Tag Archives: Refugee law

Understanding the ‘internal protection alternative’ (Part II)

Written by

by Jessica Schultz, Researcher, Chr. Michelsen Institute.

This is the second post in a two-part series on the internal protection alternative (IPA) based on Jessica Schultz ’s new book on the topic. The two blog posts were first posted on “The EU Immigration and Asylum Law and Policy” blog, and are re-posted here. The first post took Norway as a case study and this post follows up with a reflection on refugee law’s ‘surrogate’ role which states use to justify IPA practice.

The surrogate role of refugee law: a principle or preoccupation?

As described in the previous post, the ‘internal protection alternative’ (IPA) is a limit on refugee status used to exclude claimants with access to adequate protection somewhere within their countries of origin. For example, an Afghan who fled Taliban recruitment in Helmand province may be returned to Kabul if conditions are sufficiently safe there. In most jurisdictions, return must also be reasonable for the person concerned. 

Although there is no mention of the IPA in the 1951 Refugee Convention or its 1967 Protocol, this implied limit spread in state practice during the mid-1980s as a response to several factors, including the prominence of non-state persecution as a basis for refugee claims. Today, the concept has firmly taken root in many jurisdictions despite continued debate about the exact criteria for IPA application. 

The IPA is also sometimes applied to justify decisions relating to the cessation of refugee status when some area of the country of origin is deemed secure enough to accommodate returning refugees. The migration of the IPA concept from the inclusion to the cessation analysis is not broadly endorsed (see UNHCR’s position and UK jurisprudence), but there is little doubt that states are increasingly interested in identifying some safe space – anywhere – to which refugees or refugee claimants may return within their countries of origin. 

My book analyzes various ways that states and scholars have framed the IPA limit in refugee law. This post examines a common element of these efforts, the observation that international protection is ‘surrogate’, or subsidiary, to national protection provided by countries to their own citizens. As Zimmermann and Mahler explain, ‘the evolution of the [IPA] concept draws heavily on the notion of surrogacy as a basic principle of refugee law, according to which international protection only comes into play when national protection within the country of origin is not available.’ If national protection exists somewhere, the back-up remedy of protection abroad is not (or no longer) needed. 

The surrogate role of refugee law: ‘thin’ versus ‘thick’ perspectives

As Goodwin-Gill and McAdam have written, the notion of ‘surrogacy’ usefully describes the overall role of refugee law. The 1951 Convention obliges state parties to provide replacement protection, when a person can demonstrate a well-founded fear of persecution within their country of ‘origin’ (citizenship or previous residence in the case of stateless persons) (Article 1A(2)). Beyond the inclusion provision, Article 1 also contains criteria for exclusion and cessation of refugee status. These too confirm the subsidiary, or surrogate, character of Convention protection. For example, Articles 1A(2) para 2, 1C(5), 1C(6), and 1E all indicate that when there is no well-founded fear of persecution within the country of origin or another country where the claimant has a national connection, refugee status need not be recognized. However, neither the text nor other sources (including the extensive drafting history, as discussed in the book) suggest a similar limit for persons for whom protection is available in only part of a country. 

It may be helpful, then, to distinguish between a ‘thin’ surrogacy perspective and the ‘thick’ one that is sometimes referred to as a ‘principle’ of refugee law.  According to a ‘thin’ surrogacy perspective, the lack of protection by the state of origin is relevant insofar as it negates the well-founded nature of a claimant’s fear. This absence of protection is a characteristic of a person who meets the Convention criteria; it is not a condition of refugee status. A fear of persecution is well-founded because the state has failed to ensure a minimum level of security to suppress a risk of serious harm. The IPA is not compelled by this interpretation of refugee law, although there may be cases in which the claimant’s unwillingness to avail him or herself of home state protection, despitea well-founded fear, cannot be reasonably justified. 

According to a ‘thick’ surrogacy perspective, meanwhile, state protection comes into play at two separate stages in the assessment of refugee status. First, serious harms committed by non-state actors qualify as persecution for a Convention reason when the state is unable or unwilling to protect the claimant. Second, the possibility of protection elsewhere in the country of origin must be considered to establish the necessity of protection abroad. In other words, protection has a systemic aspect (related to the state’s ability and willingness to protect from the original harm) as well as a territorial one (related to the possibility of protection elsewhere). 

There is nothing in the structure of the Convention to suggest that the presence of protection somewhere in a state’s territory can defeat a claim to refugee status. Article 1A(2) of the Refugee Convention defines a refugee as someone who ‘owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable, or owing to such fear, is unwilling to avail himself of the protection of that country….’

The criteria for refugee status are satisfied, in other words, when the claimant 1) is outside his or her country of origin; 2) has a legitimate fear of persecution for a Convention ground; and 3) is unable or unwilling, owing to the well-founded fear, to avail him or herself of that country’s protection. The ‘thick surrogacy’ perspective shifts focus from the claimant’s ability or willingness to avail herself of home state protection to the ability or willingness of the state to provide it (within the country of origin).  

Despite the fact that a ‘thin’ surrogacy perspective aligns better with the text of the treaty, a ‘thick’ surrogacy perspective has gained traction in the scholarly discourse as well as in national jurisprudence. Below are some of the factors – by no means exhaustive – that help explain this development.

Factor 1: The influence of human rights law and principles on interpretation of the refugee concept

One explanation for the persistent power of ‘surrogacy’ as a justification for limiting the scope of refugee status relates to the misuse—in my view—of human rights law and principles for the purpose of interpreting the Refugee Convention. In terms of protection criteria, non-refoulement cases decided by human rights courts have influenced states’ interpretation of the 1951 Convention’s refugee concept. These do not frame the IPA as a limit on the scope of refugee status but rather as an extension of the basic risk analysis covering conditions on return. 

Procedural concepts from the human rights field have also infiltrated substantive interpretations of refugee law. Some sources refer to the duty to ‘exhaust domestic remedies’ in support of IPA practice. The ‘exhaustion of local remedies’ rule applies in human rights law to preserve the subsidiary nature of supranational decision-making bodies vis-à-vis a more democratic local organ. Despite its authoritative ring, it makes no sense in the context of refugee claims. First, the rule is usually concerned with systemic procedures (has the case been heard by the state’s highest court?) and not with the absence of protection by local state agents in certain areas. It also involves a backward-looking analysis (what has been done?) in contrast to the prospective inquiry required to assess the need for protection.  

Factor 2: The influence of other disciplines in the interpretation of the refugee concept

A second explanation for the prominence of the surrogacy discourse relates to broader developments in the field of refugee studies. In recent decades, important contributions to the refugee concept from social scientists and philosophers have filtered into legal debates. With some exceptions, these scholars (see here and here) argue that international protection should extend beyond the confines of the Geneva Convention to include persons forced to flee their countries of origin for reasons that may or may not have a Convention nexus.  

For example, in his influential article ‘Who is a Refugee?’ (1985), Shacknove posits the following definition: 

[R]efugees are, in essence, persons whose basic needs are unprotected by their country of origin, who have no remaining recourse other than to seek international restitution of their needs, and who are so situated that international assistance is possible. 

Despite the fact that Shacknove’s contribution to the field of forced migration was never meant as a restatement of refugee law, legal scholars like Hathawayand Storey have cited him in support of their view that the possibility of protection somewhere in the country of origin may, under certain conditions, defeat a claim for refugee status. Scholarly contributions, even those motivated by the desire to expand the refugee concept,  have reinforced the view that a deserving claimant is one whose compelling needs arise from the absence of adomestic alternative

Factor 3: The self-referential nature of norm development in refugee law

Without a treaty monitoring body to oversee the Convention, the situation arises in which certain ideas take on a life of their own through the interplay of jurisprudence, scholarship and policy. The concept of surrogacy has undoubtedly gained momentum through the inter-jurisdictional ‘dialogue’ in the field of refugee law and the influence of leading scholars. Its expression, however, has also been marred by circular reasoning. For example, the Canadian Supreme Court in Ward refers to the first edition (1991) of The Law of Refugee Status of Hathaway to explain the surrogate role of refugee law. Although the legal question was about whether non-state persecution was covered by the Convention’s refugee concept (and not about a potential IPA), this case appears regularly in IPA literature. Indeed, authors of the second edition of The Law of Refugee Status cite Ward as evidence of positions proposed in the previous book. 

Factor 4: Harmonization impulses versus the Refugee Convention as ‘lex specialis’

As most states operate with multiple regimes of international protection, the ‘surrogacy principle’ can be leveraged to justify the IPA as an unwritten exception to refugee status no matter what its treaty basis may be. For example, Article 8 of the EU Qualification Directive (2011) provides in general terms that internal protection may be considered ‘(a)s part of the assessment of the application for international protection’. By framing the IPA in this way, it is easier for states to justify a common IPA test in claims to Convention refugee status as well as those that relate to complementary forms of protection. As discussed elsewhere, the consequence has been that human rights courts like the ECHR are increasingly setting the standards for interpretation of the Refugee Convention – also for IPA practice – rather than the other way around.

Implications of a ‘thin’ surrogacy perspective for IPA practice

While the Refugee Convention’s purpose is to provide substitute protection, recognition as a refugee does not depend on the absence of a domestic alternative.  Nonetheless, an IPA limit may apply under certain conditions: when a claimant can relocate within the country of origin with minimal negative impact (from either an objective human rights perspective or a more subjective, humanitarian one). In these cases, extending international protection would erode the treaty’s effectiveness. In addition to individual factors, sending states must consider structural ones: whether return would reinforce policies or practices of ethnic cleansing, or stress already fragile areas. Both dimensions require a displacement-sensitive analysis. 

The scope for IPA practice when revoking refugee status is even narrower. As UNHCR has rightly maintained, where a risk of persecution persists in one region of a country, it is unlikely that the changed circumstances are ‘profound’ and ‘enduring’ as required under the cessation analysis. There are also situations in which circumstances have objectively changed but return is still unreasonable due to past experience of persecution.  

The IPA question goes to the heart of what it means to be a refugee. Is a well-founded fear of persecution within the country of origin enough to establish a refugee claim? While persecution – especially by non-state actors – may not always justify the claimant’s recourse to refuge abroad, the presumption that it could should not be diluted. The surrogacy concept in many jurisdictions has done just that, by shifting the focus of refugee status determination from the risk of persecution to the possibility of return in line with minimal human rights standards. This not only downplays the relevance of refugee experience, but it also renders the unique protection issues attached to internal displacement – a consequence of IPA practice –  largely invisible.

Understanding the ‘internal protection alternative’ (Part I)

This is the first post in a two-part series on the internal protection alternative (IPA) based on Jessica Schultz ’s new book on the topic. The two blog posts were first posted on “The EU Immigration and Asylum Law and Policy” blog, and are re-posted here. It starts with a case study on Norway and follows up with a post reflecting on refugee law’s ‘surrogate’ role which states use to justify IPA practice.

By: Jessica Schultz, Researcher and Senior Adviser, CMI

A post mortem on the demise of the reasonableness requirement: The IPA in Norway

It might surprise some readers that Norway, normally viewed as a human rights stalwart, is at the forefront of efforts to push the boundaries of refugee law in a restrictive direction. Like other European States, Norway responded to the influx of refugee claims in 2015 with a barrage of policies intended to deter and divert refugee flows. Border controls, safe third country transfers, time limits on residence, and restricted family reunification were among the measures adopted to ensure that Norway’s policies at a minimum were not more generous than those of its neighbors.  

In one area, however, Norway’s restrictions surpassed those of other states: it lowered the threshold for applying the ‘internal protection alternative’ (IPA) as a basis for denying refugee claims. IPA practice is premised on the view that refugee law comes into play when the claimant’s country of origin cannot or will not provide protection itself. If a domestic alternative to asylum abroad is accessible, safe, and reasonable, UNHCR and many states accept that a refugee claim may be refused.

Following amendments to the Immigration Act passed in 2016, this last condition, that relocation is ‘reasonable’, no longer applies. In the government’s view, the principle of non-refoulement only requires that protection against persecution is available in a return area. If it is, refugee status need not be recognized – no matter how harsh the consequences may be. Only one other jurisdiction – Australia – excludes reasonableness from the IPA assessment.

For reasons described here, ‘reasonableness’ (or proportionality) is widely-recognized as a legal requirement for application of the IPA limit. So what explains Norway’s outlier position? This post reviews the historical and political roots of Norway’s current IPA practice, including the claim that the right to refugee status is subject to a degree of state discretion. I will also discuss, as an example, the consequence of Norway’s position for unaccompanied Afghan minors and implications for other areas of refugee law.

Roots of the reasonableness test in Norway

As with other states in Northern Europe, IPA practice in Norway gained momentum in the 1990s, and evolved largely in response to claims of persecution by non-state actors. Consideration of the IPA in these early years was exceptional and informal in nature, and justified with reference to paragraph 91 of UNHCR’s 1979 Handbook. Although the 1988 Immigration Act made no mention of an IPA limit, the Ministry of Justice’s Asylum Guidelines in 1998 formally addressed, for the first time, the concept’s relation to refugee status:

In cases where the applicant will be threatened by non-state groups or individuals in certain areas of the home country, protection in Norway (either in the form of asylum or a residence permit) is normally refused if he or she will be secured protection in other (for example government-controlled) areas of the home country.

The Guidelines offered an exception when, ‘after a holistic assessment of all aspects (health issues, impact on children, links to Norway), there may be cases in which the claimant should not be compelled to relocate elsewhere in the home country despite the possibility of securing protection there.’ Notably, the ‘aspects’ mentioned depart from the ‘reasonableness’ criteria set out by UNHCR. Instead, they refer back to a separate provision of the Immigration Act concerning residence on humanitarian grounds.

From the beginning, then, the reasonableness test was deemed a matter of state discretion, to be linked to whatever criteria domestic authorities deemed to be most compelling.  The consequence was an overly narrow reasonableness assessment (excluding issues like the right to education, freedom of religion and past persecution) and a lower standard of judicial review.

Drafters of the 2008 Immigration Act aimed to realign the reasonableness test with UNHCR’s Guidelines. The Immigration Regulations that followed, however, reasserted the link between the reasonableness assessment and criteria for residence on strong humanitarian grounds. Jurisprudence remained split on the proper reference point until the issue was finally brought to the Norwegian Supreme Court in 2015.

The Supreme Court’s Internal Flight judgment

The Internal Flight case involved an Afghan family refused asylum on the basis of an IPA in Kabul. The parents were originally from Ghazni province, but had spent many years in Iran where their two daughters were born. The Board of Immigration Appeals (UNE) had concluded that their claim for asylum under the Refugee Convention was not credible, but that the family was nonetheless protected on grounds of the security situation from return to their area of origin.  

The question was then: could the family safely and reasonably relocate to another part of Afghanistan? The claimants argued that the IPA test should be interpreted in line with UNCHR´s guidance, in accordance with the intention of lawmakers. By linking the reasonableness criteria with discretionary factors instead, the Immigration Regulations overstepped their statutory basis. The Court, however, declined to rule directly on this issue. Instead, it simply confirmed that the Immigration Regulations, and the specific interpretation they codify, have a legal basis in the Immigration Act.

The Court’s refusal to address the actual criteria reflects a belief that reasonableness is not integral to the IPA concept. Why? One clue is found in Judge Utgård’s opinion, where he harkened back to the Supreme Court’s Abdi judgment from 1991. In that case, involving a sur place claim arising from the person’s voluntary activities in Norway, the Court distinguished between core areas covered by the Convention and periphery issues belonging to a state’s discretion. The subjective sur place problem occupied this peripheral zone: although Abdi was protected from refoulement, he could still be refused refugee status.

Referring to the Abdi judgment, Utgård wrote that the state has ‘broad liberty’ to regulate who has the right to refugee status in Norway. In Utgård’s view, the parameters of non-refoulement regulated by Article 33 (1) of the Geneva Convention only require that the ‘return area is accessible and safe.’ Considerations of reasonableness, on the other hand, occupy a peripheral space that can be regulated as the State sees fit. Even though Utgård’s position was obiter dictum, it was picked up by the Ministry of Justice and Security in its proposal not long afterwards to remove the reasonable conditions from the IPA test: ‘(t)he assessment here is linked to a core area for the Convention, which is protection against return to an area where the foreigner has a well-founded fear of persecution’ (emphasis added).

The ‘refugee crisis’ and removal of the reasonableness requirement in IPA practice

This proposal came as part of a package of measures announced in December 2015. According to the Ministry, the reasonableness test was essentially problematic: it had unclear scope and content; it opened for discretionary assessments that were difficult to structure; and it lead to unequal treatment of similar cases. Furthermore, the Ministry curiously claimed, ‘it is undisputed that international law does not require states to operate with the reasonableness criteria.’ In support of this statement it referred to Utgård’s minority opinion and incorrectly cited Professor Zimmermann´s well-known Commentary on the Refugee Convention. The Ministry also wrote that the ‘reasonableness’ requirement in the IPA provision of the EU Qualification Directive (Article 8) referred only to the extreme humanitarian conditions which have anyway been read into Article 3 ECHR by the ECtHR. In reality, Article 3 jurisprudence doesn’t even capture the requirements of ‘effective protection’ much less reasonableness for IPA purposes.  

Parliament approved the proposed amendment, which came into effect on October 1, 2016. The current IPA provision states that:

“[t]he right to be recognized as a refugee according to paragraph 1 does not pertain if the foreigner can receive effective protection in other parts of the country of origin than that area from which the claimant has fled”.

Consequences for refugee claimants: the case of Afghan minors

It is hard to measure the impact of the change in IPA practice on rates of recognition in Norway. One reason is that the IPA is often used as a subsidiary reason for refusing refugee status, when other aspects of the claim are unclear. Decisions typically reason that ‘even if’ the claimant is telling the truth, or the risk of persecution indeed exists, he or she could still safely relocate to a city or region within their country of origin. Therefore, statistics on the formal grounds for rejection do not capture the influence of IPA reasoning.

We do know, however, that changes to IPA practice has affected the rates of refugee status for some vulnerable groups. Families with children, single women, persons with serious illnesses and others are no longer recognised as refugees  because return to internal displacement would be unreasonable. Instead, if they are lucky, they receive a more contingent leave to remain for humanitarian reasons. The IPA rules have also affected recognition rates for unaccompanied minors (UAMs), most of whom come from Afghanistan.  Before 2016, UAMs were exempt from IPA practice since the absence of a caregiver would automatically render return ‘unreasonable’. This is no longer the case. Removal of the reasonableness requirement has resulted in the expanded use of temporary residence visas that expire at the age of 18. At that point these youths may be returned to a city (Kabul) increasingly recognized as profoundly unsafe and to a country those born in Iran or Pakistan have never even lived in.

Following a regulation change earlier in 2018 aimed at softening these harsh effects, decision-makers were instructed to review these cases to consider, among other things, whether the minor would have a network and/or resources to get along in Kabul.  These vulnerability criteria covered only a fraction of the factors relevant to a reasonableness analysis. Even so, the Immigration Directorate determined that less than half of the youths who applied met them. Many others, living precariously in Paris and elsewhere, did not meet the deadline for having their claim reconsidered.

Consequences for other dimensions of refugee law: cessation of refugee status

The concept of a refugee set out in the 1951 Refugee Convention is being squeezed not only in terms of its spatial dimension, but also its temporal one. As the Ministry of Justice reminds us, ‘international protection is subsidiary to protection in one’s own country’. In the next post, I will unpack this claim. For the time being, however, it begs the question: if refugee status can be refused on the basis of an IPA, can it also be revoked when an IPA becomes available?  In Norwegian practice, the answer appears to be positive.

In the view of the Ministry of Justice, the need for protection no longer exists when some area of the home country is safe. It has argued that implementing the IPA in these cessation cases ensures ‘equal treatment’ for all refugees from the same country, no matter what part they come from. This position not only conflates return to one’s previous residence with prolonged (domestic) displacement, but it diverges from requirements under the Refugee Convention. Article 1C (5) permits states to withdraw refugee status if, among other things, circumstances that gave rise to that status no longer exist. As the  UNHCR explains, “the changed situation must address the causes of displacement. Further, changes must be fundamental in nature, so that the refugee ‘can no longer…continue to refuse’ home state protection”. Referral to an IPA undermines both these guarantees.

Conclusion

In Norwegian practice, the focus of asylum authorities is not on the risk of persecution but on the possibility of protection somewhere, no matter how unreasonable the consequences are for the claimant. Even the threshold of  ‘effective protection’ is undermined by narrow interpretations of who can provide it, how long it may last and how big the area in which it exists needs to  be. The dynamics set in motion in 2015 create a dangerous precedent in a region where national authorities are anxious to exploit all possible arguments for refusing claims to refugee status.