The end of protection? Cessation and the ‘return turn’ in refugee law

This text first appeared on the blog of the Odysseus Academic Network and is re-posted here. Jessica Schultz is a Post-Doctoral researcher at the University of Bergen Faculty of Law, and a Researcher and Senior Adviser at the Chr. Michelsen Institute.

Photo: Marienna Pope-Weidemann via Flickr

With increasing enthusiasm, European states are reviving the Refugee Convention’s cessation provisions in service of their return-oriented refugee policies. This practice threatens the careful balance established by refugee law between the security of refugee status, on the one hand, and its impermanence on the other.

This post reviews the legal requirements for cessation of refugee status as well as how the focus on return distorts their application. Through the lens of Norwegian practice, it is possible to see how reliance on an internal protection alternative (IPA) and non-state actors of protection dilute the requirement of durable protection, especially for women and children.

As cessation practice becomes a more normalized part of refuge administration in Europe and elsewhere, it is essential that decision-makers uphold the distinction between protection from return or non-refoulement, and cessation of refugee status. As described below, cessation application requires authorities to consider factors beyond the risk of persecution or serious harm.

The context: the ‘return turn’ in refugee law

Containment measures like pushbacks, extraterritorial processing and safe third country regimes dominate public debates concerning refugees. In contrast, control efforts that reduce the length of stay in host countries have not received the same scrutiny. In recent years, states have decreased the durability of residence permits, enhanced scrutiny of refugees’ continued need for protection, and introduced new requirements for gaining permanent residence. While traditional temporary protection policies have operated in parallel to Refugee Convention protection, these newer measures affect those with Convention status as well.

Application of the cessation clauses contained in Articles 1C(5) and (6) of the Convention is an important example of how the Refugee Convention is being leveraged to advance states’ restrictive agendas.  These provisions permit termination of refugee status when the refugee …’ can no longer, because the circumstances in connection with which he has been recognized as a refugee have ceased to exist, continue to refuse to avail himself of the protection of the country of his nationality’ or, in the case of a stateless person, the country of previous residence.

While cessation has always been a possibility, the high legal threshold and administrative costs are among the reasons why most states (Germany and Australia are exceptions), have refrained from an active cessation practice when it comes to individual cases. Instead, Convention refugees have typically received a permanent residence permit either immediately or at least predictably within a relatively brief period.

This commitment to provide a stable refugee status is clearly eroding. CanadaDenmark, and Sweden have all adopted policies involving the proactive review of a refugee’s continued need for protection. In Denmark, the threat of cessation even applies to resettled refugees, who already, in theory, have a durable solution. The draft EU Qualification Regulation would introduce mandatory reviews of refugee status to be conducted in connection with the renewal of residence permits for the first time for Convention refugees (after 3 years) and for the first two times for beneficiaries of subsidiary protection (i.e. after one year and again after 3 years).

As UNHCR rightly points out, there is a danger that such reviews could slide into a ‘cessation light analysis’ where the only issue assessed is whether the risk of persecution or serious harm still exists. This ‘mirror approach’ premises cessation on the absence of conditions giving rise to refugee status, and nothing more.

Cessation in refugee law: balancing two kinds of ‘refugee time’

So what exactly is the problem? Besides the fact that refugees face a risk of premature (unsafe) return and renewed displacement, the ‘mirror approach’ reflects a distorted reading of the Refugee Convention.

As Durieux has eloquently observed, the Convention carefully balances two dimensions of ‘refugee time’. The first relates to the basket of rights and benefits that accrue as the refugee’s attachment to the host state increases. (Articles 3-34). While the Convention stops short of requiring states to naturalize refugees, this outcome is explicitly encouraged in Article 34. The second relates to the end of refugee status. While the logical consequence of long-term stay is naturalization, Article 1(C) of the Convention clearly permits the termination of status before this takes place. The two dimensions of refugee time create an internal tension between opportunities to settle in a new community and the possibility of being forced to return to one’s country of origin.

In light of the Convention’s object and purpose to provide stable, if not permanent, status, cessation must be carefully and narrowly applied in situations where durable protection is clearly in place. From a policy perspective, the requirement of durability also ensures the alignment of cessation practice with the broader solution-oriented aims of refugee protection, and commitments under the Global Compact on Refugees.

That the cessation analysis grounded in Articles 1C (5) and (6) requires more than simply the absence of persecution is also clear from the text, which refers to ‘circumstances in connection with the grant of refugee status.’ This suggests that decision-makers must determine whether, in light of changes in the country of origin, the refugee’s continued refusal to accept protection from the country of origin is still justified. As the UK House of Lords explained, ‘double guarantee’ (no risk of persecution plus adequate protection) is necessary to displace the refugee’s ‘assurance of a secure future in the host country’.

To guard against premature application, UNHCR and many jurisdictions agree that changes must be ‘fundamental, stable and durable’ – or a variation on these criteria. While dramatic events like a democratic transition can occur quickly, many of today’s refugee-producing countries are beset by complex conflicts where the consolidation of peace is a halting, difficult process.

Importantly, changes that only affect part of a country may not be ‘fundamental’ enough to warrant cessation of status. Partial cessation is particularly problematic when it involves the internal relocation of returning refugees (because the country is generally unsafe). In these cases, not only is the change not fundamental, but it is unlikely to be ‘stable and durable’ for those who are internally displaced upon their ‘return’.

The ‘return turn’ in Norwegian refugee law post-2015

Like other countries in Europe, Norwegian authorities responded to the influx of refugee claims in 2015 with policies seeking to reinforce the temporary nature of refugee protection. These included amendments to the Immigration Act increasing the time period for applying for permanent residence, and the introduction of non-protection related requirements for permanent residence including proof of economic self-sufficiency, local knowledge and language skills. In addition, changes to Internal Protection Alternative (IPA) criteria led to an increased scope for granting a time-limited, non-renewable residence permit for unaccompanied minors.

In March 2016, the Ministry of Justice and Security issued an Instruction to the Immigration Directorate (Utlendingsdirektoratet, UDI) on cessation of refugee status and revocation of residence permits. The Instruction requires that immigration authorities apply the cessation provisions in domestic law when 1) the protection needs for larger groups of refugees have fallen away or 2) there are indications in an individual case that cessation might apply. In addition, UDI must consider cessation and eventual revocation when processing an application for permanent residence.

So far in Norway, Somali nationals have been most affected by cessation practice as well as revocation for other reasons, including fraud or administrative error. In 2017, the Grand Board of Immigration Appeals held that the cessation analysis required a ‘margin of security’, meaning that the grounds for refugee status disappeared and that changes appeared to be durable. When applying this standard to the facts in Mogadishu, a majority found that despite the state’s inability to enforce the rule of law, adequate protection could be secured from the strong clan system. Therefore, refugees from Mogadishu could be safely returned.  Somali refugees continue to be the group most affected by cessation practice and the revocation of residence permits more broadly.

The Norwegian Supreme Court: precarious protection by private actors in unstable places

In 2018, the Norwegian Supreme Court decided two cases involving the cessation of refugee status and subsequent revocation of a residence permit. Both plaintiffs were Afghans granted refugee status at least in part as single women without male protection. These judgments provide more insight into interpretive dilemmas attached to cessation criteria, and their consequences for women and children especially. Here I will only discuss the first judgment, from March 2018, which clarified – in part – the legal criteria for applying the cessation provisions in Norway’s Immigration Act.

In this case, the applicant (called “A” in the judgment) and her daughter Farida had come to Norway in 2011, after being separated from their husband and father in Greece. They received refugee status because their home area in Jaghuri district was insecure and, as a single woman without male protection, A lacked an ‘internal protection alternative’ (IPA) in a safer part of the country. When A’s husband arrived in Norway, his asylum application was denied and A received notice of cessation and revocation of her right to residence.

The question before the Supreme Court was: does cessation simply require that the family no longer fulfilled the criteria for refugee status, as the Norwegian state argued? Or, consistent with the ‘double guarantee’ advocated by UNHCR and others, did it also require a showing of adequate protection on the part of the Afghan government?

Since the Immigration Appeals Board (Utlendingsnemnda, UNE) decision was based on the fact that Jaghuri was no longer unsafe, the Court found it in error for not considering whether the change in security had stabilized to a certain degree. Importantly, the Court confirmed that the criteria for cessation are not a direct mirror of the criteria for refugee status. The imperative to assess the durability of change arises from a refugee’s legal position, which confers a certain degree of security. The Court also obliquely referred to policy concerns by noting that foreigners who have adjusted themselves to residence in Norway should not be exposed to a return that can easily lead to further displacement and a new claim to refugee status.

A victory for refugee law, then, right? Not necessarily, because the decision left too many questions unanswered. What does durable protection really require? Does it depend on a functioning state apparatus? If Jaghuri was found to be an oasis of safety surrounded by insecure areas would cessation criteria be met? Does it matter that the plaintiffs are originally from there? Does the fact that A and her husband are together constitute ‘protection’ for the purpose of cessation?

This was indeed the conclusion of the Norwegian Supreme Court in another revocation case from November 2018. Although the main question here was whether, post-cessation, revocation of a residence permit would violate the applicants’ right to private life under Article 8 ECHR, the Court endorsed the decision to cease refugee status. No gender or indeed displacement-sensitive assessment of whether protection by a male relative is adequate or durable for the long term was made.

And last but not least: what does a child-sensitive cessation practice look like? Children’s best interests must be a fundamental consideration in all decisions affecting them under the Convention on the Rights of the Child. But in the Farida case, the Supreme Court failed to address this fact in its assessment of cessation criteria.

The fates of Farida and her mother A are still unresolved. After being remanded to the Immigration Appeals Board, which based its new decision on the possibility of an internal protection alternative (IPA) in Kabul, the case was appealed to Oslo District Court. The District Court held, in a mystifying judgment, that while the requirement of a ‘fundamental and stable’ change applies to the claimants’ home area, it does not apply to other areas where they could be returned. In those places, the test is simply whether a well-founded fear of persecution or real risk of ill-treatment contrary to Article 3 ECHR is currently present. In other words, the Court identified a broader scope for cessation of refugee status if ‘return’ is not to the place of previous residence.

Applying this test to the facts, the District Court found that conditions in Kabul, while certainly unstable, did not meet the threshold of persecution or serious harm under Article 3 ECHR.  If the case reaches the Supreme Court again, the Court must clarify whether the IPA and cessation concepts can be reconciled at all, and if so under what conditions. Evidence of a male protector in an IPA is not sufficient to displace the stability Farida and her mother had received through refugee status.

Three concerns: actors of protection, spaces of protection, and the passage of time

There are three interrelated issues that threaten principled application of the cessation provisions.  The first is the willingness of decision-makers to endorse non-state actors of protection for the purpose of ceasing refugee status. Clans or the institution of marriage cannot replace the durable protection provided by state or state-like authorities. The second is the seeming acceptance of the IPA as part of the cessation analysis. It is especially worrying that as the lack of an IPA is increasingly perceived as a condition of refugee status then the existence of an IPA may justify its cessation, despite lip-service to the need for fundamental and durable change. ‘Return’ to internal displacement is profoundly at odds with the imperative to promote solutions and is especially problematic for female refugees made dependent on male relatives for mobility as well as protection.

And finally, the drawn-out nature of individual cessation practice, including the appeals process, increases tension between opportunities to settle in the country of refuge and the possibility of eventual return. This has serious consequences for the well-being of individual refugees and particularly children, who may have spent their formative years in a country before cessation kicks in. Even when the ‘best interests of the child’ are assessed, they may not succeed in outweighing the state’s immigration control interests.

The protection gaps that arise from the failure to correctly implement cessation criteria create a risk of unsafe and unsustainable refugee returns. The possibility of protection somewhere else does not, on its own, justify ending it in the country of refuge.