities, including relevant Ombudsperson mandates,[226] national human rights institutions and civil society. Increasing transparency can also support other forms of non-legal accountability. Accountability, first and foremost, requires a robust system of effective independent external monitoring, which should involve democratic accountability institutions as well as regional administrative authorities, as well as other actors with relevant experience in human rights monitoring.

 

7. Conclusion

We trust that this paper provides the reader of the RLI Declaration on Externalisation and Asylum with a firm understanding of the detailed conceptual and legal analysis on which it rests. Externalisation might not be a new process in State governance, even in its impact on access to territorial asylum by aliens in need of international protection. Nonetheless, presently, it appears that processes of externalising the State functions of border control and asylum systems continue to gather momentum, driven by States and other actors particularly in the Global North but also in other parts of the world. However, whatever claims of sovereign prerogatives may be made in regulating the entry and stay of aliens to State territories and however politicised the subject matter, our analysis shows that clear international law parameters exist for such measures and must be observed if they are to be lawful in the eyes of the nations.

 

Footnotes

[1] The first four authors led the research and drafting process, with contributions from the remaining authors. The article is published in (2022) 34 International Journal of Refugee Law 120, https://doi.org/10.1093/ijrl/eeac023.

[2] Although it is not the focus of this analysis, it is important to note that domestic law in particular countries may impose additional legal constraints on externalisation.

[3] Alongside the drafting authors, the following RLI Research Affiliates contributed to discussions: Kathryn Allinson, Chloe Gilgan and Eleni Karageorgiou.

[4] The following RLI Senior Associates provided comments on earlier drafts: Cathryn Costello, Jeff Crisp, María-Teresa Gil-Bazo, Mariagiulia Giuffre, Lucy Hovil, Penelope Mathew, Violeta Moreno-Lax and Almamy Sylla. We are also grateful to RLI staff member Sarah Singer for her comments on an earlier draft of the paper.

[5] Aristide Zolberg, ‘Matters of State’ in Charles Hirschman, Philip Kasinitz, and Josh DeWind (eds), The Handbook of International Migration: The American Experience (Russell Sage Foundation 1999) 71.

[6] John Torpey, The Invention of the Passport: Surveillance, Citizenship and the State (2nd edn, CUP 2018) 12.

[7] Valsamis Mitsilegas, ‘Extraterritorial Immigration Control in the 21st Century: The Individual and the State Transformed’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control: Legal Challenges (Nijhoff 2010).

[8] Christina Boswell, ‘The ‘External Dimension’ of EU Immigration and Asylum Policy’ (2003) 79 International Affairs 619.

[9] See, for example, Asher Lazarus Hirsch, ‘The Borders beyond the Border: Australia’s Extraterritorial Migration Controls’ (2017) 36 Refugee Survey Quarterly 48; Jennifer Hyndman and Alison Mountz, ‘Another Brick in the Wall? Neo‐Refoulement and the Externalization of Asylum by Australia and Europe’ (2008) 43 Government and Opposition 249; Steven Legomsky, ‘The USA and the Caribbean Interdiction Program’ (2006) 18 International Journal of Refugee Law 677.

[10] Compare, for example, UNHCR, ‘Note on the “Externalization” of International Protection’ (28 May 2021) para 5; Bill Frelick, Ian M. Kysel and Jennifer Podkul.’”The Impact of Externalization of Migration Controls on the Rights of Asylum Seekers and Other Migrants’ (2016) 4 Journal on Migration and Human Security 190; Martin Lemberg-Pedersen, ‘Manufacturing Displacement: Externalization and Postcoloniality in European Migration Control’ (2019) 5 Global Affairs 247.

[11] Nikolas Feith Tan, ‘Conceptualising Externalisation: Still Fit for Purpose?’ (2021) 68 Forced Migration Review 8.

[12] Other contemporary examples include the externalisation of national security and defence functions.

[13] See, for example, UNHCR, ‘Note on “Externalization”’, para 5.

[14] Clearly, not all inter-State cooperation on border or asylum matters involves externalisation, which exists only where the cooperation involves one State partially or wholly shifting pertinent functions to the other.

[15] See, for example, Jeff Crisp, ‘The Cruelty of Containment’ (United Against Inhumanity, 29 April 2022) <https://www.against-inhumanity.org/2022/04/29/the-cruelty-of-containment-how-the-worlds-wealthiest-states-are-using-inhumane-and-illegal-methods-to-undermine-the-right-to-asylum/> accessed 30 April 2022.

[16] As such, we treat unilateral measures such as visa controls and carrier sanctions, as well as unilateral pushbacks on the high seas, as externalised border control measures.

[17] UN Charter (adopted 26 June 1945, entered into force 24 October 1945) 1 UNTS XVI, arts 2(7) and 2(2); Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States) (Merits) [1986] ICJ Rep 14, 106-108, paras 202-205.

[18] Convention relating to the Status of Refugees (adopted 28 Jul 1951, entered into force 22 Apr 1954) 189 UNTS 150 (Refugee Convention); Protocol relating to the Status of Refugees (adopted 31 Jan 1967, entered into force: 4 Oct 1967) 606 UNTS 267 (Protocol). In relation to Art 33, for example, a leading commentary describes ‘virtual unanimity’ on this position, except for the ‘outlier’ US case of Sale, Acting Commissioner, INS v Haitian Centers Council, 509 US 155 (1993) and possibly recent Australian practice. See Guy Goodwin-Gill, Jane McAdam and Emma Dunlop, The Refugee in International Law (4th edn, OUP 2021) 308-313. Art 40 of the Convention (the ‘territorial application clause’ or, originally, the ‘colonial clause’) differs from the jurisdiction clause found in many human rights treaties, since it concerns only dependent territories that did not have the legal personality to accede to the Convention on their own (María-Teresa Gil-Bazo, ‘Article 40’, in Andreas Zimmermann, The 1951 Convention: A Commentary (OUP 2011) 1569).

[19] Minimum Convention guarantees will be applicable to all refugees subject to a State party’s jurisdiction, with additional guarantees accruing as the refugee’s degree of attachment to that State increases along the spectrum of (i) physical presence in the territory, (ii) lawful or habitual presence, (iii) lawful stay and (iv) durable residence. See J. Hathaway, The Rights of Refugees under International Law (2nd edn, CUP 2021) 173-219.

[20] European Roma Rights Centre et al [2004] UKHL 55.

[21] See Declaration of States Parties to the 1951 Convention and/or its 1967 Protocol relating to the Status of Refugees, UN doc HCR/MMSP/2001/09 (2001) para 4, alongside analysis in Goodwin-Gill, McAdam and Dunlop, The Refugee, 300-306. For the contrary view, see Hathaway, Rights of Refugees, 435-465. Refugees who fall foul of the exception in art 33(2) of the Convention will not benefit from protection under art 33(1).

[22] Report of the United Nations High Commissioner for Refugees. Part 2, Global Compact on Refugees,

UN doc A/73/12 (Part II) (2018) (GCR). For wider discussion of responsibility-sharing duties in international refugee law, see R. Dowd and J. McAdam, ‘International Cooperation and Responsibility-Sharing to Protect Refugees: What, Why, and How?’ (2017) 66 International and Comparative Law Quarterly 863.

[23] For international armed conflict, see Common art 2 of the 1949 Geneva Conventions and art 1(4) of Protocol Additional (I) to the Geneva Conventions of 12 August 1949 and relating to the Protection of Victims of International Armed Conflicts (adopted 8 June 1977, entered into force 7 December 1978) 1125 UNTS 3 (Protocol I). For non-international armed conflicts, Common Article 3 of the 1949 Geneva Conventions applies generally and Protocol Additional (II) to the Geneva Conventions of 12 August 1949 and relating to the Protection of Victims of Non-International Armed Conflicts (adopted 8 June 1977, entered into force 7 December 1978) 1125 UNTS 609 (Protocol II) describes the threshold for the application of the additional rules contained within that protocol.

[24] See the provisions relating to non-international armed conflicts cited in the preceding footnote.

[25] Jean-Marie Henckaerts and Louise Doswald-Beck, Customary International Humanitarian Law (CUP/ICRC, 2005).

[26] See, for example, the fundamental guarantees applicable in international armed conflicts (set out in Part IV, Section III of Protocol I) and in non-international armed conflicts (set out in Common Article 3 of the 1949 Geneva Conventions and in art 4 of Protocol II), alongside the rules prohibiting the targeting of the civilian population and civilian objects (e.g. Art 48 of Protocol I, with ‘civilians’ defined in art 50).

[27] The additional guarantees relevant to these ‘protected persons’ are established in Part III of Geneva Convention relative to the Protection of Civilian Persons in Time of War (adopted 12 August 1949, entered into force 21 October 1950) 75 UNTS 287 (Geneva Convention IV).

[28] See, for example, Legal Consequences of the Construction of a Wa1l in the Occupied Palestinian Territory (Advisory Opinion) [2004] ICJ Rep 2004, para 136 and 178.

[29] United Nations Convention on the Law of the Sea (adopted 10 December 1982, entered into force, 16 November 1994) 3 UNTS 1183.

[30] Irini Papanicolupulu, ‘The Duty to Rescue at Sea, in Peacetime and in War: A General Overview’ (2016) 98 International Review of the Red Cross 491, 494.

[31] See UNCLOS, art 98(1)-(2); International Convention for the Safety of Life at Sea, (adopted 1 November 1974, entered into force 25 May 1980) 278 UNTS 1184 (SOLAS) in particular Chapter V, REG 33 (1); International Convention on Maritime Search and Rescue (adopted 27 April 1979, entered into force 22 June 1985) 118 UNTS 1405, in particular Chapter 2 (II); and International Convention on Salvage (adopted in 28 April 1989, entered into force 14 July 1996) 165 UNTS 1953.

[32] Note, however, that certain human rights treaties express a more limited field of personal application: the Convention on the Rights of the Child (adopted 20 November 1989, entered into force 2 September 1990) 1577 UNTS 3 (CRC), for example, applies only to children. Moreover, even in human rights treaties of general application, certain provisions may have a more limited personal application, such as the limitation of freedom of movement under art 12(1) of the International Covenant on Civil and Political Rights (adopted 16 December 1966, entered into force 23 March 1976) 999 UNTS 171 (ICCPR) to persons who are ‘lawfully within the territory’.

[33] Marko Milanovic, Extraterritorial Application of Human Rights Treaties: Law, Principles, and Policy (OUP 2015) 17; Anja Klug and Tim Howe, ‘The Concept of State Jurisdiction and the Applicability of the Non-refoulement Principle to Extraterritorial Interception Measures’ in Bernard Ryan and Valsamis Mitsilegas (eds), Extraterritorial Immigration Control: Legal Challenges (Martinus Nijhoff 2010) 98.

[34] Ilasçu & Ors v Moldova & Russia App no 48787/99 (ECHR, 8 July 2004) paras 312-319.

[35] This is the case even where the treaty appears to limit ‘jurisdiction’ to a State’s own territory (such as in art 2(1) ICCPR). See Human Rights Committee (HRC), General Comment No. 31, UN doc CCPR/C/21/Rev.1/Add.13 (26 May 2004) para 10; ICJ, Wall Advisory Opinion, para 111.

[36] For the American Convention on Human Rights (adopted 22 November 1969, entered into force 18 July 1978) 36 Organization of American States Treaty Series 1 (ACHR), note that the Inter-American Commission on Human Rights (IAComHR) and the Inter-American Court of Human Rights (IACtHR) have held that the test is ‘authority and control’ over a person, but this can be inferred from ‘total and exclusive de facto control’ over a location (The Environment and Human Rights, Advisory Opinion OC-23/17, IACtHR Series A No 23 (15 November 2017) paras. 81 and 93; Victor Saldaño v Argentina, IAComHR Report No. 38/99 (11 March 1999) para 21; Djamel Ameziane v US, Report No. 17/12 (20 March 2012) para 32). The IAComHR applies the same approach to the applicability of the standards set out in the American Declaration on the Rights and Duties of Man (see Rafael Ferrer-Mazorra et al v US, Report No. 51/01 (4 April 2001) paras 179-180). The African Commission on Human and Peoples Rights (AfComHR) has determined that obligations under the African Charter on Human and Peoples’ Rights (adopted 17 June 1981, entered into force 21 October 1986) 21 ILM 58 (ACHPR) exist extraterritorially on this basis too (AfComHR, General Comment No. 3 (4-18 November 2018) para 14).

[37] For the Convention against Torture (adopted 10 Dec 1984, entered into force 26 June 1987) 1465 UNTS 85 (CAT), see Committee against Torture, General Comment No. 2, UN doc CAT/C/GC/2 (24 January 2008) para 16. For the ICCPR, ICESCR and CRC, see the Wall Advisory Opinion, paras 111-113. For the European Convention for the Protection of Human Rights and Fundamental Freedoms (adopted 4 November 1950, entered into force 3 September 1953) 5 ETS 1, see Al-Skeini & Ors v UK App no 55721/07 (ECHR, 7 July 2011) paras 138-139. The same can be inferred for the ACHPR from the findings by the AfComHR re ‘effective control’ over territory by States acting extraterritorially as a basis for a range of human rights obligations in DRC v Burundi, Rwanda and Uganda, Communication No. 227/99 (May 2003) para 91. Note, however, the ECtHR has recently held that extraterritorial jurisdiction under the ECHR will not exist ‘during the active phase of the hostilities’ in armed conflict (Georgia v Russia (II) App No 38263/08 (21 January 2021) para 83)

[38] HRC, General Comment No. 31, paras 3 and 10. For the range of obligations in the CAT, extraterritorial jurisdiction arises not only due to control over territory but also where a State exercises any kind of control over ‘persons in detention’ outside its territory (CAT, General Comment No. 2, para 16).

[39] For example, a similar position might be inferred from the IAComHR findings in Djamel Ameziane.

[40] Al-Skeini, para 137.

[41] See HRC, General Comment No. 36, UN doc CCPR/C/GC/36 (30 October 2018) para 22 and AS, DI, OI and GD v Italy, UN doc CCPR/C/130/D/3042/2017 (27 January 2021) para 7.8; Committee on Economic Social and Cultural Rights, General Comment No. 24, UN doc E/C.12/GC/24 (10 August 2017) para 32; Committee on the Rights of the Child, LH et al v France, UN doc CRC/C/85/D/79/2019 (2 November 2020) para 8.5 and Sacchi et al v Argentina, UN doc CRC/C/88/D/104/2019 (8 October 2021) para 10.7.

[42] Mohammad Munaf v Romania, UN doc CCPR/C/96/D/1539/2006 (21 August 2009) para 14.2.

[43] Advisory Opinion OC-23/17, para 104(h). States are thus ‘obliged to take all necessary measures to avoid activities implemented in their territory or under their control affecting the rights of persons within or outside their territory’ (ibid, para 104(g)).

[44] AfComHR, General Comment No. 3, para 14

[45] See, for instance, Haydarie and Ors v Netherlands App no 8876/04 (ECHR, 20 October 2005); Minasyan and Semerjyan v Armenia App no 27651/05 (ECHR, 23 June 2009); Zouboulidis v Greece (no 2) App No 36963/06 (ECHR, 25 June 2009). More recent case-law confirms the idea that extraterritorial jurisdiction is created where a State exercises jurisdiction over persons abroad through procedural or legal means, for instance, in the context of judicial proceedings or a transnational arrest warrant (Romeo Castaño v Belgium App No 8351/17 (ECHR, 9 July 2017) para 42; Big Brother Watch and others v UK Apps Nos 58170/13 et al (ECHR, 13 September 2018) para 271; Güzelyurtlu and others v Cyprus and Turkey App No 36925/07 (ECHR, 29 January 2019) para 188).

[46] Issa v Turkey App no 31821/96 (ECHR, 16 November 2004). In light of the Georgia v Russia (II) decision, it may be that the ECtHR’s finding of no extraterritorial jurisdiction in Bankovic and Ors v Belgium and Ors App no 52207/99 (ECHR, 12 December 2001) can be distinguished by the fact that the impugned acts in that case took place during the phase of active conduct of hostilities in an international armed conflict.

[47] For cases where this link was not sufficiently close, see Ben El Mahi and others v Denmark App No 5853/06 (ECHR, 11 December 2006); ECtHR, Abdul Wahab Khan v UK App No 11987/11 (ECHR, 28 January 2014); MN & Ors v Belgium App No 3599/18 (ECHR, 5 May 2020).

[48] See, for example, art 4 ICCPR, art 27 ACHR and art 15 ECHR. Other treaties, such as the ACHPR, ICESCR and the law of armed conflict treaties, do not allow for derogation at all.

[49] This factual and legal context is thus wholly distinct from the potentially permissible use of derogation for extraterritorial military operations in the context of war or occupation, where the facts themselves and the existence of a lex specialis and de minimis legal framework governing the conduct of hostilities and protection of civilians might provide a justification for derogation measures. For the latter position, see Marko Milanovic, ‘Extraterritorial Derogations from Human Rights Treaties in Armed Conflict’ in Nehal Bhuta (ed), The Frontiers of Human Rights: Extraterritoriality and its Challenges, (OUP 2016).

[50] International refugee law does not provide any general legal basis for refusing to receive such applications. Art 9 of the Refugee Convention only allows a State to take essential provisional measures in time of war or other grave circumstances against a person ‘pending a determination’ by that State that the person is in fact a refugee and that the measures are necessary in the individual case.

[51] Case Concerning the Factory at Chorzów (Germany v Poland) (Merits) [1928] PCIJ Rep Series A No 17; British Claims in the Spanish Zone of Morocco (Great Britain v Spain) [1924] 2 R.I.A.A. 615.

[52] H.L.A. Hart, The Concept of Law (OUP 1961). In international law, some ‘self-contained’ regimes, such as human rights law, contain both primary obligations as well as secondary responsibility frameworks (International Law Commission (ILC), Fragmentation of International Law, UN doc A/CN.4/L.682 (13 April 2006) 68).

[53] ILC Draft Articles on State Responsibility with Commentaries, vol II (Part Two) Yearbook of the International Law Commission 2001; ILC Draft Articles on the Responsibility of International Organizations, Yearbook of the International Law Commission 2011/II (2).

[54] ARSIWA, art 1; ARIO, art 1.

[55] ARSIWA, art 2(b); ARIO, art 4(b).

[56] ARSIWA, arts 20-25; ARIO, arts 20-25. There are six circumstances where responsibility will not accrue despite an act being internationally wrongful: consent, self-defence, countermeasures, force majeure, distress and necessity. Each has a high threshold for application, and art 26 ARSIWA (and art 26 ARIO) makes it clear that none of these circumstances can be relied on if to do so would conflict with a peremptory norm of general international law. Moreover, defences cannot be used where the primary rules already take them into account (e.g. international human rights law, refugee law and the laws of war take account of necessity already).

[57] Case Concerning Rainbow Warrior Affair (New Zealand v France) [1990] UNRIAA, Vol XX, 215, 251, para 75.

[58] ARSIWA, art 3; ARIO art 5.

[59] ARSIWA, art 2; ARIO, art 4.

[60] Case Concerning Oil Platforms (Iran v USA) [2003] ICJ Rep 161, 803, para 4, emphasis added.

[61] Goodwin-Gill, McAdam and Dunlop, The Refugee, 4, emphasis added.

[62] ARSIWA, art 4. The test of ‘complete dependence’ for de facto State organs is established in Military and Paramilitary Activities in and against Nicaragua, ICJ Rep 1986, para 110. Art 6 ARIO specifies that international organisations are responsible for the conduct of their ‘organs or agents’.

[63] ARSIWA Commentary, art 4, 40.

[64] ARSIWA, art 4; Efthymios Papastavridis, ‘Rescuing Migrants at Sea and the Law of State Responsibility’ in Thomas Gammeltoft–Hansen and Jens Vedsted-Hansen (eds), Human Rights and the Dark Side of Globalisation (Routledge 2017) 172.

[65] ARSIWA, art 7; ARIO, art 8, establishes the same ‘if the organ or agent acts in an official capacity and within the overall functions of that organization’.

[66] ARSIWA Commentary, 42; French-Mexican Claims Commission, Caire Claim [1929] 5 RIAA 516.

[67] Roper (USA) v Mexico [1927] IV RIAA 145; James Pugh (Great Britain, Panama) [1933] III RIAA 1439.

[68] James Crawford, Brownlie’s Principles of Public International Law (9th edn, OUP 2019) 528.

[69] ARSIWA, art 5. No parallel provision appears in ARIO.

[70] ARSIWA, art 8; Prevention and Punishment of the Crime of Genocide (Bosnia v Serbia) [2007] ICJ Rep 2007, 43, 208, para 400. No parallel provision appears in ARIO.

[71] ARSIWA, art 6. Similarly, a parallel provision in ARIO establishes that: ‘The conduct of an organ of a State or an organ or agent of an international organization that is placed at the disposal of another international organization shall be considered under international law an act of the latter organization if the organization exercises effective control over that conduct’ (art 7).

[72] ARSIWA, art 16. The aid or assistance must be given not only with a view to facilitate the wrongful act but it must actually do so. Moreover, there is no requirement that the aid or assistance ‘should have been essential to the performance of the internationally wrongful act; it is sufficient if it contributed significantly to that act’. (ARSIWA Commentary, 66). Otherwise, neither ‘fault’ nor ‘intent’ is usually required by the secondary rules on responsibility (Crawford, Brownlie’s Principles, 538). Art 14 is the parallel provision in ARIO concerning where an international organisation aids or assists a State or another international organisation.

[73] ARSIWA Commentary, 67.

[74] James Hathaway and Thomas Gammeltoft-Hansen, ‘Non Refoulement in a World of Cooperative Deterrence’ (2015) 53 Columbia Journal of Transnational Law 235, 258. Indeed, it suffices that the State turned a blind eye to any breach of obligations, despite having access to credible information (Miles Jackson, Complicity in International Law (OUP 2015) 54).

[75] See discussion in Vincent Chetail, ‘Sovereignty and Migration in the Doctrine of the Law of Nations: An Intellectual History of Hospitality from Vitoria to Vattel’ (2016) 27 European Journal of International Law 901–922.

[76] For example, the UN Special Rapporteur on the Human Rights of Migrants (UNSRHRM) defines pushbacks as measures that result in migrants, including asylum seekers, ‘being summarily forced back, without an individual assessment of their human rights protection needs, to the country or territory, or to sea, whether it be territorial waters or international waters, from where they attempted to cross or crossed an international border’ (UN Human Rights Council (UNHRC), Report on Means to Address the Human Rights Impact of Pushbacks of Migrants on Land and at Sea: Report of the Special Rapporteur on the Human Rights of Migrants, UN doc A/HRC/47/30 (12 May 2021) para 34).

[77] UNHRC, Report of the Special Rapporteur on Torture, UN doc A/HRC/37/50 (26 February 2018) para 51.

[78] These non-entrée practices have been increasingly common since at least the 1980s (Atle Grahl Madsen, ‘Identifying the World’s Refugees’ (1983) 467 The Annals of the American Academy of Political and Social Science11, 20).

[79] See, respectively, Liliana Lyra Jubilut, Camila Sombra Muiños de Andrade and André de Lima Madureira, ‘Humanitarian Visas: Building on Brazil’s Experience’ (2016) 53 Forced Migration Review 76; and Regional Comprehensive Protection & Solutions Framework (MIRPS), II Annual Report 2019 (29 January 2020) 101.

[80] UNHRC, Report on Pushbacks, para 100.

[81] Amnesty International, Pushed to the Edge: Violence and Abuse against Refugees and Migrants along the Balkans Route (Amnesty International 2019).

[82] Amnesty International, America: Pushback Practices and their Impact on the Human Rights of Migrants and Refugees (February 2021) <https://www.amnesty.org/en/wp-content/uploads/2021/05/AMR0136582021ENGLISH.pdf> accessed 10 March 2022.

[83] See, UNHRC, Report on Pushbacks, para 57.

[84] Constantin Hruschka, ‘Hot Returns Remain Contrary to the ECHR: ND & NT before the ECHR’ (EU Immigration and Asylum Law and Policy, 28 February 2020) <https://eumigrationlawblog.eu/hot-returns-remain-contrary-to-the-echr-nd-nt-before-the-echr/> accessed 11 January 2022.

[85] See, for example, ‘Poland seals Belarus Border Crossing in Migrant Standoff’ (DW, 9 November 2021) <www.dw.com/en/poland-seals-belarus-border-crossing-in-migrant-standoff/a-59763332> accessed 11 January 2022.

[86] Ariel G. Ruiz Soto, ‘One Year after the U.S.-Mexico Agreement: Reshaping Mexico’s Migration Policies’, (Migration Policy Institute Briefing, June 2020) <www.migrationpolicy.org/research/one-year-us-mexico-agreement>; -, ‘Honduras Thwarts Attempts by Thousands of Guatemalan Migrants to Reach US’ (Statecraft, 19 January 2021) < https://www.statecraft.co.in/article/honduras-thwarts-attempts-by-thousands-of-guatemalan-migrants-to-reach-us>; both accessed 14 March 2022.

[87] Legal Centre Lesvos, Crimes Against Humanity in the Aegean (LCL 2021).

[88] Human Rights Watch (HRW), ‘Submission to the Special Rapporteur’s Report on Pushback Practices and Their Impact on the Human Rights of Migrants’ (1 February 2021) <https://www.hrw.org/news/2021/02/01/human-rights-watch-submission-special-rapporteurs-report-pushback-practices-and> accessed 11 January 2022.

[89] Stephen Legomsky, ‘The USA and the Caribbean Interdiction Program’ (2006) 18 International Journal of Refugee Law 677.

[90] European Parliament LIBE Committee, Report on the Fact-finding Investigation on Frontex concerning Alleged Fundamental Rights Violations (European Parliament 2021).

[91] Asher Lazarus Hirsch, ‘The Borders Beyond the Border: Australia’s Extraterritorial Migration Controls Get access Arrow’ (2017) 36 Refugee Survey Quarterly 48.

[92] Natasha Yacoub, Nikola Errington, Wai Wai Nu and Alexandra Robinson, ‘Rights Adrift: Sexual Violence against Rohingya Women on the Andaman Sea’ (2021) 22 Asia-Pacific Journal on Human Rights and the Law 96.

[93] Hathaway, The Rights of Refugees, 390; UNHRC, Report on Pushbacks, para 67.

[94] UNHRC, Report of the Special Rapporteur on Torture, para 54.

[95] EU Commission, ‘EU Trust Fund for Africa Adopts €46 Million Programme to Support Integrated

Migration and Border Management in Libya’ (July 28, 2017).

[96] Memorandum d’intesa sulla cooperazione nel campo dello sviluppo, del contrasto all’immigrazione illegale, al traffico di esseri umani, al contrabbando e sul rafforzamento della sicurezza delle frontiere tra Io Stato della Libia e la Repubblica Italiana. (February 2017, Renewed in 2020); Memorandum of Understanding Between the Government of National Accord of the State of Libya and The Government of The Republic of Malta in the Field of Com batting Illegal Immigration’ (28 May 2020).

[97] Hathaway, The Rights of Refugees, 390; BVMN et al, ‘Systematic Human Rights Violations’.

[98] HRW, ‘Submission to the Special Rapporteur’s Report’, para 34, 37

[99] European Parliament LIBE Committee, Report on Frontex; HRW, ‘Submission to the Special Rapporteur’s Report’, para 55, 76.

[100] See, for example, Patrick Kingsley and Haley Willis, ‘Latest Tactic to Push Migrants from Europe? A Private, Clandestine Fleet’ New York Times (New York, 30 April 2020).

[101] Amnesty International, Malta: Waves of Impunity (Amnesty International 2020) 12.

[102] See the ratio in European Roma Rights Centre et al.

[103] In the case of rejected visa applicants who fear refoulement at the hands of the other State where they are located, ECtHR jurisprudence suggests that the operation of visa controls by a destination State is unlikely to represent a causal connection sufficiently close for jurisdiction to exist under the ECHR (MN & Ors v Belgium, paras 118-124). This position may not hold true for other human rights treaties and, in relation to the ECHR, has attracted criticism from scholars (Vladislava Stoyanova, ‘M.N. and Others v Belgium: no ECHR protection from refoulement by issuing visas’ (EJIL: Talk!, 12 May 2020) <https://www.ejiltalk.org/m-n-and-others-v-belgium-no-echr-protection-from-refoulement-by-issuing-visas/> accessed 7 January 2022.

[104] Art 9, ILC Draft Articles on Expulsion of Aliens, vol II (Part Two) Yearbook of the International Law Commission 2014. See, more generally, Jean-Marie Henckaerts, Mass Expulsion in Modern International Law and Practice (Nijhoff 1995); and Lena Riemer, The Prohibition of Collective Expulsion in Public International Law (Berlin, 2020).

[105] The prohibition of collective expulsion can be found, inter alia, in art 13 ICCPR, art 4 Protocol 4 to the ECHR, art 12 (5) ACHPR, art 22(9) ACHR, art 26(2) ArCHR, and art 22 International Convention on the Protection of the Rights of All Migrant Workers and Members of their Families (adopted 18 December 1990, entered into force 1 July 2003) 2220 UNTS 3 (ICRMW). Note that art 13 ICCPR is limited to aliens who are lawfully in the territory.

[106] HRC General Comment No 15 (1986) para 10.

[107] UNHRC, Report of the Special Rapporteur on Torture, para 34. Recent case law from the ECtHR reading in purported exceptions to this rule based on ‘fault’ in the behaviour of the asylum seekers is inconsistent as to relevant factors and, in some cases, risks rendering the prohibition meaningless (compare AA et al v North Macedonia App Nos 55808/16 et al (ECHR, 5 April 2022), Shahzad v Hungary App No 12625/17 (ECHR, 8 July 2021), and ND and NT v Spain App Nos 8675/15 and 8697/15 (ECHR, 13 February 2020)).

[108] See, particularly, art 33(1) Refugee Convention and art II(3) OAURC. Among many regional soft law standards, see also art III Asian-African Legal Consultative Organization, Bangkok Principles on the Status and Treatment of Refugees (adopted 24 June 2001) (Bangkok Principles).

[109] The prohibition on refoulement is articulated by international human rights law in art 3 of the CAT, art 22(8) of the ACHR and art 16 of the International Convention for the Protection of All Persons from Enforced Disappearance (adopted 20 December 2006, entered into force 23 December 2010) 2716 UNTS 3; and it can be inferred from provisions that articulate a general prohibition on torture, inhuman and degrading treatment, such as art 7 ICCPR and art 3 ECHR (see, for example, Chitat Ng v Canada UN doc CCPR/C/49/D/469/1991 (7 January 1994) and Soering v UK App No 14038 (ECHR, 7 July 1989)). The laws of war protect against refoulement of civilian protected persons, e.g. art 45, para 4, GCIV.

[110] Art 12(2) ICCPR, Art 5(ii) ICERD; art 10(2) CRC; art 8(1) ICMW; art 2(2) ECHR Protocol No. 4; Art 22(2) ACHR; art 12(2) ACHPR; art 27(1) Arab Charter on Human Rights (revised version adopted 23 May 2004, entered into force 15 March 2008) (2006) 24 Boston University International Law Journal 147 (ArCHR). This right is subject to limitations. It is also expressed by soft law, including art 13(2) of the Universal Declaration of Human Rights (adopted 10 December 1948) UNGA Res 217A(III) (UDHR) and art 15 ASEAN, Human Rights Declaration (adopted 19 November 2012) (ASEAN HRD). In international armed conflict, GCIV protects this right for aliens in the territory of a party to the conflict (art 35) and for protected persons who are not nationals of the State whose territory is occupied (art 48).

[111] See, for example, Article 18 EUCFHR, Art 27(1) ArCHR. The right to seek and be granted asylum in accordance with international law is protected in art 22(7) ACHR and art 12(3) ACHPR; and the right to seek and enjoy asylum is reflected in soft law in art 14 UDHR, art XXVII ADHR, art 16 ASEAN HRD and Art II(1) Bangkok Principles. For a broader discussion of the impact of pushbacks on ‘access to protection’ as an overarching concept, see M. Giuffré, The Readmission of Asylum Seekers under International Law (Bloomsbury 2020) 35-128.

[112] Art 7 ICCPR; art 6 and 37(a) CRC; art 3 ECHR; art 5(2) ACHR; art 5 ACHPR; art 13(a) ArCHR; and in soft law by art 5 UDHR, art I ADHR and art 14 ASEAN HRD. In armed conflict: common art 3, 1949 Geneva Conventions, art 75(2)(a)-(b) Protocol I and art 4(2)(a) and (e) Protocol II.

[113] Art 2(3) ICCPR; art 13 ECHR; art 10 ACHR; arts 9 and 16 ACHPR; art 12 and 23 ArCHR; and in soft law by art 8 UDHR and art 5 ASEAN HRD

[114] Art 6 ICCPR; art 6 CRC; art 2 ECHR; art 4 ACHR; art 4 ACHPR; art 5 ArCHR; and in soft law by art 3 UDHR, art I ADHR and, art 11 ASEAN HRD. In armed conflict: common art 3, 1949 Geneva Conventions, art 75(2)(a) Protocol I and art 4(2)(a) Protocol II.

[115] This act is prohibited under broader human rights protection of the right to life and specifically through treaties such as the International Convention for the Protection of All Persons from Enforced Disappearance (adopted 20 December 2006, entered into force 23 December 2010) 2716 UNTS 3.

[116] Pushbacks from within a State’s own land or sea territory might also be considered externalised in a limited set of circumstances, such as where they are carried out on the basis of agreement by another State to receive the people summarily pushed back as part of the border control arrangements of the first State.

[117] For example, under arts 19(2) and 21 1(h) UNCLOS, a foreign vessel that has compromised its right of innocent passage (a passage not prejudicial to the peace, good order and security of the coastal State) by intending to load and unload persons contrary to the immigration laws of the coastal State may be subject to interception by the coastal State. However, some scholars argue that a vessel carrying people who intend to seek asylum cannot be equated with ‘loading and unloading’ of people contrary to immigration laws. Violeta Moreno-Lax ‘Protection at Sea and the Denial of Asylum’ in Cathryn Costello et al (eds) The Oxford Handbook of International Refugee Law (OUP 2021) 489; Mark Pallis, ‘Obligations of States towards Asylum Seekers at Sea: Interaction and Conflicts Between Legal Regimes’ (2002) 14 International Journal of Refugee Law 329, 357. In any event, such interceptions need to comply with human rights and refugee law obligations.

[118] Nora Makard, ‘The Right to Leave by Sea: Legal Limits on EU Migration Control by Third Countries’ (2016) 27 European Journal of International Law 591, 600; Guy Goodwin-Gill, ‘The Right to Seek Asylum: Interception at Sea and the Principle of Non-Refoulement’, (2011) 23 International Journal of Refugee Law, 443–457.

[119] Subject to the right of innocent passage in UNCLOS, art 17.

[120] UNCLOS, art 33. See Maria Gavouneli, Functional Jurisdiction in the Law of the Sea (Nijhoff 2007).

[121] UNCLOS art 92-111. See Efthimios Papastavridis, The Interception of Vessels on the High Seas, Contemporary Challenges to the Legal Order of the Oceans (Hart 2013).

[122] UNCLOS, 98(1), 1974 SOLAS Convention Chapter V, REG 33 (1), 1979 SAR Convention Chapter 2 (II).

[123] SOLAS Convention Chapter V, REG 33 (1)

[124] IMO Circular 194/2009 Principles Relating to Administrative Procedures for Disembarking People Rescued at Sea, para 3. Although a ship may serve as a temporary place of safety, survivors of rescue cannot be held on board a ship or a rescue unit indefinitely (IMO Resolution Guidelines on the Treatment of Persons Rescued at Sea, MSC.167(78), adopted on 20 May 2004 para 6.14; UNHCR, Background on the Protection of Asylum Seekers and Refugees Rescued at Sea, (18 March 2002) para 12).

[125] IMO Circular 194/2009 ‘Principles Relating to Administrative Procedures for Disembarking People Rescued at Sea’, at para 3. Some scholars take the view that there is a legal obligation on the part of the coastal state in whose SAR zone the rescue takes place to allow for disembarkation (see Violeta Moreno Lax, ‘Seeking Asylum in the Mediterranean: Against a Fragmentary Reading of EU Member States’ Obligations Accruing at Sea’ (2011) 23 International Journal of Refugee Law 174, 198-200).

[126] See text accompanying notes 71-74 above.

[127] Any ‘instigation, support or participation’ by a State in delegating its border governance obligations ‘may give rise to complicity in or joint responsibility for unlawful pullback operations and the resulting human rights violations, including torture and ill-treatment’ (UNHRC, Report of the Special Rapporteur on Torture, para 57).

[128] See Hirsi Jamaa and Ors v Italy, App No 27765/09 (ECHR, 23 February 2012) – the word “expulsion” should be interpreted “in the generic meaning, in current use (to drive away from a place)” and thus applies to pushbacks. This approach is confirmed in ND & NT v Spain App Nos 8675/15 and 8697/15 (ECHR, 13 February 2020).

[129] By analogy with the findings in the Nicaragua case (para 115), a State’s participation, ‘even if preponderant or decisive’, in the financing, organizing, training, supplying and equipping of the non-State actor, or the selection of its targets, or the ‘planning of the whole of its operation’, is still insufficient in itself to attribute responsibility to the State.

[130] Art 14, ARIO. However not all non-State actors will fulfil the definition of an international organisation and thus allow the application of the rules set out in ARIO.

[131] Giorgos Christides et al, ‘EU Border Agency Frontex Complicit in Greek Refugee Pushback Campaign’ Der Spiegel (Hamburg, 23 October 2020).

[132] Mariana Gkliati, ‘The Next Phase of the European Border and Coast Guard: Responsibility for Returns and Pushbacks in Hungary and Greece’ (2022) 7 European Papers 171.

[133] Mariagiulia Giuffré, ‘State Responsibility Beyond Borders: What Legal Basis for Italy’s Push-backs to Libya?’ (2012) 24 International Journal of Refugee Law 692. For readmission arrangements between Vietnam and Australia in the context of interception at sea, see Ben Doherty, ‘Vietnamese Asylum Seekers Forcibly Returned by Australia Face Jail’ The Guardian (23 May 2016).

[134] Neither can internal regulations, such as the Frontex Sea Operations Regulation, which must be applied without prejudice to binding international legal obligations.

[135] In ND & NT v Spain, the ECtHR argued that pushing back migrants who had managed to scale the Melilla border fence did not amount to a collective expulsion due to the migrants’ own ‘culpable conduct.’ It found that the applicants had ‘deliberately take[n] advantage of their large numbers and use[d] force, […] such as to create a clearly disruptive situation which is difficult to control and endangers public safety’. The court decided that ‘the lack of individual removal decisions can be attributed to the fact that the applicants, if they indeed wished to assert rights under the Convention, did not make use of the official entry procedures existing for that purpose, and was thus a consequence of their own conduct.’ (paras 201, 231).

[136] See text accompanying notes 48-50 above.

[137] Volker Türk and Madeline Garlick, ‘From Burdens and Responsibilities to Opportunities: The Comprehensive Refugee Response Framework and a Global Compact on Refugees’ (2016) 28 International Journal of Refugee Law 656–678, 657.

[138] This includes the guarantees in arts 3, 13, 16(1), 20, 22, 29, 33 and 34 of the Convention.

[139] According to GCIV, civilian protected persons ‘may be transferred by the Detaining Power only to a Power which is a party to the present Convention and after the Detaining Power has satisfied itself of the willingness and ability of such transferee Power to apply the present Convention’ (art 45, para three); transfers out of occupied territory are prohibited (art 49, para one).

[140] Noll, Fagerlund and Liebaut, Study on Feasibility.

[141] See, for example, Arthur Helton, ‘Refugee Determination under the Comprehensive Plan of Action: Overview and Assessment’ (1993) 5 International Journal of Refugee Law 544.

[142] As such, asylum seekers need not have arrived in the territory of the destination State, merely under its jurisdiction (i.e. including where exercised extraterritorially).

[143] For example, Australia tried to send refugees from Nauru to Cambodia and the USA (Memorandum of Understanding between the Government of the Kingdom of Cambodia and the Government of Australia, Relating to the Settlement of Refugees in Cambodia (26 September 2014); Australia–United States Resettlement Arrangement (2021) .

[144] On STC concepts in general, see Eleni Karageorgiou, Luisa Feline Freier and Kate Ogg, ‘The Evolution of Safe Third Country Law and Practice’ in Cathryn Costello et al (eds) The Oxford Handbook of International Refugee Law (OUP 2021).

[145] Other STC arrangements (such as readmission agreements) that involve no corresponding duty on the part of the receiving State to process asylum claims are not considered as third country processing and fall outside the scope of this section.

[146] Note that the Dublin system has also been categorized as a ‘first country of arrival’ system (e.g. Hathaway, Rights of Refugees, 330).

[147] The core of this framework is the so-called Dublin Regulation (Regulation (EU) No 604/2013 of the European Parliament and of the Council of 26 June 2013 establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person).

[148] In 2021, Denmark changed its law to allow for the transfer of asylum seekers outside the EU for the purposes of both asylum processing and protection of refugees in a third country (L 226 Forslag til lov om ændring af udlændingeloven og hjemrejseloven (Indførelse af mulighed for overførsel af asylansøgere til asylsagsbehandling og eventuel efterfølgende beskyttelse i tredjelande), Lovforslag som vedtaget, 3 June 2021).

[149] European Commission, ‘EU–Turkey Statement’ (18 March 2016).

[150] Memorandum of Understanding between the Government of the United Kingdom of Great Britain and Northern Ireland and the Government of the Republic of Rwanda for the Provision of an Asylum Partnership Arrangement (13 April 2022).

[151] Agreement between the Government of Canada and the Government of the United States of America For Cooperation in the Examination of Refugee Status Claims from Nationals of Third Countries (signed 5 December 2002, effective 29 December 2004).

[152] See the US Immigration and Nationality Act 1965, § 1158. Only the Guatemala ACA went into operation.

[153] Daniel Ghezelbash, Refuge Lost: Asylum Law in an Interdependent World (CUP 2018) 75-76; Azadeh Dastyari, United States Migrant Interdiction and the Detention of Refugees in Guantánamo Bay (CUP 2015) 47.

[154] According to the agreement, the US ‘shall exercise complete jurisdiction and control over and within said areas’, whilst recognising ‘the ultimate sovereignty of the Republic of Cuba over the above described areas of land and water’ (Agreement between the United States and Cuba for the Lease of Lands for Coaling and Naval stations (23 February 1903) arts 1 and 3).

[155] Treaty between the United States of America and Cuba (29 May 1934) art 3.

[156] Australia-Malaysia Asylum Seeker Transfer Agreement (27 July 2011). The Agreement was struck down by the Australian High Court and never entered into operation.

[157] MOU between the Government of the Independent State of Papua New Guinea and the Government of Australia, relating to the transfer to and assessment of persons in PNG, and related issues’ (8 September 2012).

[158] MOU between the Republic of Nauru and the Commonwealth of Australia, relating to the transfer to and assessment of persons in Nauru, and related issues (3 August 2013).

[159] UNHCR, ‘Statement on Australia-Cambodia Agreement on Refugee Relocation’ (26 September 2014) <https://www.unhcr.org/en-au/news/press/2014/9/542526db9/unhcr-statement-australia-cambodia-agreement-refugee-relocation.html> accessed 11 January 2022.

[160] UNHCR, Mission to the Republic of Nauru (14 December 2012) <https://www.refworld.org/docid/50cb24912.html>; UNHCR Mission to Manus Island, Papua New Guinea 15-17 January 2013 (4 February 2013) <http://www.refworld.org/docid/5139ab872.html>; both accessed on 11 January 2022.

[161] Madeline Gleeson and Natasha Yacoub, ‘Cruel, Costly and Ineffective: the Failure of Offshore Processing in Australia’, Kaldor Centre Policy Brief No. 11 (2021).

[162] UNHRC, Report of the Special Rapporteur on the Human Rights of Migrants on his Mission to Australia and the Regional Processing Centres in Nauru, UN doc A/HRC/35/25/Add.3 (24 April 2017).

[163] Senate Legal and Constitutional Affairs References Committee, ‘Incident at the Manus Island Detention Centre from 16 February to 18 February 2014’ (December 2014); Australian Human Rights Commission, ‘The Forgotten Children: National Inquiry into Children in Immigration Detention’ (2014); Senate Select Committee on the recent allegations relating to conditions and circumstances at the Regional Processing Centre in Nauru, ‘Taking Responsibility: Conditions and Circumstances at Australia’s Regional Processing Centre in Nauru’ (31 August 2015); Senate Legal and Constitutional Affairs References Committee, ‘Serious Allegations of Abuse, Self-Harm and Neglect of Asylum Seekers in Relation to the Nauru Regional Processing Centre, and Any Like Allegations in Relation to the Manus Regional Processing Centre’ (April 2017).

[164] The UN reports focus on breaches of rights in the ‘regional processing centres’ in Nauru and on Manus Island and the possibility of holding Australia accountable (see, for example, UNHRC, Report of the Special Rapporteur on the Human Rights of Migrants on his Mission to Australia, paras 72–84; HRC, ‘Concluding Observations on the Sixth Periodic Report of Australia’, UN doc CCPR/C/AUS/CO/6 (1 December 2017) paras 35–36).

[165] See, for example, the UK Nationality and Borders Act 2022, s 29.

[166] Guy Goodwin-Gill, ‘The Right to Seek Asylum, Interception at Sea and the Principle of Non-Refoulement’ (2011) 23 International Journal of Refugee Law 443.

[167] For example, the ECtHR is clear that diplomatic assurances, even given in good faith, cannot be relied upon per se but must rather take a form and content adequate to protect the transferee’s ECHR rights (see, for example, Othman (Abu Qatada) v UK App No 8139/09 (ECHR, 17 January 2012).

[168] Plaintiff M70/2011 v Minister for Immigration and Citizenship para 60.

[169] See notes 108-109 above.

[170] See, for example, R v SSHD, ex parte Adan, R v SSHD, ex parte Aitseguer [2000] UKHL 67 and TI v UK App No 43844/98 (ECHR, 7 March 2000).

[171] For example, an MOU between Australia (a State party to the Refugee Convention) and Malaysia (not a party to the Convention) was struck down by the High Court of Australia in 2011 because the relevant rights of individuals transferred by Australia could not be assured in Malaysian domestic law or in practice (Plaintiff M70/2011 v Minister for Immigration and Citizenship [2011] HCA 32, paras 125–126). See also James Hathaway, ‘The Michigan Guidelines on Protection Elsewhere’ (2007) 28 Michigan Journal of International Law 207, para 3. It would appear that contingent rights under the Refugee Convention are pegged to national law standards in the receiving State rather than the transferring State.

[172] See section 5.2.3 below.

[173] See text accompanying notes 41-47 above.

[174] For instance, in Othman v UK, the ECtHR points to the critical importance of an independent monitoring mechanism as a means to ensure human rights treaty standards are not violated in such contexts.

[175] See, mutatis mutandis, the provisions for the transfer of civilian protected persons in paragraph 2, art 45 GCIV, which require that the transferring State must ‘take effective measures to correct the situation or…. request the return’ of the persons transferred if the receiving State fails to abide by treaty standards relating to their treatment.

[176] Azadeh Dastyari and Daniel Ghezelbash, ‘Asylum at Sea: The Legality of Shipboard Refugee Status Determination Procedures’ (2020) 32 International Journal of Refugee Law 1.

[177] Refoulement is contrary to international refugee and human rights law (see notes 108-109 above.).

[178] See text accompanying notes 169-170 above.

[179] It will be important to check that those selected for transfer do not in fact hold nationality of the transferring State (and thus cannot ‘arbitrarily deprived of the right to enter his own country’, e.g. art 12(4) ICCPR) or a right (acquired or otherwise) under domestic law to enter or remain in the country. This will be especially important where persons travelling (irregularly) to the transferring State are intercepted outside its territory en route, as suggested by the attempts to reach the UK using whatever means possible on the part of some British nationals and residents who found themselves in Afghanistan at the time of the 2021 Taliban takeover.

[180] For example, where the transfer would amount to a disproportionate interference with rights to family and private life protected under international human rights law (see, for example, on art 8 ECHR, Osman v Denmark App No 38058/09 (ECtHR, 14 June 2011)).

[181] See section 5.2.3 below.

[182] Art 26 Refugee Convention; arts 9 and 12(1) ICCPR; art 37(b) CRC; arts 6 and 12(1) ACHPR; arts 7 and 22(1) ACHR; arts 14 and 26(1) ArCHR; art 5 ECHR and art 2(1) ECHR Protocol 4; and in soft law arts 9 and 13 UDHR, arts XXV and VIII ADHR, and arts 12 and 15 ASEAN HRD.

[183] The jurisprudence of human rights treaty bodies suggests that this is a binding requirement under ICCPR (HRC, Concluding Observations on Estonia, UN doc CCPR/CO/77/EST (31 March 2003) para 13), ACHPR (ACommHPR, Kenneth Good v Botswana (2010), Communication No. 313/05, para 169) and ACHR (IACtHR, Pacheco Tineo Family v Bolivia, IACtHR Series C No 272 (25 November 2013) paras. 159(e)–(f)), and ECHR (Vilvarajah & Ors v UK App Nos 13163/87 et al (30 October 1991) para 126).

[184] Goodwin-Gill and McAdam, The Refugee, 529–530; UNHCR, Handbook on Procedures and Criteria for Determining Refugee Status, UN doc HCR/IP/4/Eng/REV.1 (1979, reissued 2019) para 189; UNHCR Executive Committee Conclusion No. 21 (1981) para d, referring to “effective implementation”.

[185] See notes 108-109 above.

[186] D.J. Cantor, ‘Reframing Relationships: Revisiting the Procedural Standards for Refugee Status Determination in light of Recent Human Rights Treaty Body Jurisprudence’ (2015) 34 Refugee Survey Quarterly 79.

[187] See text accompanying notes 183-186 above. The human rights basis for this is also articulated, inter alia, in MSS v Belgium and Greece App No 30696/09 (ECHR, 21 January 2011) paras 358-9, and Ilias & Ahmed v Hungary App No 47287/15 (ECtHR, 14 March 2017) paras 128-38.

[188] On the non-refoulement principle, see notes 108-109 above. Due process guarantees in the expulsion context exist for refugees ‘lawfully in the territory’ (art 32 Refugee Convention and, as a soft law example, art V(4) Bangkok Principles) and, under human rights law, for aliens ‘lawfully in the territory’ (art 13 ICCPR, art 22(6) ACHR, art 12(4) ACHPR, art 1, ECHR Protocol 7, art 26(2) ArCHR).

[189] See the provisions relating to the forcible transfer of civilian protected persons in paragraph 2, art 45 GCIV, which require that the transferring State must satisfy itself of the ‘willingness and ability’ of the receiving State to apply that treaty and the legal rules contained therein.

[190] For example, arrangements such as the Dublin system of transfers between contracting EU States clearly permit these transfers only where both States are bound by the same common standards on asylum (see, for example, Regulation 604/2013 establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person (recast) [2013] OJ L180 (Dublin III)). In a different regional context, this same principle was the basis for the Australian High Court striking down the transfer agreement between Australia and Malaysia (Michelle Foster, ‘The Implications of the Failed Malaysian Solution: The Australian High Court and Refugee Responsibility Sharing at International Law’ (2012) 13 Melbourne Journal of International Law 395).

[191] See, for example, Ilasçu & Ors v Moldova & Russia.

[192] See text accompanying notes 36-40 above. Note that Australia’s ‘significant levels of control and influence exercised…over the operation of the offshore regional processing centres, including over their establishment, funding and service provided therein amount to effective control’ such as to trigger jurisdiction (HRC, ‘Concluding Observations on the Sixth Periodic Report of Australia’, paras 35–36; see also CAT, ‘Concluding Observations on the Fourth and Fifth Periodic Reports of Australia’, UN doc CAT/C/AUS/CO/4-5 (23 December 2014)).

[193] See text accompanying notes 68-74 above.

[194] However, under the ECHR, where extraterritorial jurisdiction is exercised only over the person rather than the location, then the range of rights might be tailored to the specific situation (Al-Skeini, para 137).

[195] See text accompanying notes 41-47 above.

[196] See text accompanying notes 41-47 and 71-74 above.

[197] ARSIWA, art 6. See the parallel situation regarding IOs in art 7 ARIO.

[198] See notes 71-74 above.

[199] See Hathaway, ‘Michigan Guidelines on Protection Elsewhere’. In this regard, the ratification by a receiving State of treaties such as the Refugee Convention is at best indicative, and not determinative, of implementation (UNHCR, ‘Legal Considerations regarding access to protection and a connection between the refugee and the third country in the context of return or transfer to safe third countries’ (April 2018) <https://www.refworld.org/docid/5acb33ad4.html> accessed 13 March 2022, 2).

[200] For these purposes, a State is normally presumed to exercise jurisdiction throughout its own territory (Ilasçu, para 312).

[201] See notes 108-109 above.

[202] Art 32(1) Refugee Convention. Art 32(2)-(3) also requires observing due process in any expulsion and giving reasonable time to the refugee to seek admission to another country. In soft law, similar standards are articulated in art V Bangkok Principles.

[203] There is strong albeit not unanimous affirmation that ‘lawful’ presence is triggered by the application for asylum and persists until that claim is rejected (see, for example, Hathaway, Rights of Refugees, 196-212).

[204] UNHCR, Guidance Note on Bilateral and/or Multilateral Transfer Arrangements of Asylum-seekers (May 2013) para 1.

[205] UNHCR Executive Committee Conclusion No 15 (1979), paras (h)(iv) and (e). Similarly, in EU secondary law, the Asylum Procedures Directive (art 38(2)(a) requires, in the application of STC rules, that regard be had to a ‘connection’ with a country that would make it ‘reasonable’ for the person to go there.

[206] See text accompanying note 22 above. For an analysis of the bad faith interpretation of Australia’s international obligations through its offshore processing policy, see Penelope Mathew, ‘Australian Refugee Protection in the Wake of the Tampa’ (2002) 96 American Journal of International Law 661, 673.

[207] D. Oliver, Government in the United Kingdom: The Search for Accountability, Effectiveness and Citizenship (OUP 1991) 26.

[208] ARSIWA, arts 29-31, 34-37; ARIO, arts 29-31, 34-37. The general international law concept of reparation, and its constituent components of restitution, compensation and satisfaction, have both tended to be read into the concept of remedies under specialised human rights treaty regimes. See, for example, 2005 UN Basic Principles and Guidelines on the Right to a Remedy and Reparation for Victims of Gross Violations of International Human Rights Law and Serious Violations of International Humanitarian Law – UNGA res 60/147 (21 March 2006).

[209] ARSIWA, arts 42–48; ARIO, arts 43-50.

[210] Externalisation practices might also be prosecuted as crimes against humanity before the International Criminal Court or UN ad hoc tribunals (as well as in domestic courts). Rome Statute of the International Criminal Court (adopted 17 July 1998, entered into force 1 July 2002) 2187 UNTS 3 arts 7(2)(d), 7(1)(d); see also Itamar Mann, ‘Border Crimes as Crimes Against Humanity’ in Cathryn Costello, Michelle Foster and Jane McAdam (eds), Oxford Handbook of International Refugee Law (OUP 2021).

[211] ACHPR Protocol on the Establishment of an African Court on Human and Peoples’ Rights (adopted 10 June 1998, entered into force 25 January 2004) Art 30; ECHR, art 46; ACHR, art 68.

[212] See, for example, ICCPR, art 41; ACHPR, art 47; ACHPR Protocol, art 5; ACHR, art 45; ECHR, art 33.

[213] See, for example, ICCPR Optional Protocol, arts 1-2; ACHPR, art 55; ACHR, art 44, 46; ECHR, arts 34-35. Under ECHR art 35(1), State complainants are equally required to show exhaustion of domestic remedies.

[214] Jasmine Coppens, Interception of Migrant Boats at Sea in Violeta Moreno-Lax and Efthimios Papastavridis (eds), Boat Refugees and Migrants at Sea: A Comprehensive Approach: Integrating Maritime Security with Human Rights (Brill 2016) 199.

[215] See, for example, the case before the Supreme Court of Victoria (Australia), Kamasaee v Commonwealth of Australia & Ors (Approval of settlement) [2017] VSC 537 (6 September 2017).

[216] Supreme Court of Papua New Guinea, Namah v Pato [2016] PGSC 13, SC1497 (26 April 2016).

[217] ECtHR jurisprudence includes Hirsi Jamaa and Ors v Italy, as well as MK & Ors v Poland, App Nos 40503/17 et al (ECHR, 23 July 2020), DA & Ors v Poland, App No 51246/17 (ECHR, 8 July 2021), Shahzad v Hungary App No 12625/17 (ECHR 8 July 2021), MH & Ors v Croatia App Nos 15670/18 and 43115/18 (ECHR, 18 November 2021). See also C‑808/18 Commission v Hungary (CJEU, 17 December 2020).

[218] The Haitian Centre for Human Rights et al v US, IAComHR Report No. 51/96 (13 March 1997); Namah v Pato (Minister for Foreign Affairs and Immigrations) and Ors (2016) PJSC 13; Kamasaee v. Commonwealth, [2017] SCI 2014 06770.

[219] Mariana Gkliati and Jane Kilpatrick, ‘Crying Wolf Too Many Times: The Impact of the Emergency Narrative on Transparency in Frontex Joint Operations’ (2022) 17 Utrecht Law Review 57, 64.

[220] However, Greece is bound by the equivalent art 19 of the EU Charter for Fundamental Rights.

[221] See text accompanying notes 100-101 above; also Council of Europe Commissioner for Human Rights, ‘Commissioner urges Malta to meet its obligations to save lives at sea’ (11 May 2020) <https://www.coe.int/en/web/commissioner/-/commissioner-urges-malta-to-meet-its-obligations-to-save-lives-at-sea-ensure- prompt-and-safe-disembarkation-and-investigate-allegations-of-delay-or-no> accessed 12 April 2022.

[222] See Stichting Mothers of Srebrenica and Ors v Netherlands and UN, Final appeal judgment, LJN: BW1999, ILDC 1760 (NL 2012)

[223] Art. 268 TEU; Asteris and Others v Greece and EEC; CJEU 14 January 1987, C-281/84, ECLI:EU:C:1987:3 (Zuckerfabrik Bedburg v Council and Commission); CJEU 29 July 2010, C-377/09, ECLI:EU:C:2010:459 (Hanssens-Ensch v European Community).

[224] Mariana Gkliati and Jane Kilpatrick, ‘Losing Sight of an Agency in the Spotlight: Frontex Cooperation with Third Countries and their Human Rights Implications’ (2021) 68 Forced Migration Review 18.

[225] Hungarian Helsinki Committee, ‘M.H. and Others v. Croatia: Third party intervention on behalf of the Hungarian Helsinki Committee’ (8 October 2018) <https://helsinki.hu/wp-content/uploads/3rd-party-intervention-in-MH-v-Croatia.pdf> accessed 10 January 2022. Individuals may record pushbacks on their phones, such evidence may be lost where these individuals’ property is confiscated or destroyed. See Border Violence Monitoring Network, ‘”Unverifiable Information from Unknown Migrants”? – First Footage of Push-Backs on the Croatian-Bosnian Border’ (16 December 2018) <www.borderviolence.eu/proof-of-push-backs/> accessed 11 January 2022.

[226] For example the Greek Ombudsman, in investigating alleged pushbacks from Greece to Turkey, asked the Greek police to provide relevant information on their involvement, but was met with a categorical denial of pushback allegations (Greek Ombudsman, ‘Own Initiative Investigation: Alleged pushbacks to Turkey of foreign nationals who had arrived in Greece seeking international protection’ (Statewatch, 31 December 2021) <www.statewatch.org/media/2325/gr-ombudsman-pushbacks-interim-report-4-21.pdf> accessed 1 January 2022).