Compulsory Secondary Movement and Article 32 of the Refugee Convention

August 28, 2019
Yuval Livnat
Tel Aviv University, Faculty of Law, Refugee Rights Program


“Safe third countries” policies (better termed compulsory secondary movement policies) have been gaining popularity since the 1990s. Are they in compliance with Article 32 to the Convention relating to the Status of Refugees (“the Convention”), which sets limits on contracting states’ power to deport refugees to non-persecutory countries? Oddly enough, there is little academic and judicial discussion of this legal question. As one legal scholar put it: “Article 32 of the 1951 Convention is a forgotten article.”[1] In this article, I will argue that Article 32 is highly relevant to the legal review of compulsory secondary movement policies. I will further argue that Article 32 should be interpreted and applied by resorting to the (subsidiary) “object and purpose”[2] of the Convention, which is restoring refugees’ stability in life. In addition, I will also discuss the subsidiary purpose of the Convention to promote cooperation among states in order to fairly allocate the burden of refugees’ protection.

Article 32

Article 32, labeled “expulsion,” provides that “[t]he contracting states shall not expel a refugee lawfully in their territory, save on grounds of national security or public order.” It stands alongside Article 33, which articulates the renowned right of refugees to non-refoulment. Article 33 is broader than Article 32 in the sense that it applies to any refugee, while Article 32 applies only to “lawfully present” refugees. In a different sense, however, Article 32 is broader than Article 33: while the latter prohibits a contracting state from deporting a refugee “to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion,” the former prohibits a contracting state from deporting a (legally present) refugee to any country whatsoever.

Compulsory secondary movement policies are regularly designed to deport asylum seekers to non-persecutory “third” countries. These policies further aim to refrain from indirect refoulment (i.e. the policies endeavor to ensure that the “third” country would not, in turn, deport the asylum seekers to their country of origin or any other country where they might face persecution). Hence, Article 32 is highly relevant for evaluating the legality of such policies and plans. Surprisingly, neither Western countries, which espouse such policies, nor the United Nations High Commissioner for Refugees (“UNHCR”), pay much attention to Article 32.

“Lawfully present” refugees

One possible explanation for the apparent oversight of Article 32 is the vagueness of the “lawful presence” requirement. As a result of this vagueness, some contracting states have adopted the position that the “legality” of a refugee’s presence is determined by exclusively resorting to the host state’s own domestic laws.[3] This position is problematic in its own right because – as explained below – it confers more power to states than is granted by the Convention. It becomes even more problematic when a contracting state revokes a refugee’s lawful presence status after the refugee acquired that status pursuant to domestic laws (despite the fact that the person is a recognized refuge or awaits such recognition). In this situation, the contracting state could – so this troubling position holds – freely expel the refugee without any Article 32 ramifications since the refugee is no longer “legally present.”

This problematic position, expressed by courts in countries such as the United Kingdom,[4] the United States,[5] and Germany[6] (but see Australia[7]), is certainly wrong. However, a position that holds that the phrase “lawful presence” should be interpreted without any deference to contracting states’ domestic law also seems flawed. In what follows, I will argue – following other refugee law scholars – that an interpretation of the “lawful presence” requirement should give some deference to domestic immigration laws. Such deference should end when domestic laws run against the Convention’s purpose. I will diverge, however, from other scholars in delineating the precise criteria that fulfill the “lawful presence” requirement and justify interference with domestic law. In particular, I reject Professor Hathaway’s claim that, once a person makes an asylum application, he becomes “legally present” in the host country for the duration of his refugee status determination (“RSD”) process. Instead, I suggest a more nuanced and multi-faceted test. This test takes into account, among other considerations, the duration of stay of the asylum seeker in the host country seeking to deport him to a “safe third country,” the ties that the asylum seeker has to the third country, and the stability which the asylum seeker could possibly enjoy in the third country (and, as an ancillary matter, the question of whether the planned transfer reflects burden-sharing or burden-shifting among states).

Interpreting “Lawful Presence” within the Meaning of Article 32: Combination of Domestic and International law

In a previous RefLaw article, I argued that ambiguous terms of the Convention, as in any convention, should be understood according to the rules of interpretation detailed in Articles 31-32 of the Vienna Convention on the Law of Treaties (“VCLT”), “and without taking colour from distinctive features of the legal system of any individual contracting state.”[8]

This is indeed the general rule. It is a rule of interpretation.[9] However, sometimes there are treaty provisions that explicitly or implicitly direct contracting states to apply their own domestic law, respect domestic considerations, or use their own discretion somewhat free from international law guidance while applying a specific treaty provision.[10] It follows that the contracting states acting in accordance with these principles are, therefore, applying an international norm. Hence, this is not really an exception to the general rule according to which national courts should make “every effort to interpret [a treaty] as it would be interpreted by an international tribunal and avoiding interpretations influenced by national interests.”[11] This is the case with the term “lawful presence,” which is at the heart of this article.

The term “lawful presence” appears throughout the Convention (and establishes what Hathaway labels the third “level of attachment” of the refugee to the country of asylum). Just like the term “lawful stay” (the fourth “level of attachment”), “lawful presence” is an interesting example in which the treaty itself directs the contracting states, at least in part, to resort to their own domestic laws while implementing their obligations under the Convention. Hathaway acknowledges this, as do other legal scholars of refugee law.[12] He writes that, “[a]s a starting point, the logic of deference to national legal understandings of lawful presence is clearly sensible,”[13] but adds that such deference cannot be absolute. This is so, he writes, because absolute deference “could result in [a] refugee never being in a position to secure more than [the] rights defined by the first two of the five levels of attachment agreed to by the state parties.”[14] Similarly, Ulrike Davy writes:

It is, on the one hand, undeniable that the 1951 Convention refers back to domestic law when certain rights are made dependent on the condition that the refugees’ presence be ‘lawful’. In principal, it is a matter of the contracting states to lay down the rules for lawfulness of presence. On the other hand, if refugees are to be protected from the arbitrariness of contracting States – and that is one of the main goals of the 1951 Convention – then the term “lawfulness” must also have some kind of international autonomous meaning… A deliberate withholding of rights undermines the purposes of the 1951 convention, at least if the refugees remain in the country for a considerable time.[15]

I agree that interpretation of the term “lawful presence” should be comprised of a combination of domestic and international norms; in other words, absolute deference should not be given to domestic immigration law. However, I respectfully disagree with the reasoning Hathaway and Davy offer for this proposition. I also respectfully disagree with what should be considered the most important criterion for “lawful presence,” which Hathaway offers as the pending asylum application. In the rest of this article, I will present this disagreement and offer a substitute test for “lawful presence.”

When is deference to domestic law overdone?

Both Hathaway and Davy seem to worry that contracting states would intentionally withhold the granting of formal legal status to the asylum seeker in order to prevent him from enjoying “third level of attachment” rights in their territory. However, if states do so, they violate VCLT Article 26, which provides, “Every treaty in force is binding upon the parties to it and must be performed by them in good faith.” In other words, if the reason for withholding legal status is a deliberate intent to prevent refugees from enjoying rights under the Convention, then the problem would be with the performance of the Convention rather than with its interpretation. Even if the correct interpretation of “lawful presence” is complete deference to domestic law, if the sole reason a contracting state withholds legal status is to deny asylum seekers “third level attachment” rights, they would be violating VCLT Article 26.

Still, this does not solve the underlying issue of how to interpret the term lawful presence.[16] The answer lies in the Convention’s subsidiary purpose, which is to allow refugees to restore stability in their lives (sometimes also referred to as the quest for a “durable” or “sustainable” solution). An additional subsidiary purpose of the Convention, which also carries some normative weight in the interpretation of Article 32, is promoting international cooperation among states in the area of refugee protection. I will discuss each of these two subsidiary purposes of the Convention below.

The Refugee Convention’s Subsidiary Purposes

It is a common understating that Article 33 is the “heart” of the Refugee Convention. The principle purpose of the Convention is, no doubt, to prevent a refugee’s return or transfer to a persecutory country. The country to which the refugee has (sometimes illegally) entered is asked to forgo its power of sovereignty and act as a “surrogate country” to a refugee-stranger. In other words, the country is asked to grant the refugee a set of basic human rights that his own country failed to grant him. Indeed, as the High Court of Australia ruled, the chief purpose of the Refugee Convention is “to impose obligations on the signatories to the Convention to provide protection and equality of treatment for the nationals of countries who cannot obtain protection from their own countries.”[17]

However, treaties often have more than one purpose.[18] While it is clear that the Convention’s main purpose is to protect refugees from persecution on account of one of the enumerated five grounds, I will discuss two subsidiary purposes of the Convention: restoring refugees’ sense of stability, and promoting burden-sharing through international cooperation. These subsidiary purposes, I will argue, are relevant for interpreting the term lawful presence within the meaning of Article 32.

Restoring Refugees Stability as a Subsidiary Purpose of the Refugee Convention

An important subsidiary purpose of the Convention is to offer refugees stability. A refugee, by definition, is a person who had to cross at least one international frontier in his quest for asylum. Such a person, so the argument goes, should be given a chance to restore stability in his life. The main purpose and the subsidiary purpose of the Convention are connected because, by requiring countries not to return a refugee to another country where he might face persecution, the Convention already offers the refugee some kind of stability. However, this is the bare minimum from the perspective of stability. It might not be enough to provide the refugee with a proper and enduring sense of stability that can lead to the restoration of psychological well-being. The Convention reflects this understanding and offers further stability-enhancing measures beyond mere protection from refoulment.

This purpose is advanced, first and foremost, by Article 34, which provides that “[t]he Contracting States shall as far as possible facilitate the assimilation and naturalization of refugees.” [19] By including this Article, the framers intended to provide refugees with a substantial sense of stability, despite the fact that – by definition, and by explicit indication of Article 1C (“the cessation clause”) – a refugee’s status is ephemeral and discontinues once “the circumstances in connection with which he has been recognized as a refugee have ceased to exist.”[20] If a refugee is, after all, naturalized – as Article 34 encourages contracting states to do – then Article 1C becomes irrelevant, and a refugee can experience the same sense of stability as a citizen, at least as far as his legal status is concerned.

Another indication that the Convention goes beyond merely providing protection from persecution and towards enduring stability for refugees is the “compelling reasons” caveat in Articles 1C(5) and 1C(6).[21] The framers found it important to indicate that, in some circumstances, a refugee would not be returned to his country of citizenship even when the risk he once faced there had dissipated. Like Article 34, the “compelling reason” exception does not serve the purpose of protection from persecution; both Articles go beyond that. Specifically, these Articles serve an array of humanitarian considerations, which include at their core restoring refugees’ stability and psychological well-being.

Article 32 serves the same purpose. By prohibiting a refugee’s expulsion to non-persecutory states (once the refugee acquires “lawful presence” in the host state), the Article offers the refugee peace of mind that his wandering days have come to an end; the host country will not remove him to another country except for in rare cases where he poses a risk to public security or public order. The term “legally present” as it appears in Article 32 should be interpreted in light of this purpose.

Burden-Sharing and Promoting International Cooperation as a Subsidiary Purpose of the Refugee Convention

An additional subsidiary purpose of the Refugee Convention is to promote cooperation among states in matters of refugee protection, in order to establish a fair system of responsibility- and burden-sharing. This principle is reflected in the fourth recital of the Convention’s preamble (“Considering that the grant of asylum may place unduly heavy burdens on certain countries, and that a satisfactory solution of a problem of which the United Nations has recognized the international scope and nature cannot therefore be achieved without international co-operation.)”[22]

As most compulsive secondary movement plans are based on seemingly consensual agreements among sovereign states, such plans could allegedly be considered to be a good practice of international cooperation aimed to promote fairer burden-sharing. After all, the receiving country seemingly agrees to accept the asylum seekers, a fact that supposedly proves cooperation rather than coercion between the parties to the secondary movement arrangement. Moreover, sometimes such an agreement is part of a broader pact between the countries, which may also involve country A remunerating country B for agreeing to take upon itself the cost of admitting the transferred asylum seekers into its territory. To invoke an analogy to a contract signed under a laissez faire regime, there is no reason to interfere with a free understanding between two autonomous actors.

Yet the matter is more complicated than this. First, international agreements under international law are very different from financial contracts under contract law, and it is far from obvious that the analogy holds. Since the fourth recital specifically mentions the need to tackle “unduly heavy burdens on certain countries” through international cooperation, one could argue that an agreement which shifts an additional burden to an already heavily burdened country runs against the purpose of the Convention, and is therefore unlawful, at least in extreme situations of burden-shifting. In other words, the argument might be that it is not enough that the agreement is consensual, it must also advance the Convention’s subsidiary purpose of burden-sharing (and that this understanding influences the manner in which we should interpret the tern “lawful presence” within the meaning of Article 32, as suggested below).

Second, even if the analogy holds, just as a transaction under a laissez faire regime could be unfair (and legally invalid) due to duress, exploitation or unconscionability, so – one could argue – an agreement between two sovereign counties for the transfer of asylum seekers could be deemed unfair and invalid for similar reasons.

More importantly, just as an agreement could be deemed unlawful because of its implication on third parties (“externalization”), so can an agreement between two countries for a compulsory transfer of asylum seekers from one country to the other be deemed unlawful due to a violation of the transferees’ rights granted to them under the Convention or under any other human rights instrument. In other words, if the agreement is more in line with the nature of “burden-shifting” than “burden-sharing,” and the overburdened country could not assure the transferees a set of rights as prescribed to them in the Convention and other human right instruments, or the transfer would cause them further instability due to a volatile status in the overburdened country, this could render the agreement illegal, and – as suggested below – a violation of Article 32. After all, the subsidiary purpose of the Convention is to encourage international cooperation that would ensure better, not worse, protection of asylum seekers. One could further argue that the Convention’s subsidiary purpose to promote international cooperation is subordinated to the Convention’s subsidiary purpose to restore refugees’ stability. An agreement between two countries to transfer refugees from Country A, where their status could potentially be stable (e.g. due to good economic conditions, low number of asylum seekers in the country, and decent human rights protection), to Country B, where their status would be volatile (e.g., due to poor economic conditions, high number of asylum seekers in the country, and poor human rights protection), might advance the Convention’s purpose to encourage international cooperation among states but run against the more important purpose of restoring refugees’ stability (and in even worse cases, against the Convention’s purpose to protect refugees).

“Lawful Presence” under Article 32 – the Criteria

Let us now turn to the criteria for “lawful presence.” As mentioned above, Hathaway, Davy, and other legal scholars concede that the starting point for legality of presence should be determined according to the host state’s immigration laws. I agree. The reason for this position stems from the ordinary meaning of the term lawful presence in light of basic principles of international law.[23] Sovereignty under international law entails, first and foremost, a state’s authority to adapt its own immigration and naturalization laws.[24] When the Refugee Convention prohibits the expulsion of a foreigner under certain circumstances, it is an important limitation on a state’s sovereignty. But the general rule remains that states have the power to design their own immigration laws, including the regularization of lawful versus unlawful presence. As Article 32 does not explicitly dictate otherwise, this should remain the starting point. However – and here again I agree with Hathaway, Davy and others – this cannot be the end point. When interpreting the ambiguous phrase lawful presence, one must choose the interpretation that aligns with the context of the Convention’s provisions and promotes the Convention’s purpose (alongside general principles of international law like state sovereignty).

Article 32 deals with expulsion to non-persecutory states. Hence, the primary purpose of the Convention – protection from persecution – is irrelevant for its interpretation. However, the Convention’s subsidiary purposes, i.e., restoring stability and promoting burden-sharing through international agreements, are highly relevant. Hence, while the starting point for interpreting lawful presence within the context of Article 32 is indeed deference to domestic law, if domestic law deviates too far from these purposes, deference should cease.

Now we get to the bottom line – when does domestic law deviate too far from the Convention’s purpose(s), and how should the phrase lawful presence be interpreted in such a case? At this very practical level, I diverge from Professor Hathaway.

Critique of Hathaway’s Criteria

Hathaway argues that a refugee is legally present in one of three scenarios: First, if a contracting state admits the asylum seeker into its territory, the refugee is lawfully present for the duration of the authorization of such admission; second, as long as the asylum seeker’s application for asylum is being processed (including through an appeal period), he is lawfully present; third, if the contracting state decides to forgo RSD procedures altogether and instead employs, for example, a “temporary protection” regime, then the refugee is lawfully present.[25]

My main concern is with the second criterion, which is the most prominent of the three case types, as Hathaway himself acknowledges.[26] I think such a criterion for “lawful presence” is wrong, at least as a stand-alone criterion, for the following reasons.

First, there seems to be a significant gap between, on the one hand, Hathaway’s discussion regarding the primary role of domestic law in interpreting lawful presence and, on the other hand, this criterion. “Interpretation of the notion of ‘lawful presence,’” writes Hathaway, “should … look primarily to domestic legal requirements, interpreted in the light of the small number of international legal understandings on point.” [27] However, granting lawful presence to each asylum seeker who files an asylum application, potentially the moment he sets foot in the contracting state and against the contracting state’s own immigration laws and regulations, constitutes a significant intrusion on domestic law. It means completely setting domestic law aside in favor of, as Hathaway puts it, a norm anchored in the Refugee Convention. If all one has to do in order to gain legal status is to ask for asylum, this is quite a serious encroachment on domestic immigration law and policy. Consider the rights that flow from the Convention once legal status is established: protection from expulsion to a non-persecutory country (Article 32), freedom of residence and internal movement (Article 26), and the right to self-employment (Article 18).

True, a state is free to exercise quick RSD proceedings that would potentially screen out undeserving applicants. However, with the rise in the number of refugees worldwide, quick proceedings are becoming more and more unattainable, even by countries seeking to perform their responsibilities toward refugees in good faith. I agree with Hathaway that granting conditional rights to asylum seekers who await a decision in their applications could incentivize states to speed up RSD proceedings (or forgo them altogether in favor of some kind of collective protection), which is a good thing. Hence, my suggested criteria also consider the duration of time (awaiting a decision on the asylum application) as a highly relevant factor. However, granting an asylum seeker “lawful presence” and a bundle of meaningful rights based on this status solely by filing an application (which could happen the minute the person entered the country) gives absolutely no deference to domestic immigration laws.

Second, Hathaway argues that his position that a refugee is legally present throughout the course of his RSD determination is derived from Article 31.[28] This, however, is a non loquitor. Article 31(1) deals with immunity. Specifically, it provides that “[t]he Contracting States shall not impose penalties, on account of their illegal entry or presence, on refugees who, coming directly from a territory where their life or freedom was threatened in the sense of [A]rticle 1, enter or are present in their territory without authorization, provided they present themselves without delay to the authorities and show good cause for their illegal entry or presence.”[29] It provides that a refugee who complied with the requirements specified therein (“coming directly,” “without delay”) would be immune from prosecution or other sanctions, as long as his asylum application is processed. However, a right and an immunity are two separate things. One could perpetrate an offence, for example, yet be immune from being prosecuted for it. Hence, immunity, while implying non-liability, does not necessarily imply right-holding or lawfulness. Hathaway is certainly correct that Article 31 infers that the status of a refugee who entered into the territory of a contracting state illegally may subsequently be regularized. However, it is an overstatement to say that the making or lodging[30] of an asylum application, in and of itself, immediately regularizes the refugee’s status and renders his presence legal.

Moreover, refugees that came indirectly to the host state (i.e., passing through a transit state) and those who “presented themselves to the authorities” (i.e., applied for asylum) with delay are left with an unclear status when they rely on Article 31 for “lawful presence” status. Consider, for example, a refugee that enters the territory of the state without authorization after passing through a transit state or a refugee that enters the territory of the state without authorization and waits to file an asylum application. In these examples, when and how would the refugees gain “lawful presence” in the host state? What if the host state fails to make a decision on their (delayed) asylum application in a timely fashion?

Third, Hathaway’s first scenario, that lawful presence is granted upon authorized admission and throughout such authorization, is internally incongruent. If filing an asylum application makes an asylum seeker, ipso facto, legally present, as long as the application is still pending, it follows that Hathaway’s first scenario comes into play without an asylum application being lodged at all. Otherwise, the first scenario becomes redundant. However, this would be a strange criterion for lawful presence for refugee law concerns. It is a common understanding that a refugee carries the initial and minimal duty to ask for asylum. As a general rule, a contracting state is not expected to guess or assume that an immigrant is a refugee or encourage him to ask for asylum. Hence, the mere legal entry and sojourn, absent an asylum application, cannot make an immigrant a legally present refugee.

Fourth, establishing a legal status based on the formality of filing an asylum application might lead to an absurd, not to say macabre scenario. Specifically, an immigration inspector may attempt to assert that he is contemplating sending the refugee to a safe third country, even before the refugee has the chance to ask for asylum. The motivation for the inspector’s assertion may be to avoid the formation of the third level of attachment,[31] which would accordingly prevent the host country from expelling the asylum seeking under, e.g., a compulsory secondary movement plan.

Finally, Hathaway’s criterion seems to be at odds with his own later writings. Recently, for example, he reasserted his plan for global responsibility sharing of refugee law.[32] He proposed that, when an asylum seeker stays five years in a host state, and neither repatriation nor local integration is viable, then he should be resettled in a third country. While the question of voluntary versus involuntary resettlement is unclear according to this plan, it seems sensible that the asylum seeker would not have the right to veto resettlement under such a plan.[33] Moreover, if after five years the asylum seeker has not yet been offered local integration, one might wonder how voluntary his decision to leave to a suggested third country would be. Nevertheless, Hathaway writes that the suggested plan “would require no amendment of the Refugee Convention.”[34] However, if lawful presence consolidated at the moment an application for asylum was lodged, how could an asylum seeker be compulsorily or constructively expelled to a third country five years later? And how is this compatible with Hathaway’s claim that filing an asylum application makes a person lawfully present in the host country, thus triggering the Article 32 right to non-expulsion?

The Suggested Criteria

I argue that the term lawful presence in Article 32 should be interpreted in light of the Convention’s subsidiary purpose to restore refugees’ stability and psychological well-being, and (to a lesser degree) in light of the Convention’s subsidiary purpose to promote burden-sharing through international cooperation. These purposes serve as Dworkinian principles[35] rather than as rules in the sense that they carry normative weight. Such normative weight should be balanced by opposing rules and principles, and could potentially be overridden by them, depending on how “heavy” each side of the scales is. For example, an opposing consideration could be respect of a state’s sovereign power to decide its own immigration rules and to sign legally binding international agreements with other states, including agreements for the transfer or admission of asylum seekers.

Given the two subsidiary purposes of the Convention analyzed above, the following are, I argue, the key[36] substantive[37] factors that should be taken into account when considering whether an asylum seeker could be expelled to a non-persecutory[38] country:

  • The duration of time that the asylum seeker is physically present in the host state (and, in cases where the asylum seeker has waited to file an asylum application – the duration of time that has passed since such filing),
  • Whether the asylum seeker has any ties in the third country (as compared to ties in the country wishing to transfer him), including for example biographical ties (e.g., previous habitation there), family ties, and cultural or linguistic ties[39],
  • The stability of the legal status of the asylum seeker in the “third” country (e.g., whether the asylum seeker would be granted citizenship or permanent residency status, or whether the status would be temporary and volatile),[40]
  • Whether the compulsory secondary movement agreement is a burden-sharing or burden-shifting arrangement (this factor could reflect on the previous one, as explained above while discussing the “externalization” argument, but could also stand on its own).[41]

The first three factors stem from the Convention’s subsidiary purpose to restore refugees’ sense of stability and psychological well-being. The fourth emanates from the Convention’s subsidiary purpose to promote burden-sharing of refugees’ protection by international cooperation. Some of them also have specific manifestations in the Convention. For example, the idea that the longer the refugee’s presence in the country of asylum the broader the set of rights such country owes him is manifested in Article 17(2)(a) of the Convention, which provides that “restrictive measures imposed on aliens or the employment of aliens for the protection of the national labour market shall not be applied to a refugee who… has completed three years’ residence in the country.”[42] Similarly, Article 7(2) of the Convention provides: “After a period of three years’ residence, all refugees shall enjoy exemption from legislative reciprocity in the territory of the Contracting States.”[43]

The second factor, giving regard to the asylum seeker’s ties to the “third country” or lack thereof, is also reflected in the Convention. Consider Article 17(2)(b)-(c), which provides that “restrictive measures imposed on aliens or the employment of aliens for the protection of the national labour market shall not be applied to a refugee who… has a spouse possessing the nationality of the country of residence” (sub-section (b)) or “who has one or more children possessing the nationality of the country of residence” (sub-section (c)).[44]

As mentioned above, the idea of durable stability is manifested in several articles of the Convention, particularly Article 34, and the vision of burden-sharing through international cooperation is reflected in the Convention’s forth recital.

Let us deliberate on these criteria by exploring some real-life examples of compulsory secondary movement plans and rules. Consider the Dublin Regulations, for instance. The Regulations, which are an agreement among EU Member States and certain other European states, deal with the compulsory transfer of an asylum seeker from one European state (“the subsequent state”) to another, where the asylum seeker was present before (“the previous state”). Such transfer can be done within the first six months of his stay in the subsequent state,[45] i.e. before the asylum seeker put down deep roots in that state. The movement is into another (the previous) European state, most of which are non-persecutory in the broad sense used herein.[46]

Further, these European states typically also grant refugees a stable recognized legal status for the duration of the RSD procedures and enable, once the refugee is recognized as such, permanent status and naturalization. Since the asylum seeker passed through the previous state, to which he is being transferred, on his way to the subsequent state, some biographical ties might also exist (but this admittedly is not always the case). On the other hand, one could argue that the Dublin Regulations look more like a burden-shifting arrangement rather than a burden-sharing one, given that many asylum seekers arrive to Western Europe though the Mediterranean Sea by boats, entering Greece and Italy first, before continuing to other European countries.[47] Still, considering all of the factors, including taking domestic (here regional) law as the starting point, which is the default rule according to the suggested legal analysis, it seems that the Dublin Regulations are in compliance with Article 32.

The same is true regarding the “safe third [non-EU] country” scheme under Article 38 of the EU Asylum Procedure Directive.[48] Here too, the decision to reject an asylum application based on the safe third country concept can only be made during the initial stages of the asylum seeker’s reception in the Member State.[49] Moreover, according to Article 38(a)(2) of the Directive, the Member State must also consider whether there is “a connection between the applicant and the third country concerned on the basis of which it would be reasonable for that person to go to that country.”[50] This last criterion is often mentioned by Courts, especially in the United Kingdom, as particularly substantial in deciding whether to deny a request for asylum based on a “safe third country” solution. As the British Upper Tribunal wrote in 2010:

[T]he type of case with which we are concerned here, involving intended expulsion of a refugee, tends only to arise as a matter of international state practice in situations where the person concerned has some connection with the third state which is said to be safe, based on nationality, prior residence, marriage, entitlement to residence, historical ties etc. It does not arise simply because there is a safe third country somewhere. Within the EU, the requirement of a connection is now stipulated in [the Asylum Procedures Directive].[51]

Regarding the stability of the refugee’s status in the third country, the Directive requires at a minimum that the asylum seeker could “request refugee status and, if found to be a refugee . . . receive protection in accordance with the Geneva Convention.”[52] Considering all of the factors I suggest above, and in particular on the basis of the required connection between the asylum seeker and the third country, this EU scheme also appears to comply with Article 32.

However, some compulsory secondary movement plans are more legally contentious. Two countries that have more problematic policies are Australia[53] and Israel.[54]

Australia initially signed bilateral agreements with Papua New Guinea (PNG) and Nauru for the temporary transfer of so-called “boat people” approaching Australia’s shores. The purpose of this agreement was to complete RSD processing in Regional Processing Centers (RPCs) on those third countries’ territories.[55] In 2013, Australia extended its agreements with PNG[56] and Nauru[57] to prolong the stay of the RPC recognized refugees. This extension supposedly occurred on the basis of such refugees’ consent to stay in these two countries. In 2014, Australia signed a now-expired agreement with Cambodia in which the latter agreed to admit recognized refugees from Nauru RPCs for a prolonged stay.[58] Here too, the transfer (from Nauru to Cambodia) was supposedly voluntary. NGOs and academics argued, however, that the refugees’ stay was involuntary in either PNG or Nauru.[59] Further, they argued that the transfer from Nauru to Cambodia was involuntary as well.[60] Australia’s top politicians made it clear that it would not allow any RPC recognized refugee to enter Australia.[61] At the same time, the RSD procedures in the RPCs were extremely long and the living conditions in the RPCs were very poor.[62] Hence, so goes the claim, the refugees had no choice but to agree to stay in PNG or Nauru, or transfer to Cambodia, simply out of the necessity to leave substandard RPCs.

Australia’s initial agreements with PNG and Nauru are problematic from the perspective of restoring a refugee’s stability in life. The mere fact that the agreements regulate the transfer of refugees to another country where they have no pre-existing ties is enough to cause instability. Even further, the initial plan proposed returning the refugees to Australia once they were recognized as refugees. In no way can this transfer from Australia to PNG and Nauru and later on, supposedly, back to Australia be viewed as restoring a refugee’s sense of stability. The same is obviously true of the likely involuntary transfer from Nauru to Cambodia. In this sense, were the conditions in PNG and Nauru acceptable, a continued stay there would have been more defensible. However, the refugee’s status in those countries was volatile. The countries only grant the recognized boat people that agreed to stay there a very limited set of rights.

Given the huge economic gaps between Australia and PNG, Nauru and Cambodia, and despite the low number of refugees seeking asylum in these countries, it is quite obvious that these agreements cannot be justified as a burden-sharing agreement; rather, they are clearly burden-shifting.[63]

Taking these considerations into account, Australia’s plans seem to violate Article 32.[64] Consider the effect of adopting Hathaway’s criteria as opposed to my suggested criteria. According to Hathaway, no lawful presence was formed in this situation because the transfer to the offshore process was immediate and possibly took place before the boat people asked for asylum. According to my suggested criteria, lawful presence was formed and therefore the scheme is in violation of Article 32.

We next turn to Israel’s agreements with Rwanda and Uganda. Between 2006 and 2013, around 60,000 Eritrean and Sudanese asylum seekers entered Israel through its then open southern border.[65] Upon their entry, Israel granted them temporary protection under the rule of non-refoulment, but until 2013 did not enable them to file for asylum.[66] Israel’s stance is that the vast majority of Eritrean and Sudanese who came to Israel are not refugees within the Convention’s definition. Hence, its asylum authority rejects over 99% of asylum applications of these nationals (which is the rejection rate in Israel generally).[67]

Around 2014, Israeli immigration officials began offering Eritrean and Sudanese nationals the sum of $3,500 if they agreed to go to Rwanda or Uganda. The offer was based on a “safe third country” agreement that Israel claimed to have signed with each of these countries, but remained confidential.[68] The “offer,” which was initially left for the Eritrean or Sudanese national to accept or reject, was soon to be accompanied by a threat that a refusal would lead to unlimited immigration detention. However, the authorities at this point applied the “Rwanda/Uganda or detention” scheme only to Eritreans and Sudanese who either never applied for asylum (yet, as mentioned, enjoyed temporary protection under the rule of non-refoulment), or those relatively few whose applications were denied (and continued to enjoyed temporary protection as well). Once the threats materialized, the “Rwanda/Uganda or detention” scheme was challenged in Court.[69]

The Israeli District Court ruled that Rwanda and Uganda were both “safe third countries,” and therefore, in principle, there was no legal barrier to approving the prolonged detention of Eritrean and Sudanese nationals who refused to be transferred to either of those countries. The Supreme Court affirmed with regard to Rwanda (at that point expulsions to Uganda were suspended, and the Court therefore did not discuss the Ugandan case). However, the Supreme Court found (after reviewing the Israel-Rwanda agreement in camera) that Rwanda agreed to accept only Eritreans and Sudanese who voluntarily agreed to leave Israel and enter its territory. Hence, the Court ruled, Israeli authorities could not detain the Eritrean and Sudanese nationals in the hope that they would ultimately “consent” to go to Rwanda, as this would not be an exercise of free will, and the agreement conditioned the transfer on free will. The petitioners, it should be noted, did not raise Article 32 arguments, supposedly because the “Rwanda/Uganda or detention” scheme did not apply to asylum seekers awaiting their decisions, only to those who either never applied for asylum or applied and were rejected. The Court also remained silent on this matter.[70]

Later on, at the beginning of 2018, Israel claimed to revise the agreements with Rwanda and Uganda in a way that would allow the involuntary transfer of Eritrean and Sudanese to their territories. Around the same time, Israel’s Population and Immigration Authority promulgated a new regulation that stipulated that Eritrean and Sudanese nationals, who (a) never filed for asylum, (b) filed and were denied asylum, or (c) filed for asylum subsequent to the date January 1st 2018, must leave to one of the safe third countries or otherwise they could be put in immigration detention. Because the regulation now applied to people who lodged an asylum application, Article 32 potentially comes into play.

At the end of the day, seemingly due to public pressure, Rwanda and Uganda backed away from their new agreements with Israel. While the 2018 regulations are still in force, they are ineffective at this point in time, as Israel has not signed new agreements with countries that agree to admit asylum seekers for Israel without their consent. The Supreme Court, which started deliberating the case, did not rule on it because it became increasingly obsolete. Had the Court ruled on the matter, however, I believe it should have ruled that the expulsion plan was in violation of Article 32, at least with regard to the asylum seekers who were awaiting decisions on their applications, regardless of the timing of lodging the application.

Practically all Eritrean and Sudanese asylum seekers entered Israel between 2005 and 2013. In 2013, Israel completed building a fence along its Southern border. Since then, only a handful of asylum seekers have been able to pass through it. Hence, in January 2018, most Eritrean and Sudanese nationals have been living in Israel for at least five years. Many of these individuals have been in Israel for over a decade. As mentioned, Israel did not allow Eritrean and Sudanese asylum seekers to apply for asylum until 2013 (on the grounds that they were immune from deportation due to their temporary protection status), so it would be problematic and unconscionable to penalize them for delaying an application for asylum. Moreover, many of these individuals believed that their initial registration with UNHCR’s office was equivalent to lodging an application with the Israeli authorities.

Arguably, then, the Israel-Rwanda and Israel-Uganda agreements purported to transfer Eritrean and Sudanese asylum seekers. These asylum seekers were in Israel for a considerable amount of time. During this period, Israel failed to process their asylum applications while declaring that they were non-deportable under the principle of non-refoulment. After such a long period, asylum seekers had put down roots in Israel. Some learned Hebrew, became part of the work-force, married, had children, and integrated into Israel despite public pressure and governmental efforts to prevent that from happening. It is highly destabilizing to transfer these individuals to a third country after such a long stay. In addition, the Eritrean and Sudanese asylum seekers had no previous or existing ties to the two countries in the agreement. Israel did not claim that according to the agreements it signed with Rwanda and Uganda the transferees would enjoy permanent status there, and according to information gathered by UNHCR and NGOs, the status of those who left Israel to go to one of these countries, under the agreements, was volatile.[71] Like in the Australian case, Rwanda and Uganda are both significantly less economically well-off than Israel, and both host significantly higher numbers of refugees than Israel. So, as UNHCR indeed stated,[72] this plan was a burden-shifting rather than a burden-sharing plan. Considering all the factors, this plan was a clear violation of Article 32.

Challenges to the suggested Criteria

Several challenges could be directed against my suggested interpretation of Article 32’s lawful presence component. I will address two of them here.

One challenge is that (a) my suggestion that “legally present” should be interpreted in consideration of factors like the ties that the refugee has to the third country infers that the suggested interpretation is applicable to the usage of the term in Article 32 only, and not in other articles of the Convention that use the same term; and (b) that such discrepancy (interpreting the term “lawful presence” differently in different articles) is (i) legally wrong or (ii) cumbersome and unwarranted.

I agree with proposition “a.” My suggested interpretation would indeed be (at least partly) irrelevant and inapplicable to the other two provisions that include the phrase lawful presence, those Articles dealing with internal freedom of movement (Article 26) and self-employment (Article 18).

Preposition b(i) is itself wrong. VCLT Articles 31-32 allow the possibility that identical words or terms can be interpreted differently.[73]

Regarding proposition b(ii), it may indeed make a jurist’s job easier if the term lawful presence was interpreted identically throughout the Convention. However, there is no rule of interpretation that considers whether it would be easier and nicer to have a single meaning of a term throughout a treaty. VCLT Article 31 refers the jurist-interpreter to the ordinary meaning of words in their context and in light of the purpose of the Convention. But the immediate context of a term might change from provision to provision, even while the wider context of the treaty remains. Also, as argued in detail, a convention might have more than one purpose. And while one article that mentions the term could mainly advance purpose A of the treaty (e.g. protection of refugees), a second article that mentions the same term could mainly advance purpose B of the treaty (e.g. restoring refugees’ stability). Hence, while it might be preferable to have a single interpretation of the same term throughout a treaty, this cannot be a decisive argument.

A second, more exigent challenge is the claim that the suggested criteria could indeed support a warranted legal rule (if we could return in time to draft Article 32 anew), but it has nothing to do with existing law, i.e. it is not an interpretation of Article 32, as it is drafted. The suggested criteria, the challenge goes, simply departs too far from the ordinary meaning of the term legal presence.

The critique could continue by using the following example: take, for example two asylum seekers who are candidates for an expulsion to a “safe third country.” Both entered the country (which now wishes to expel them) at the same time, filed their asylum applications together, and they both await a decision on their application. However, asylum seeker A has past ties with the country of intended transfer while asylum seeker B does not. Could it be that only one of them (asylum seekers B) is a “refugee lawfully in the territory,” while the other is not? Similarly, take an asylum seeker who is a candidate for a compulsory transfer to Country A, where his status would be volatile, or alternatively to Country B, where his status would be stable. Is it possible that the same asylum seeker will be “lawfully in the territory,” within Art. 32’s meaning, when his expulsion to Country A is reviewed yet “unlawfully in the territory” when his expulsion to Country B is under consideration?

Surprising as it may sound to some readers, my answer to both these hypothetical questions is in the affirmative. There is nothing analytically or conceptually wrong with a contextual, nuanced and multi-faceted – rather than static – understanding of the term lawful presence within the context of Article 32, i.e. a reading that takes into account factors from the country where the refugee is physically present, the country of intended expulsion, and the refugee himself (and his relation to each of these two countries). If the VCLT rules of interpretation call for such an interpretation – and I claim that they do – then this is the correct interpretation of the term.

Legal rules, unlike rules of nature, are of a normative, rather than a descriptive character. Lawfulness (as in the term lawful presence), by its very nature, is a changeable concept, reliant on the contingent rules of legality in a specific setting. There is, therefore, nothing “unnatural” or inherently wrong in a flexible and adaptable understanding of the term lawful presence, if this is what humans (people who hold an authoritative power to craft and interpret law) decide it to be. This could well be the “ordinary meaning” to be given to the term lawful presence in its context and in light of the Convention’s purpose.

Modern legal systems are comprised of both “primary rules” and “secondary rules,” as H.L.A Hart wrote,[74] and the secondary rules contain rules of interpretation. In international law, the rules of interpretation are codified in Articles 31-32 of VCLT. My interpretation is in accordance with these rules. The starting point in interpreting the term lawful presence is a presumption of deference to domestic law. This presumption derives from the principle of sovereignty, and in accordance with VCLT Article 31(3)(c), which dictates that in interpreting a treaty, “relevant rules of international law” (i.e. the rule of respecting sovereignty of all states) should be taken into account, together with the context of the treaty’s terms. The considerations that could overturn this presumption, like a (long) duration of stay, (no) ties to the third country and (in)stability of status to be accrued in the third country, as well as the consideration of burden-sharing vs. burden-shifting, all stem from the immediate context of Article 32 and the aforementioned subsidiary purposes of the Convention; i.e. they are adopted as part of an interpretation of the term “lawful presence” in its context (of Article 32, i.e. expulsion) and in light of the Convention’s purpose(s), according to VCLT Art 31(1). While this might be a creative reading of the ordinary meaning of the term lawful presence, in the context of Article 32 and in light of the Convention’s purposes, it is, I argue, a viable one.[75]


In this article, I suggest a contextual, nuanced and multi-faceted interpretation of the term lawful presence, which appears in Article 32 of the Refugee Convention. Compulsory secondary movement plans and agreements, gaining popularity in the last several decades, must be compatible with Article 32, and such compatibility should be achieved and judicially reviewed, I argue, by adopting the suggested interpretation of the Article.

I suggest a number of criteria that should be considered when interpreting Article 32. Most of these criteria stem from the desire to minimize further erosion in refugees’ (already undermined) sense of stability.

My suggested reading of Article 32 already has some anecdotal underpinnings in existing Article 32 and safe third country jurisprudence.[76] But some polishing is in order. My hope is that this article will help to elucidate the array of relevant considerations that have to be taken into account when considering the transfer of an asylum seeker to a third country. Compulsory removal of any person to a different country is a brutal measure, and much more so when that person is a refugee who already fled his home country. Hence, transfers should not be taken lightly, and certainly transfers must uphold proper procedural safeguards that ensure a refugee’s rights. There are certainly more factors, procedural measures, and creative ideas to be considered beyond the few I have mentioned. I hope future writings will deal with these pressing matters to help alleviate the difficult conditions of refugees worldwide and to restore their sense of stability.

I wish to thank Shany Bar-Tuvia, Aelad Cahana, Tally Kritzman-Amir, and Maayan Niezna for their valuable comments on a previous draft of this article.

[1] Ulrike Davy, Art. 32 1951 Convention, in The 1951 Convention Relating to the Status of Refugees and its 1967 Protocol: A Commentary 1277, 1324 (Andreas Zimmermann ed., 2011).

[2] I refer to the phrase “object and purpose” in the context of Article 31 of the Vienna Convention of the Law of Treaties. Vienna Convention on the Law of Treaties art. 31(1), May 23, 1969, 1155 U.N.T.S. 331 <> last accessed August 28, 2019 [hereinafter Vienna Convention]. The difference between “object” and “purpose” is minor, and “[p]ractice shows that courts and tribunals have tended to treat the term ‘purpose’ as a single but broad remit.” Richard K. Gardiner, Treaty Interpretation 193 (2d ed. 2015). Therefore, I will use, hereinafter, the shortened term “purpose” instead of “object and purpose.”

[3] See infra notes 4-6.

[4] See, e.g., Secretary of State for the Home Department v. ST (Eritrea) [2010] EWCA (Civ) 643 (Eng.); RR (Refugee – Safe Third Country) Syria v. Secretary of State for the Home Department [2010] UKUT 422 (IAC) (Eng.); IH v. Secretary of State for the Home Department (Eritrea) [2009] UKAIT 00012 (Eng.).

[5] See, e.g., Kan Kam Lin v. Rinaldi, 361 F. Supp. 177 (D.N.J. 1973), aff’d per curiam, 493 F.2d 1229 (3d Cir.), cert. denied, 419 U.S. 874 (1974). It should be stressed, though, that while the wording of this decision (and others) imply that U.S. domestic law alone governs the legality of any immigrant’s presence in the U.S. (including for Article 32 matters), in this case (and others), the asylum seeker overstayed his visa, and made an asylum claim only after deportation procedures were applied against him.

[6] See, e.g., Yugoslav Refugee (Germany) Case, 26 Int’l L. Rep. 496, 498 (Ger. FASC, Nov. 25, 1958) (ruling that when a refugee enters Germany illegally, he would become “legally present” for Article 32 concerns only when he acquires lawful residency in the country); see also Bayerischer Verwaltungsgerichtshof [Higher Administrative Court of Bavaria] July 10, 2009, 10 ZB 09.950 (ruling that (recognized) refugees lose their rights under Article 32 as soon as they have been served with an expulsion order (ignoring their rights under Article 32), since the expulsion order renders their presence unlawful). But cf. Bundesverwaltungsgericht [Federal Administrative Court)] Sept. 30, 1958, I C 172/57 (holding that administrative authorities must not decide upon expulsion as long as the question of refugeehood was not settled). The author of this article does not speak German and relies on subsidiary sources for an understanding of these cases. Davy, supra note 1; James C. Hathaway, The Rights of Refugees Under International Law (2005).

[7] See Rajendran v The Minister for Immigration & Multicultural Affairs [1998] FCA 1085 (Austl.).

[8] Yuval Livnat, Can Israel Freely Interpret the Refugee Convention (When it Comes to Palestinian Asylum Seekers?), RefLaw (July 11, 2016) last accessed August 28, 2019; Regina v. Home Secretary for the Home Department, ex parte Adan (HL) [2001] 2 AC 516-517 (Eng.).

[9] “[N]ational authorities enjoy comparative institutional advantages over international courts with regard to fact-finding and fact-assessing exercises, but not in relation to norm-interpretation projects. As a result, a general margin of appreciation doctrine should mainly govern fact-intensive law-application decisions and not norm-intensive law-interpretation processes, whose ultimate elaboration should remain the exclusive province of the international judiciary.” Yuval Shany, Towards a General Margin of Appreciation Doctrine in International Law?, 16 Eur. J. Int’l L. 907, 913 (2006) last accessed August 28, 2019 (distinguishing between interpreting a norm of international law and applying it, and allowing state a limited margin of appreciation in the latter endeavor, but not in the former).

[10] Consider, for example, Article 2(3) of the International Covenant on Economic, Social and Cultural Rights, which provides: “Developing countries, with due regard to human rights and their national economy, may determine to what extent they would guarantee the economic rights recognized in the present Covenant to non-nationals.” International Covenant on Economic, Social and Cultural Rights art. 2(3), Dec. 16, 1966, 993 U.N.T.S. 3 <> last accessed August 28, 2019.

[11] The Activities of National Judges and the International Relations of Their State, Inst. Int’l L. art. 5(3) (1993) <> last accessed August 28, 2019.

[12] See. e.g., Violeta Moreno-Lax, The Legality of the ‘Safe Third Country’ Notion Contested: Insights from the Law of Treaties, in Migration and Refugee Protection in the 21st Century: International Legal Aspects 663, 704 (Guy S. Goodwin-Gill & Philippe Weckel, eds. 2015) last accessed August 28, 2019.

[13] Hathaway, supra note 6, at 177.

[14] Id.

[15] Davy, supra note 1, at 1304.

[16] The good faith requirement is part of Article 31(1) too. Hence, if one interprets “lawful presence” by granting complete deference to domestic law, knowing that the state purports to withhold lawful presence status in order to avoid granting rights or in order to grant the state with such a possibility in the future, then one interprets the term in bad faith. Still, the question would remain: what is the correct interpretation, in good faith?

[17] Minister for Immigration and Multicultural Affairs v Respondents S152/2003 (2004) 222 CLR 1 para. 53 (Austl.).

[18] Gardiner, supra note 2, at 215 (quoting US—Import Prohibition of Certain Shrimp and Shrimp Products, WTO Doc. WT/DS58/AB/R ¶ 17 (adopted Oct. 12, 1998) <> last accessed August 28, 2019).

[19] Convention Relating to the Status of Refugees art. 34, July 28, 1951, 189 U.N.T.S. 152 <> last accessed August 28, 2019 [hereinafter Refugee Convention].

[20] Id. at art. 1(C)(5).

[21] While the language of the “compelling reasons” exception refers only to statutory refugees (under Article 1(A)(1) of the Convention), it is a common understating that its spirit applied to “convention refugees” (under Article 1(A)(2)) too. See, e.g., United Nations High Comm’r for Refugees, Handbook and Guidelines on Procedures and Criteria for Determining Refugee Status ¶ 136 (2d ed. 1992, reissued 2019) <> last accessed July 16, 2019 [hereinafter UNHCR Handbook].

[22] Refugee Convention, supra note 19, at pmbl.

[23] See Vienna Convention, supra note 2, at art. 31(3)(c).

[24] See generally Rogers Brubaker, Citizenship and Nationhood in France and Germany 21-49 (1992); John Torpey, The Invention of the Passport 1-20 (2000).

[25] Hathaway, supra note 6, at 173-184, 658.

[26] Id. at 175.

[27] Hathaway, supra note 6, at 177.

[28] Id. at 178 (“where persons seeking recognition of refugee status meet the requirements of Art. 31… their presence must be deemed lawful”).

[29] Refugee Convention, supra note 19, at art. 31(1).

[30] The UNHCR Comments discuss the difference between “making” and “lodging” an asylum application. “UNHCR welcomes the distinction between the “making” of an application for international protection and the ‘lodging’ of the application. The “making” of an application does not involve any formalities, but simply refers to a wish expressed by a third country national or stateless person, who can be understood to seek international protection… UNHCR further welcomes the obligation for Member States to promptly register an application for international protection no later than three days from when it is made.” United Nations High Comm’r for Refugees, Comments on the European Commission’s Proposal for an Asylum Procedures Regulation – COM (2016) 467, 25 (2016) <> last accessed August 28, 2019. Similarly, Hathaway argues that an asylum seeker fulfills Article 31’s requirement to present himself before the host state’s authorities “without delay” by presenting himself before any officials, even the wrong ones (i.e. officials not belonging to the office dealing with asylum applications) and “advising them of his situation.” See Hathaway, supra note 6, at 390. For convenience sake, I will use hereunder the term “filing” to include both “making” and “lodging” of an asylum application, within the meaning given to these terms by UNHCR.

[31] According to the reasoning of the UK House of Lords in Regina v. Immigration Officer at Prague Airport and another ex parte European Roma Rights Centre and others [2004] UKHL 55 this does not seem to be in violation of VCLT Article 26 (performance in good faith). European Roma Rights Centre v. Immigration Officer at Prague Airport [2002] EWHC 1989 (Eng.) <> last accessed August 28, 2019; R (European Roma Rights Centre) v. Immigration Officer at Prague Airport, [2005] 2 AC 1 (Eng.) <> last accessed August 28, 2019.

[32] James C Hathaway, The Global Cop-Out on Refugees, 30 Int’l J. Refugee L. 591 (2019) <> last accessed August 28, 2019. The plan suggested in this article echoes the one presented in one of Hathaway’s earlier article. James C. Hathaway & R. Alexander Neve, Making International Refugee Law Relevant Again: A Proposal for Collectivized and Solution-Oriented Protection, 10 Harv. Hum. Rts. J. 115 (1997) <> last accessed August 28, 2019.

[33] Hathaway & Neve, supra note 31. (Hathaway and Neve purport to sketch a plan that would create cooperation and responsibility sharing among states and it would be odd if this sophisticated and calculated allocation of burden and responsibility could be circumvented by individual asylum seekers asylum “shopping.”).

[34] Hathaway, supra note 31, at 599.

[35] Ronald Dworkin, Taking Rights Seriously 15-45 (1977).

[36] This is a non-exhaustive list. As mentioned, these factors carry normative weight that should be balanced against the weight of opposing rules and principles. Not all of the enumerated factors have to be met in order for a compulsory secondary movement to be in accordance of Article 32, i.e. these are not necessary conditions, they are instead relevant considerations.

[37] I do not deal here with procedural safeguards that could be required as part of a review of the legality of compulsory second movement agreements, such as the transparency of the agreement.

[38] By “non persecutory,” I mean, at the minimum, a country which would not refoul the asylum seeker to his country of nationality, respects basic human rights, would allow the asylum seekers who enter its territory under a compulsory secondary movement agreement to file for asylum, and – if found eligible – be provided with the set of rights enumerated in Articles 2-34 of the Convention.

[39]In its 2018 position paper on the return or transfer to third countries, UNHCR repeatedly argues that “[r]equiring a connection between the refugee or asylum-seeker and the third state is not mandatory under international law” (a proposition I would not wholly disagree to, as I argue that existing ties to the third country is a key factor that must be taken into account, with other relevant considerations, while contemplating expulsion, but not a pre-condition for legal expulsion), yet concludes that “[i]n follow up to relevant conclusions of UNHCR’s Executive Committee, UNHCR though has consistently been advocating for a meaningful link or connection to exist that would make it reasonable and sustainable for a person to seek asylum in another state.” 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, UNHCR (2018) <> last accessed August 28, 2019. I think UNHCR’s position here is partly correct: this should not be a mere recommendation. Existing ties to the third country, while neither a necessary nor a sufficient condition to deem compulsory secondary movement plans legal, is a factor that must be considered and weighed with and against other considerations in the decision making process regarding such legality. Interestingly, and as a side point only, it should be mentioned that the Human Right Committee, in its General comment No. 36 (2018) on article 6 of the International Covenant on Civil and Political Rights, on the right to life, has referred to previous ties between a person and the country to which s/he is forcibly removed, as a relevant criterion, under certain circumstances, from the perspective of ICCPR Art. 6: “it would be inconsistent with article 6 … to deport an individual to an extremely violent country in which he has never lived, has no social or family contacts and cannot speak the local language.” Hum. Rgts. Committee, General Comment No. 36 (2018) on Article 6 of the International Covenant on Civil and Political Rights, on the Right to Life para. 30 (Oct. 30, 2018) <> last accessed August 28, 2019. In their global plan, Hathaway and Neve also suggest that refugees would preferably be resettled in the same region from which they originally came. Hathaway & Neve, supra note 31.

[40] In its 2018 position paper, supra note 38, UNHCR does not directly deal with the question of temporary versus permanent status in the “third country.” It is satisfied by saying that “[w]hen it has not yet been determined by the third state whether the person seeking protection is a refugee… upon return or transfer to the third country the person needs to be… authorized to remain in the country, until and unless a final negative determination of the asylum-seeker’s claim to refugee protection is rendered. During this time the asylum-seeker is lawfully in the state… As such, although the enjoyment of the right to self-employment under Article 18 of the 1951 Convention may be delayed for a limited period of time, it cannot be denied over the long-term because of government delays in the asylum procedures. Once it is determined the person is a refugee, s/he should be granted lawful stay and have access to the corresponding rights of the 1951 Convention….”

[41] In its 2018 position paper, supra note 38, UNHCR argues that “transfers to third countries should be aimed at enhancing burden- and responsibility-sharing and international/regional cooperation, and not be burden shifting.”

[42] Refugee Convention, supra note 19, at art. 17(2)(a).

[43] Id. at art. 7(2).

[44] Id. at art. 17(2)(b)-(c).

[45] Regulation (EU) 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 Stateless Person, 2013 O.J. (L 180/31) art. 29(1)-(2) <> last accessed August 28, 2019.

[46] See supra note 37. When a member state was found not to fulfill this requirement, the ECtHR justifiably forbad the expulsion of the asylum seeker. See MSS v. Belgium and Greece, Eur. Ct. H.R. (2011) <> last accessed August 28, 2019.

[47] This is a complex matter, which I will not delve into here. Many factors have to be taken into consideration, including the voluntary (i.e. not required according to Dublin Regulations) reception of over a million refugees, many of them Syrians, by Germany during 2014-2017. See Codrina Csesznek, Asylum in Germany During The New Exodus – A Secondary Analysis on Official Reports, 10 Bull. Transilvania Univ. of Braşov 141 (2017) <> last accessed August 28, 2019.

[48] Directive 2013/32, of the European Parliament and of the Council of 26 June 2013 on Common Procedures for Granting and Withdrawing International Protection, 2013 O.J. (L 180/60) [hereinafter EU Directive] <> last accessed August 28, 2019.

[49] Id. at art 31(3-5).

[50] Id. at art 38(2)(a).

[51] RR (Refugee – Safe Third Country) Syria v. Secretary of State for the Home Department [2010] UKUT 422 para. 11 (IAC) (Eng.); see also IH v. Secretary of State for the Home Department (Eritrea) [2009] UKAIT 00012 para. 22-24 (Eng.) (similar reliance on personal ties to the third country, and reference to the Asylum Procedure Directive).

[52] EU Directive, supra note 47, at art 38(1)(e)

[53] Note that Australia espoused, as elaborated below, quite extreme and harmful compulsory secondary movement plans despite its Court’s ruling that an asylum seeker who filed an asylum application is “lawfully present” in its territory. See supra note 7.

[54] For a good overview and discussion of Australian and Israeli compulsory secondary movement plans see Shani Bar Tuvia, Australian and Israeli Agreements for the Permanent Transfer of Refugees: Stretching Further the (Il)legality and (Im)morality of Western Externalization Policies, 30 Int’l J. Refugee L. 474 (2018) <> last accessed August 28, 2019.

[55]See id. at 475.

[56] Regional Resettlement Arrangement, Austl.-Papua N.G., July 19, 2013 <> last accessed August 28, 2019.

[57] Memorandum of Understanding, Relating to the Transfer to and Assessment of Persons in Nauru, and Related Issues, Nauru-Austl., Aug. 3, 2013 <> last accessed August 28, 2019.

[58] Memorandum of Understanding Relating to the Settlement of Refugees in Cambodia, Cambodia-Austl., Sep. 29, 2014 <> last accessed August 28, 2019.

[59] See Bar Tuvia, supra note 53, at 491-92.

[60] See, e.g., The Australia-Cambodia Refugee Deal, Kaldor Cent. for Int’l Refugee L. 4-5 (last updated Aug. 2018) <> last accessed August 28, 2019.

[61] Peter O’Neill, Prime Minister, Papua N.G., Joint Press Conference on Regional Resettlement Arrangement (July 19, 2013) <> last accessed August 28, 2019.

[62] UNHCR, Submission by the Office of the United Nations High Commissioner for Refugees on the Inquiry into the 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 16 (Nov. 12, 2016) <> last accessed August 28, 2019.

[63] For UNHCR’s criticism of Australia’s agreements with PNG and Nauru due to burden-shifting see id. For UNHCR’s similar criticism of Australia agreement with Cambodia see UNHCR, Submission by the United Nations High Commissioner for Refugees for the Office of the High Commissioner for Human Rights’ Compilation Report 1 (Dec. 2018) <> last accessed August 28, 2019 (“On September 26, 2014, the Governments of Cambodia and Australia reached an agreement to relocate refugees from Nauru to Cambodia, thus shifting Australia’s responsibility toward individuals who had sought the country’s protection to Cambodia, a country significantly less developed and less equipped to absorb those in need of international protection.”)

[64] Australia’s disregard of Article 32 is also reflected in Section 36(3) of its Migration Law, which provides, in very broad terms: “Australia is taken not to have protection obligations in respect of a non-citizen who has not taken all possible steps to avail himself or herself of a right to enter and reside in, whether temporarily or permanently and however that right arose or is expressed, any country apart from Australia…” Migration Act 1958 (Cth) (Austl.).

[65] Population & Immigr. Authority, Pol’y Plan. Dept., Foreign Citizens in Israel Report for 2016 (Apr. 2016), <> last accessed August 28, 2019.

[66] Reuven (Ruvi) Ziegler, No Asylum for ‘Infiltrators’: The Legal Predicament of Eritrean and Sudanese Nationals in Israel, 29 J. Immigr., Asylum and Nat’lity L. 172, 181 (2015) <> last accessed August 28, 2019.

[67] Christiano d’Orsi, To Stay or to Leave? The Unsolved Dilemma of the Eritrean Asylum-Seekers in Israel, 59 Harv. Int’l L.J. 83, 89 (2018) <> last accessed August 28, 2019.

[68] The Israeli Supreme Court ruled that the confidentiality was in accordance with Israeli law due to the risk that publication of the agreements would harm Israel’s foreign relations with each of these countries. See Admin. Req. 5164/15 Sagitta v. Minister of Interior (2015) (Nevo Legal Database (by subscription, in Hebrew) (Isr.). Israeli media did, however, openly refer to Uganda and Rwanda as the two countries with which the agreements were signed.

[69] Admin. Pet. 5126-07-15 Sagitta v. Minister of Interior (2015) (Nevo Legal Database (by subscription, in Hebrew) (Isr.).; Admin. App. 8101/15 Sagitta v. Minister of Interior (2017) (Nevo Legal Database (by subscription, in Hebrew) (Isr.).. For a summary of these decisions, see D’Orsi, supra note 67, at 93-98.

[70] Bar Tuvia argues that Article 32 was nonetheless relevant to this scenario, but I will not delve into this question. See Bar Tuvia, supra note 53, at 503-04.

[71] See Supreme Court’s decision in Saggita, supra note 69, at para. 70 to President Naor’s opinion.

[72] See D’Orsi, supra note 67, at 90.

[73] This is especially true in the case of treaties, such as the Refugee Convention, “where there have been many negotiators, sometimes with different groups working on different parts of the text, and sometimes using several languages, some of which may have a greater and more nuanced range of words on a particular topic than do other languages.” Gardiner, supra note 2, at 209.

[74] H.L.A. Hart, The Concept of Law 79-100 (2nd ed. 1994).

[75] If I fail to persuade the reader that this is a viable interpretation of the term lawful presence, it is still possible to make legal use of my insights as to the two subsidiary purposes of the Convention, and their relevance to judicial review of compulsory secondary movement plans. After all, such plans should not only meet the Convention’s requirements (i.e. Article 32) but also – at least according to some countries’ domestic law – constitutional and administrative law requirements. The question whether an asylum seeker’s right to dignity (under constitutional law), for example, is violated under a secondary movement plan could depend on the application of some of the criteria mentioned above. The criteria could be similarly used for the question whether the agreement, which was signed by state bureaucrats, is reasonable and proportionate (under administrative law).

[76] See, e.g., supra note 38-40, 50.

Suggested Citation: Yuval Livnat, Compulsory Secondary Movement and Article 32 of the Refugee Convention, RefLaw (August 28, 2019),


Submit A Comment

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>