IMMINENCE IN REFUGEE AND HUMAN RIGHTS LAW: A MISPLACED NOTION FOR INTERNATIONAL PROTECTION

Abstract

This article is an output of a major research project examining the notion of imminence in the law on international protection. It is the first piece of scholarship to identify an emerging trend, namely the introduction of imminence—whether invoked implicitly or explicitly—as a potential barrier to refugee status or complementary protection. The article analyses the jurisprudence of relevant international bodies and courts and critiques the validity of this notion as a tool for assessing States’ protection obligations.

I. INTRODUCTION

Time is everywhere in refugee and human rights law.¹ But in assessing whether someone faces a real chance of being persecuted or subjected to other serious harm if removed, the immediacy or ‘imminence’ of that harm should not be the basis on which protection is granted or denied.² A decision-maker who took such an approach would fall into legal error.³

We are struck, however, by several types of protection cases in which the nearness in time of harm seemed to play a role, such as those relating to the (future) impacts of climate change and to deterioration of health over time. We are curious about why, and how, the ‘imminence’ of harm was factored into decision-making in this context—either implicitly or explicitly. What was common to these cases was that the alleged harm was not felt acutely now, but, it was argued, it would have deleterious consequences in the future.

Decision-makers may have imported the notion of imminence, understood in terms of time (when harm may occur), from other areas of international law, such as the law of self-defence.⁴ Yet, even in that context, there is ‘little scholarly consensus on what is properly meant by “imminence” in the context of contemporary threats’.⁵ Indeed, we are surprised by how little jurisprudential discussion there has been of the notion, finding that it has instead been promoted through academic debate rather than legal instruments or case law. In the ‘protection of civilians’/peacekeeping context, where the United Nations (UN) Security Council has traditionally directed peacekeepers to protect civilians from ‘imminent threat of physical violence’⁶—by ‘us[ing] force to protect civilians who were on the brink of being harmed’⁷—the notion has again proven to be problematic and poorly understood,⁸ at times creating confusion about when to act. For this reason, ‘imminence’ is now rarely invoked in that context.⁹ Our concern is that an already unclear notion may be unwittingly transposed into the law on international protection, resulting in poor decision-making, legal error, and, ultimately, negative outcomes for refugees and others in need of protection.

For these reasons, this article engages with issues that go well beyond academic inquiry alone: they have immediate and concrete consequences for people whose protection claims are rejected because of a lack of legal clarity about the relevance, nature, and scope of the notion of imminence. A series of cases in New Zealand has already started to delineate the potential of refugee and human rights law to protect people displaced from small Pacific Island countries by the impacts of climate change, disasters, and environmental degradation.¹⁰ While no one has yet been granted protection on these grounds, this jurisprudence provides the most comprehensive analysis of the issues to date. ‘Imminence’ has been discussed specifically in this context, with the Immigration and Protection Tribunal (Tribunal) observing (rightly, in our view) that it imposes no higher threshold than the well-founded fear test in refugee law, and requires ‘no more than sufficient evidence to establish substantial grounds for believing the appellant would be in danger’.¹¹ However, in other jurisdictions, imminence has operated inappropriately as an additional hurdle for people seeking international protection.¹²

This article is the product of a project examining the notion of imminence in the protection of refugees and other forced migrants. That project is based on the hypothesis that the notion has been introduced haphazardly, and virtually unnoticed, into refugee and human rights law, with no legal or conceptual foundation. While at times decision-makers’ use of the term may simply be loose language that has no particular import in the decision itself, at other times, it is being used to deny protection. Furthermore, there is a danger that loose language may risk the notion being used more substantively.

Beyond its scholarly contribution, we hope that this conceptual article will help to sharpen the focus of decision-makers when considering issues of time and risks of harm in protection cases, and the implications of using the notion of imminence in deciding whether or not individuals require international protection.

II. SCOPE AND METHODOLOGY

The Refugee Convention remains the ‘cornerstone of the international refugee protection regime’.¹³ However, due to its relatively restrictive definition—which requires a person to establish a ‘well-founded fear’ of being persecuted for reasons of his or her race, religion, nationality, political opinion, or membership of a particular social group—it has been supplemented by international and regional human rights law, which has stepped in to fill certain gaps by protecting people against return to a ‘real risk’ of (inter alia) being subjected to torture or other cruel, inhuman or degrading treatment or punishment, or being arbitrarily deprived of their life (known as complementary protection).¹⁴

This article begins by exploring whether the evidentiary standards relevant to assessing claims for international protection, namely ‘well-founded fear’ and ‘real risk’, inherently contain an imminence component. It then examines jurisprudence from international and supranational courts and bodies assessing claims for international protection in which the notion of imminence has formed part of the reasoning, either explicitly or implicitly. The jurisdictions examined are the UN Human Rights Committee (UNHRC), the UN Committee against Torture (UNCAT), the European Court of Human Rights (ECtHR), and the Court of Justice of the European Union (CJEU).¹⁵ They were chosen because: (a) they represent decision-making with respect to non-return to a ‘real risk’ of harm (which is directly relevant to refugee law and the ‘well-founded fear’ test), drawing on international law and/or European Union (EU) law; (b) they invoke the notion of imminence in their decisions, whether implicitly or explicitly; and (c) their decisions are publicly accessible, enabling an in-depth assessment of relevant decisions from each jurisdiction.

Each selected jurisdiction has its own treaty framework and distinctiveness. The UNHRC, a body of independent experts that monitors States’ implementation of the International Covenant on Civil and Political Rights (ICCPR),¹⁶ can consider individual complaints about alleged violations by States that are parties to its First Optional Protocol.¹⁷ In the present context, the most relevant allegations arise in relation to Articles 6 and 7 of the ICCPR, which, respectively, protect the right to life, and the right not to be subjected to torture or cruel, inhuman or degrading treatment or punishment.

The UNCAT, established by Article 17 of the Convention against Torture (CAT),¹⁸ may ‘receive and consider communications from or on behalf of individuals’ subject to the jurisdiction of any State that has made a relevant declaration pursuant to Article 22. For present purposes, the relevant provision is Article 3 which provides that ‘[n]o State Party shall expel, return (“refouler”) or extradite a person to another State where there are substantial grounds for believing that he would be in danger of being subjected to torture’.

In Europe, Articles 2 and 3 of the European Convention on Human Rights (ECHR) respectively protect the right to life, and the right not to be subjected to torture or inhuman or degrading treatment or punishment.¹⁹ Both have been acknowledged as extending to protect individuals from removal to a real risk of such harm²⁰—in other words, they have a non-refoulement component. This protection may be present in other rights as well: indeed, the ECtHR has expressly acknowledged this possibility with respect to Article 4 (prohibition of slavery and forced labour),²¹ Article 5 (right to liberty and security),²² Article 6 (right to a fair trial),²³ Article 7 (no punishment without law),²⁴ Article 8 (right to respect for private and family life),²⁵ and Article 9 (right to freedom of thought, conscience and religion).²⁶ However, whereas in a claim based on Articles 2 or 3 an applicant must show that there are ‘substantial grounds’ for believing that he or she would face a real (‘foreseeable’)²⁷ risk of being subjected to the relevant harm,²⁸ a claim based on the other provisions demands a much higher standard—a ‘flagrant breach’ or a ‘disproportionate breach’ of a right—which has rarely been met.²⁹ In many cases, violations of these other rights will already be encapsulated by the prohibition on removal to ‘inhuman or degrading treatment’ in Article 3,³⁰ which is why such cases are the primary focus in Part IV below.³¹

In the EU, the CJEU, through its preliminary reference procedure, answers questions referred to it by the national courts of EU Member States seeking clarification of the interpretation of EU law.³² The cases pertinent to the present article relate primarily to the interpretation of the Qualification Directive,³³ which protects third-country nationals recognized as refugees³⁴ and beneficiaries of ‘subsidiary protection’.³⁵ The latter are people who cannot be removed because there are ‘substantial grounds … for believing’ that they ‘would face a real risk of suffering serious harm’,³⁶ namely (a) the death penalty or execution; (b) torture or inhuman or degrading treatment or punishment; or (c) a ‘serious and individual threat to a civilian's life or person by reason of indiscriminate violence in situations of international or internal armed conflict’.³⁷ A significant difference between EU law and the ECHR is that the grant of refugee status or subsidiary protection status under the Qualification Directive results in a specific set of rights and benefits for beneficiaries, of which protection against refoulement is just one. The situation is notably different under the ECHR, which only provides for protection against refoulement.³⁸

III. TIME AND RISK ASSESSMENT IN INTERNATIONAL PROTECTION CASES

The notion that ‘time’ is relevant to the assessment of international protection is not new; rather, it is central to many aspects of both the refugee definition and the broader protection regime.³⁹ As Durieux has observed, refugee situations ‘are exceptional circumstances in international relations’ and, as such, ‘are necessarily temporary’.⁴⁰ Indeed, the UN High Commissioner for Refugees (UNHCR), the body vested with the responsibility to supervise the application of the Refugee Convention, was initially established as a temporary entity with a three-year mandate, which was subsequently extended in five-year increments until 2004 when it finally became open-ended.⁴¹ The cessation clauses of the Refugee Convention envisage that an individual's need for international protection may be temporary: for instance, refugee status may end if the circumstances in the country of origin change fundamentally.⁴² Contemporary debates about temporary protection, accelerated procedures, and durable solutions all invoke the concept of time as a central factor in refugee protection.

In terms of the substantive assessment of who is a refugee, time permeates every aspect of the refugee status determination (RSD) process:

On the one hand, time points backwards because refugee claims are grounded in what has happened in the past. … RSD, however, also faces forward in time. It reaches into the future and inquires as to the predicament of the claimant in his or her country of origin. Refugee status is recognised precisely because the future there for the claimant is anticipated to be one in which some form of serious harm will accrue to him or her at some projected point in time arising from a failure of the state of origin to protect the enjoyment of his or her fundamental human rights by reason of a Convention-protected ground.⁴³

The maxim that assessing a well-founded fear of persecution involves a forward-looking assessment of risk is well entrenched, yet there is little analysis of how far forward the assessment may extend, or what the role of time is in that prospective assessment. In light of the lack of explicit consideration of these issues, we begin our analysis by reflecting on whether there is there anything inherent in either the ‘well-founded fear’ or ‘real risk’ tests—the so-called ‘standard of proof’ in international protection claims—that provides answers to these questions.

The well-founded fear test in refugee law clearly entails a prospective and forward-looking assessment. Indeed, it has been stated that any refugee determination is necessarily an ‘essay in hypothesis, an attempt to prophesy what might happen to the applicant in the future, if returned to his or her country of origin’.⁴⁴ Therefore, a degree of speculation about future risk does not preclude a protection need from being recognized. However, other than these broad forms of guidance, it is striking that while there is a significant body of both jurisprudence and scholarship analysing a number of contentious aspects of the well-founded fear test, comparatively little has been written about the role of time itself.⁴⁵ Instead, debates about the ‘well-founded fear’ test have centred on: (a) whether it entails a subjective notion or is entirely objective;⁴⁶ (b) the necessary degree of likelihood to establish that a fear/risk is ‘well-founded’;⁴⁷ and (c) evidentiary issues, including past persecution and generalized violence.⁴⁸

It is in relation to the degree of likelihood that one would assume a consideration of time would most likely be found. In perhaps the most cited decision on this topic, the United States Supreme Court in Cardoza-Fonseca explained that:

There is simply no room in the United Nations’ definition for concluding that because an applicant only has a 10% chance of being shot, tortured, or otherwise persecuted, that he or she has no ‘well-founded fear’ of the event happening … [A] moderate interpretation of the ‘well-founded fear’ standard would indicate ‘that so long as an objective situation is established by the evidence, it need not be shown that the situation will probably result in persecution, but it is enough that persecution is a reasonable possibility’.⁴⁹

That the well-founded fear test requires only ‘a reasonable possibility’ of persecution, rather than the likelihood of harm, is an approach that has been widely replicated, with a variety of interchangeable tests effectively amounting to one of ‘real chance’.⁵⁰ While excluding risks that amount to ‘sheer speculation’, situations where there is only a ‘bare possibility’ of harm, or where the risk is ‘so slight that it could be discounted’,⁵¹ it is an appropriately liberal test that reflects the protective objective of the refugee definition and the inherent challenges in establishing a future risk of persecution with any certainty.

Analysis of the ‘well-founded fear’ test has tended not to invoke concepts of directness of harm, necessity, or any indication of when a risk needs to eventuate. It is possible that a risk becomes more ‘remote’ or speculative, and thus less likely, the further into the future it is anticipated, but imminence—in the sense of timing—has not and cannot be assumed to constitute an inherent aspect of the well-founded fear test. Rather, that test would appear to be sufficiently open-textured to accommodate a longer-frame assessment of future risk. It thus appears capable of encompassing the evolving nature of many contemporary forms of slower-onset harms which may present less immediate, but no less serious, risks to human rights.

How does the well-founded fear test relate to the parallel ‘real risk’ test in the context of non-refoulement under human rights law (complementary protection)? While the Refugee Convention itself provides the test for refugee law, namely, a well-founded fear of being persecuted, there is no clear test provided in most of the relevant human rights treaties due to the fact that the non-refoulement obligations derived from the ICCPR and ECHR are implied from other, primary obligations (such as the right not to be subjected to torture). The exception is CAT, where Article 3 explicitly provides that no State shall expel or return a person to another State ‘where there are substantial grounds for believing that he would be in danger of being subjected to torture’.

Just as the well-founded fear test has spawned a considerable literature, the real risk test has similarly stimulated analysis of matters such as whether there is a subjective or purely objective element,⁵² whether the evidentiary threshold should vary depending on the gravity of harm,⁵³ and how precisely the relevant standard of likelihood of harm should be framed. In relation to the last of these, there has been some debate as to whether the ‘real risk’ test represents a higher threshold for applicants than the well-founded fear test,⁵⁴ although the better view is that they should be equivalent.⁵⁵ Insofar as there is a further gloss on the meaning of real risk, it is the notion of foreseeability—not imminence that is most often invoked.⁵⁶ Yet, foreseeability in and of itself does not denote a particular timeframe.

To the extent that there has been any assessment of the role of time in the context of the real risk test, it has focused on the need for a forward-looking assessment from the date of determination. The ECtHR, for example, is explicit that while historical events may shed light on present or future events,⁵⁷ current circumstances are determinative. Often, but not always, this formulation is accompanied by the ECtHR declaring that it makes an ex nunc assessment (‘from now on’). For example, in SHH v United Kingdom, it stated:

In order to determine whether there is a real risk of ill-treatment in this case, the Court must examine the foreseeable consequences of sending the applicant to Afghanistan, bearing in mind the general situation there and his personal circumstances … A full and ex nunc assessment is called for as the situation in a country of destination may change over the course of time. Even though the historical position is of interest insofar as it may shed light on the current situation and its likely evolution, it is the present conditions which are decisive and it is therefore necessary to take into account information that has come to light since the final decision taken by the domestic authorities.⁵⁸

Thus, an ex nunc test requires that the future risk be assessed at the date of the decision-maker's consideration of the case.⁵⁹ In other words, procedurally it requires an assessment of future risk today, but it should not foreclose the substantive assessment. As the Grand Chamber of the ECtHR recalled in JK and Others v Sweden, the principle of an ex nunc evaluation of the circumstances requires that the ‘assessment must focus on the foreseeable consequences of the applicant's removal to the country of destination in the light of the general situation there and his or her personal circumstances’.⁶⁰ The court further explained

that it is assessing the applicants’ situation from the present-day point of view. The main question is not how the Swedish immigration authorities assessed the case at the time (that is, when the Migration Agency and the Migration Court took their decisions on 22 November 2011 and 23 April 2012 respectively) but rather whether, in the present-day situation, the applicants would still face a real risk of persecution for the above-mentioned reasons if removed to Iraq.⁶¹

The CJEU, too, undertakes an ex nunc test. In EU law, the assessment of facts and circumstances is regulated by Article 4 of the Qualification Directive. According to the CJEU, this

‘assessment’ takes place in two separate stages. The first stage concerns the establishment of factual circumstances which may constitute evidence that supports the application, while the second stage relates to the legal appraisal of that evidence, which entails deciding whether, in the light of the specific facts of a given case, the substantive conditions laid down by Articles 9 and 10 or Article 15 of Directive 2004/83 for the grant of international protection are met.⁶²

More specifically regarding timing, the Qualification Directive provides that the assessment of all relevant facts is to be carried out on an individual basis, as ‘at the time of taking a decision on the application’.⁶³ This ex nunc assessment of facts and circumstances is further enshrined in the Procedures Directive, which requires that ‘Member States shall ensure that an effective remedy provides for a full and ex nunc examination of both facts and points of law, including, where applicable, an examination of the international protection needs pursuant to [the QD (recast)]’.⁶⁴

One context in which there is explicit reference to a minimum future period is in relation to actors of protection.⁶⁵ Of particular relevance to the present study is the fact that protection from persecution or serious harm under EU law should be ‘effective and of a non-temporary nature’.⁶⁶ While there is no finite future period specified, the fact that ‘protection’ against persecution or serious harm must be non-temporary and durable strongly points to the need for a temporal assessment that looks well beyond the present, or immediate, risk of harm, and into the future.⁶⁷

Having thus established that there is no conceptual reason why imminence—in the sense of timing—should be used to limit a State's protection obligations, we now turn to consider how some judicial bodies have nonetheless begun to import this notion into the assessment of international protection claims.

IV. ‘IMMINENCE’ IN INTERNATIONAL PROTECTION: THE JURISPRUDENCE OF INTERNATIONAL COURTS AND OVERSIGHT BODIES

In examining jurisprudence on imminence in international protection, we focused on cases where the notion of imminence was explicitly invoked, or where the court or body used the concept of timing of future harm as a relevant factor in the assessment of the protection claim.⁶⁸

The notion of imminence has been referred to and used explicitly in all jursidictions examined for this research, and, in particular, has been engaged in some cases to reject a claim for international protection.⁶⁹ For example, in at least 18 cases in our study, the ECtHR relied on the absence of an imminent risk to reject the application.⁷⁰ Although it has not been invoked as a ground for refusal of protection by the CJEU,⁷¹ that court does recognize considerations of time, seriousness of harm, and likelihood/probability as key elements in the interpretation of EU law relating to protection.⁷²

While the number of cases that have invoked the notion in an explicit manner to date is small in the context of the overall caseload of these jurisdictions, it is nonetheless significant that it has been invoked at all, given that it does not appear to have a solid foundation in traditional principles of risk assessment in the law on international protection. Similarly, although some decision-makers have expressed scepticism about whether the invocation of a standard of ‘real and imminent risk’ (as has been adopted in a line of ECtHR authority primarily concerning Dublin transfers) is simply a result of ‘nice differences of language’,⁷³ the reality is that it has been used to deny protection—and the danger is that it could be adopted in future cases. This is particularly concerning given the lack of focused discussion about why it is used, and whether there is any principled basis for doing so.

Analysis of the case law from our selected jurisdictions reveals that while supranational courts and international oversight bodies apply certain general principles (eg substantial grounds for believing and real risk), their decisions lack detailed reasoning or information about how near or far in time a risk of future harm needs to be, what type of evidence is required to substantiate it, and why (when explicitly referred to) ‘imminence’ is used as a test at all, given that it is not expressly contained in any of the relevant legal frameworks.

The analysis below identifies the two key contexts in which these challenges are most acute. The first is where the notion of imminence has been transplanted from a procedural setting (where, arguably, it has an appropriate role to play) to a substantive assessment (where, arguably, it does not). The second concerns protection claims involving a risk of harm that may manifest in the medium- or long-term, such as medical cases where a person's health may decline over a period of time and environmental deterioration linked to the impacts of climate change. Such cases inevitably raise policy considerations about the role of the receiving State in contributing to/mitigating adverse impacts, and the responsibility of the host State to protect people from potential (but perhaps mitigable) harm.

A. Imminence: The Dangers of Transplantation from Procedural to Substantive Issues

The context in which the notion of imminence is most likely to be invoked explicitly is in assessing whether an applicant has met procedural requirements related to standing or ‘victimhood’. This is particularly pronounced in the international and regional adjudicatory human rights bodies where a complainant must establish that he or she has standing to bring a complaint.⁷⁴

For example, Article 1 of the ICCPR's First Optional Protocol provides that the UNHRC can consider complaints from individuals ‘who claim to be victims of a violation by that State Party of any of the rights set forth in the Covenant’.⁷⁵ Article 2 further provides that ‘individuals who claim that any of their rights enumerated in the Covenant have been violated and who have exhausted all available domestic remedies may submit a written communication to the Committee for consideration’. Where a complainant does not meet the criteria of a ‘victim’, the claim is deemed inadmissible.⁷⁶

The UNHRC has long taken the view that an individual can only claim to be a victim of a violation ‘if he or she is actually affected’.⁷⁷ In 1993, it introduced a further gloss on this test by stating that to be a victim, a person ‘must show either that an act or an omission of a State party has already adversely affected his or her enjoyment of such right, or that such an effect is imminent’.⁷⁸ In that case, the UNHRC was not persuaded that, ‘at the relevant period of time’, the continuing deployment of nuclear weapons in the Netherlands placed the authors—a group of 6,588 concerned Dutch citizens—in the position of ‘victims’.⁷⁹

The same approach was adopted in the 2006 decision of Aalbersberg v Netherlands, which involved a group of 2,085 Dutch citizens who claimed a violation of Article 6 (the right to life) by virtue of the Netherlands’ willingness to use nuclear weapons and to ‘cooperate with such use’.⁸⁰ The UNHRC framed the question of standing in terms of whether ‘the State's stance on the use of nuclear weapons presented the authors with an existing or imminent violation of their right to life’.⁸¹ No such case was made out and the UNHRC therefore dismissed the claim as inadmissible.⁸²

The rationale for the use of imminence in relation to standing is manifestly clear: the international and regional adjudicatory bodies do not have the power to issue advisory opinions or engage in moot adjudications.⁸³ Further, and perhaps more importantly, where a person is not currently a ‘victim’ for jurisdictional purposes, it is of course possible that the circumstances may change and hence permit a subsequent application to be made once exposure to harm is, indeed, imminent. This was explicitly acknowledged in an analogous context in H v Norway, where the ECtHR struck out the claim but nonetheless reassured the applicant that ‘should her legal situation in Norway change and a risk of expulsion become imminent, it remain[ed] open to her to lodge a new application with the Court’, which could include ‘a request for an interim measure’.⁸⁴

However, while the notion of imminence may be justified in the limited context of admissibility, our analysis reveals that there is sometimes a conflation of two conceptually distinct issues: the question of standing/victimhood, and the substantive question of the human rights violation. In the latter context, assuming that there is an imminent risk of removal, extradition, deportation, or expulsion, the assessment of State responsibility—and hence of likely future harm—must be undertaken at that moment and cannot be deferred to a later time when the evidence is clearer or more compelling. As long as the removal is imminent—and thus the person's standing or ‘victimhood’ is made out—the substantive question turns on the likelihood of harm resulting from such removal, and arguably not on precisely how soon after removal it may manifest. To conflate these two fundamentally different contexts and issues, especially without explanation as to the justification for doing so,⁸⁵ is highly problematic.

For example, just two weeks after its decision in Aalbersberg, the UNHRC conflated these separate issues by finding a claim to be inadmissible because, inter alia, the applicant had not adduced sufficient evidence in support of his contention that he would be ‘exposed to a real and imminent risk of violations of articles 6 and 7 of the Covenant if deported to Pakistan’.⁸⁶ Similarly, in another largely contemporaneous decision, the UNHRC found a claim to be inadmissible, in part because the complainant had failed to adduce sufficient evidence that he would be ‘exposed to a real and imminent risk of violations of articles 6 and 7 of the Covenant’.⁸⁷ In these admissibility decisions, the UNHRC apparently applied the notion of imminence—from a test for standing—to the substantive consideration of the alleged violations.

Similarly, in Europe, the ECtHR has held that even where all the formal admissibility conditions have been met, it ‘may nevertheless declare [an application] inadmissible for reasons relating to the examination on the merits’.⁸⁸ For instance, in a number of decisions involving Dublin transfers of asylum seekers from one EU Member State to another, the ECtHR has found that it has not been ‘demonstrated’ or ‘established’ that an applicant's ‘future prospects if returned …, whether taken from a material, physical or psychological perspective, disclose a sufficiently real and imminent risk of hardship severe enough to fall within the scope of Article 3’ of the ECHR.⁸⁹ In reaching these decisions, at times the court curiously relied on its reasoning in cases such as Ireland v United Kingdom ⁹⁰ and Jalloh v Germany,⁹¹ neither of which referred to ‘real and imminent risk’. The rationale for introducing an imminence test has not, therefore, been articulated or justified.

B. Imminence and Slow-Onset Harm: Particular Challenges

The timing of future harm is relevant in a range of factual contexts, but the two areas in which the issue is most likely to be raised in contemporary jurisprudence are risks associated with deteriorating health on return to the country of origin, and those associated with the impacts of climate change in the country of origin.

Returning to the departure point for our inquiry, the key challenge is whether it is possible or appropriate to identify a basis on which an international protection application can be rejected if the harm feared by the applicant is likely to manifest over the longer term. Is there (and should there be) a principled way of defining the relevant time period in assessing risk of future harm? How would variables such as the potential mitigating impacts of disaster risk reduction measures be accommodated within the analysis? As standard refugee cases show, context is extremely important and the rules cannot be too prescriptive: the focus is not on certainty of harm, but whether there is a real risk of it.⁹²

By way of analogy, in situations relating to generalized violence, decision-makers must carry out a detailed analysis of all possible factors and consequences, including potential risk-mitigating factors and their attendant effects. As UNHCR's Guidelines on International Protection No 12 (UNHCR's Guidelines) relating to situations of armed conflict and violence note:

When assessing the risk, it is important to take into account the fluctuating character of many contemporary situations of armed conflict and violence. … For example, even if the level of violence at the time of decision-making is relatively low, over time the situation of armed conflict and violence may change, increasing the degree of risk establishing a well-founded fear. … Similarly, even if violence has not yet broken out in a particular part of the country, it may be foreseeable that the violence will spread there, taking into account the overall context and history of the situation of armed conflict and violence, the trajectory and mapping of the violence, the power dynamics at play and other conditions in the applicant's country of origin. The effects of past violence may also still rise to the level of persecution, despite a temporary suspension of hostilities, and need to be assessed carefully. In addition, the implementation of peace and demobilization agreements may lead to new armed actors filling vacuums of power, or to the consolidation of groups composed of former members who have not disarmed and reintegrated into society. This also requires a detailed analysis that constantly evolves in response to local developments in the country of origin.⁹³

Furthermore, the ‘long-term consequences’ of armed conflict or violence, ‘such as demolition of vital infrastructure, insecurity and abject poverty’,⁹⁴ can amount to persecution or serious harm, while

[t]he systematic denial of food and medical supplies, the cutting of water supplies and electricity, the destruction of property or the militarization or closure of hospitals and schools may also constitute serious human rights or IHL violations that affect whole communities.⁹⁵

UNHCR's Guidelines are very useful in highlighting multi-causal movements, the unpredictable nature of armed conflict, its frequently shifting patterns over time, and the long-term consequences of conflict. Notably they do not purport to explain where to draw the line in relation to future harm precisely because this is not a decisive factor. Rather, they state that decision-makers should look at the trajectory of harm (violence, in that context) in deciding whether it is ‘foreseeable’ that it will spread into an area where there is no sign of violence at the time of the risk assessment.⁹⁶ The focus of the assessment remains on foreseeability of harm, and the Guidelines suggest factors to guide that assessment. They highlight that even where mitigating measures might intervene (such as the implementation of peace and demobilization agreements), there may still be negative outcomes. In other words, the presence of ‘multiple’ possibilities, each of which involves speculation, does not foreclose the possibility that protection is warranted. Rather, the flexibility and open-ended nature of the well-founded fear and real risk tests provides the necessary framework for decision-makers to assess protection claims.

Against this backdrop, we now turn to two contexts in which these issues have arisen, namely health deterioration and movement in the context of the adverse impacts of climate change.

1. Health deterioration

The context in which the question of imminence of harm has arisen most squarely to date concerns cases involving deterioration of health after removal. The most developed jurisprudence on this issue emanates from the ECtHR. Its case law spans decades but has not developed consistently.⁹⁷ In the seminal decision of D v United Kingdom in 1997, the court held that the proposed removal of a citizen of St Kitts from the United Kingdom in the ‘advanced stages of a terminal and incurable illness’⁹⁸ would amount to a violation of Article 3 of the ECHR on the basis that:

The abrupt withdrawal of these facilities will entail the most dramatic consequences for him. It is not disputed that his removal will hasten his death. There is a serious danger that the conditions of adversity which await him in St Kitts will further reduce his already limited life expectancy and subject him to acute mental and physical suffering.⁹⁹

While imminence of harm was not an explicit consideration in that case, it was central (albeit implicit) in a subsequent decision in 2008, namely N v United Kingdom. There, the court rejected an analogous claim on the basis that:

the quality of the applicant's life, and her life expectancy, would be affected if she were returned to Uganda. The applicant is not, however, at the present time critically ill. The rapidity of the deterioration which she would suffer and the extent to which she would be able to obtain access to medical treatment, support and care, including help from relatives, must involve a certain degree of speculation, particularly in view of the constantly evolving situation as regards the treatment of HIV and AIDS worldwide.¹⁰⁰

This decision had the effect of precluding successful claims by applicants seeking to resist removal due to health-related issues.¹⁰¹ However, in the 2016 judgment of Paposhvili v Belgium, the ECtHR adopted a less rigid approach. In that case, the Grand Chamber accepted a longer time frame for the risk assessment, noting that ‘the impact of removal on the person concerned must be assessed by comparing his or her state of health prior to removal and how it would evolve after transfer to the receiving State.’¹⁰² Nevertheless, the applicant must still be ‘seriously ill’ at the time of removal, and while

not at imminent risk of dying, would [nevertheless] face a real risk, on account of the absence of appropriate treatment in the receiving country or the lack of access to such treatment, of being exposed to a serious, rapid and irreversible decline in his or her state of health resulting in intense suffering or to a significant reduction in life expectancy.¹⁰³

This is reminiscent of the language employed by the CJEU in the 2014 Abdida decision in which the court referred to ‘the seriousness and irreparable nature of the harm that may be caused by the removal’, namely ‘a serious risk of grave and irreversible deterioration in his state of health’.¹⁰⁴ More recently, the CJEU has considered this issue in some depth, citing the test adopted by the ECtHR in Paposhvili; namely, ‘a serious, rapid and irreversible decline in his state of health’.¹⁰⁵ The CJEU concluded that in assessing the obligation of non-refoulement in Article 4 of the Charter of Fundamental Rights of the European Union, return is not permitted ‘where such removal would result in a real and demonstrable risk of significant and permanent deterioration in the state of health of the person concerned’.¹⁰⁶ The use of the word ‘deterioration’ by the CJEU appears to imply that an applicant's medical condition need not be at a critical stage at the time of the decision.

The concession in Paposhvili that a decision-maker must consider the evolving situation post-transfer is significant, although there is no precise framing of the outer limits of the relevant temporal period. However, it clearly allows for a longer-term assessment than the more limited test articulated in N v United Kingdom. This is certainly how Paposhvili has been understood by the English Court of Appeal, which recently affirmed that ‘it is clear’ that the case ‘relaxes the test for violation of Article 3 in the case of removal of a foreign national with a medical condition’, but that ‘it does so only to a very modest extent’.¹⁰⁷ The Court of Appeal's interpretation assumes that imminence is still a requirement; it has merely been extended from ‘imminence of death’¹⁰⁸ as the criterion to ‘a real risk of a serious and rapid decline in … health resulting either in intense suffering (to the Article 3 standard) or death in the near future’.¹⁰⁹

The Paposhvili decision is consistent with an analogous 2015 decision of the UNHRC. There, Canada was found to be in violation of Article 7 of the ICCPR for deporting a mentally ill person to Jamaica, notwithstanding Canada's arguments that the potential longer-term consequences of deportation were ‘mere conjecture’ and ‘mere speculation’ because it was ‘just as likely’ that the situation would not be as dire as predicted.¹¹⁰ On the other hand, in a case where there was no evidence of the applicant's health deteriorating on the available evidence at the time of the determination, yet arguably a longer-term risk of death in the DRC, UNCAT found this aspect of the claim inadmissible.¹¹¹

In sum, the health context suggests that while the jurisprudence has become more open to the assessment of a claim within a longer time frame, the notion of imminence, in the sense of timing, is still a present and limiting factor that influences the scope of States’ obligations to provide international protection.

2. Climate change

The other key context in which the question of timing of harm arises concerns claims linked to the adverse impacts of climate change. While the jurisprudence is far less developed than in the health context, such claims are ripe for future litigation and therefore warrant analysis.

The most comprehensive consideration of the future impacts of climate change in the protection context has been undertaken by decision-makers in New Zealand.¹¹² While the notion of ‘imminence’ has been invoked explicitly in such cases (based on the UNHRC cases on standing),¹¹³ the Tribunal has observed that ‘[i]mminence should not be understood as imposing a test which requires the risk to life be something which is, at least, likely to occur’.¹¹⁴ Rather, it is comparable to the well-founded fear test in refugee law, requiring ‘no more than sufficient evidence to establish substantial grounds for believing the appellant would be in danger’.¹¹⁵ However, it is important to note that the Tribunal found that the forward-looking nature of the protection inquiry demands that disaster risk reduction and adaptation strategies be taken into account in assessing future risks.¹¹⁶ As subsequently expressed by a member of the tribunal in an extra-curial capacity, the longer the future time frame, the less predictive certainty there is and the more speculative the risk assessment becomes because of the potential for mitigation, adaptation, and other risk-reducing actions.¹¹⁷ Accordingly, in the cases considered to date, the climate-change-related harms were assessed as being insufficiently ‘present’, or the applicants not ‘in danger’, in part due to the potential for intervening adaptive measures.¹¹⁸ This conclusion is debatable given that, as recognized in the UNHCR Guidelines related to armed conflict and violence, the possibility of multiple potential outcomes does not automatically overcome the need for international protection, particularly when the trajectory of harm is moving in a clear direction.

In this regard, it is instructive to consider assessments by the Intergovernmental Panel on Climate Change (IPCC) of the likelihood of certain future climate change risks (which could affect displacement). The findings listed below are based on the IPCC's evaluation of the underlying scientific evidence and the degree of expert agreement, the combination of which determines the level of ‘confidence’ ascribed (ranging from ‘very low’ to ‘very high’).¹¹⁹ The term ‘virtually certain’ (in the last bullet point) means that the probability is 99–100 per cent.¹²⁰ According to the IPCC:

• • Impacts from recent climate-related extremes, such as heat waves, droughts, floods, cyclones, and wildfires, show the significant vulnerability and exposure of many human systems to current climate variability (very high confidence).¹²¹

• • It is very likely that heat waves will occur more often and last longer, and that extreme precipitation events will become more intense and frequent in many regions. The ocean will continue to warm and acidify, and global mean sea level will rise.¹²²

• • Low-lying areas are at risk from sea-level rise, which will continue for centuries even if global mean temperature is stabilized (high confidence).¹²³

• • It is virtually certain that global mean sea-level rise will continue for many centuries beyond 2100 (the amount will depend on future emissions).¹²⁴

• • Over the course of the twenty-first century, climate change is expected to lead to increases in ill-health in many regions, especially in developing countries with low incomes (high confidence).¹²⁵

• • The negative impacts of climate change on crop yields, across a wide range of regions and crops, have been more common than positive impacts (high confidence).¹²⁶

• • Climate change is projected to increase risks in urban areas for people, assets, economies and ecosystems, including from heat stress, storms and extreme precipitation, inland and coastal flooding, landslides, air pollution, drought, water scarcity, sea level rise and storm surges (very high confidence). ‘These risks are amplified for those lacking essential infrastructure and services or living in exposed areas.’¹²⁷

The high degree of likelihood expressed in each finding above would certainly meet the well-founded fear threshold in refugee law.¹²⁸ As discussed above in Part III, it is not disputed that a ‘well-founded fear’ of being persecuted can equate to a less than 50 per cent chance of harm (even as low as 10 per cent) if the risk is plausible and reasonable.¹²⁹ Speculation about what ‘might’ happen in the future is thus a central component of determining refugee status.¹³⁰ However, decision-makers place limits on how pre-emptive flight must be to warrant protection, which differ depending on the particular individual's circumstances and turn on ‘the relation between the nature of the persecution feared and the degree of likelihood of its happening’.¹³¹ The assessment of the intensity, severity, and nature of future harm, based on its foreseeability in light of the individual's circumstances, is the crucial factor.¹³²

The IPCC has noted that ‘[t]ransformations in economic, social, technological and political decisions and actions can enhance adaptation and promote sustainable development (high confidence)’,¹³³ and ‘[i]nnovation and investments in environmentally sound infrastructure and technologies can reduce GHG [greenhouse gas] emissions and enhance resilience to climate change (very high confidence).’¹³⁴ However, in the context of a protection claim, the future mitigation/adaptation possibilities remain speculative and uncertain, and do not detract from the current trajectory of adverse climate change impacts. The competing possibilities thus need to be evaluated in light of the less demanding standard of proof applied in refugee and human rights law.

In conclusion, while the climate change context presents particular challenges, the assessment of an applicant's need for international protection should rest on uniform principles of refugee and human rights law, albeit with sufficient flexibility to respond to different factual contexts.

The two examples discussed above highlight the difficulty of identifying a precise time frame to assess potential future harm. Indeed, it is likely impossible and inappropriate to attempt to devise with any precision a time frame that delimits the scope of inquiry. After all it is not clear why State responsibility for future human rights violations following removal should be limited to those occurring within an immediate or short-term period.

V. DOES IMMINENCE HAVE ANY PLACE IN INTERNATIONAL PROTECTION?

As mentioned above, there is little discussion in the case law as to the rationale and justification for invoking an imminence assessment. However, one notable issue that has emerged from a close reading of the case law for this article is that, although ‘well-founded fear’ has generally not been overlaid with the additional gloss of necessity, directness, or even foreseeability, these concepts are prevalent in the jurisprudence on ‘real risk’. Indeed, it appears that it is the introduction of additional factors by some supranational courts and other international oversight bodies, such as ‘necessary and foreseeable’,¹³⁵ which has sometimes resulted in an imminence-like assessment. Where courts invoke a foreseeability test alone, imminence does not appear to be relevant. But where the concept of necessity (or similar) is introduced, if the risk is not imminent then it seems difficult for decision-makers to conclude that a necessary consequence of removal is the violation of human rights.¹³⁶

The problem is that ‘necessary’ and ‘foreseeable’ are conceptually distinct, and it is not clear where the necessity test originated or what its principled basis could be. To take the jurisprudence of the UNHRC as an example, General Comment 31 sets out the basis for an implied non-refoulement obligation, stating that:

Moreover, the article 2 obligation requiring that States Parties respect and ensure the Covenant rights for all persons in their territory and all persons under their control entails an obligation not to extradite, deport, expel or otherwise remove a person from their territory, where there are substantial grounds for believing that there is a real risk of irreparable harm, such as that contemplated by articles 6 and 7 of the Covenant, either in the country to which removal is to be effected or in any country to which the person may subsequently be removed. The relevant judicial and administrative authorities should be made aware of the need to ensure compliance with the Covenant obligations in such matters.¹³⁷

There is no mention of ‘necessity’ in this context, yet from the 1990s onwards, the UNHRC has sometimes described the relevant test as whether, ‘as a necessary and foreseeable consequence of deportation’, there will be a violation.¹³⁸ The jurisprudence is inconsistent, however, and our analysis has revealed that where foreseeability is the test, a claim is more likely to succeed.¹³⁹ By contrast, when additional factors such as whether the harm would be a ‘necessary’ consequence of removal are introduced, the claim is more likely to be rejected.

A good illustration of the issue is provided by Ng v Canada. That case concerned a man facing execution on return to the United States, and the UNHRC found that, with respect to the potential method of execution, the sending State could ‘reasonably foresee that Mr Ng, if sentenced to death, would be executed in a way that amounts to a violation’ of Article 7 of the ICCPR.¹⁴⁰ This was so despite strong dissenting views which pointed to the reasons why the harm might not eventuate, including that the applicant had not yet been ‘tried, convicted or sentenced’.¹⁴¹ Had the test been one of certainty or necessity, then these factors would have dictated that the claim fail.¹⁴² However, as one of the dissenting members explained, foreseeability, not certainty, is key:

Although it is impossible to predict a future event, it must be understood that whether or not a person is a victim depends on whether that event is foreseeable—or, in other words, on whether, according to common sense, it may happen, in the absence of exceptional events that prevent it from occurring … .¹⁴³

A similar outcome can be seen where the ECtHR focused squarely on foreseeability.¹⁴⁴ In the extradition case of Rustamov v Russia, the court stated that:

[R]equesting an applicant to produce ‘indisputable’ evidence of a risk of ill-treatment in the requesting country would be tantamount to asking him to prove the existence of a future event, which is impossible, and would place a clearly disproportionate burden on him. In this respect it further reiterates its constant case-law to the effect that what should be assessed in this type of case are the foreseeable consequences of sending the applicant to the receiving country.¹⁴⁵

Further, where the UNCAT has ignored the language of ‘imminence’ that has sometimes been invoked by complainants,¹⁴⁶ preferring instead to rely on whether a risk of harm is ‘real, foreseeable and personal’,¹⁴⁷ the claim has been more likely to succeed.¹⁴⁸ Indeed, as Wouters observes, although the UNCAT used the terms ‘foreseeable’ and ‘necessary’ in the first individual case regarding Article 3 of the CAT,¹⁴⁹ (thus requiring that the risk of torture be both a foreseeable and necessary consequence of the removal), it has never used the term ‘necessary’ again. In Haydin v Sweden, the UNCAT pointed out that ‘the requirement of necessity and predictability should be interpreted in the light of its general comment on the implementation of Article 3’—in other words, rejecting the notion that these concepts imposed any higher standard than ‘real risk’.¹⁵⁰

If the evidence is strong, an event or occurrence can be foreseeable even if it is not likely to manifest in the short term. A risk can be real notwithstanding the fact that its manifestation is some time away. Of course, foreseeability itself does not provide a template for a definable future period, since the ‘temporal limits embodied in the notion of reasonably foreseeable future are unclear’,¹⁵¹ but it is nonetheless fundamentally different from certainty, likelihood, or probability.

When factors such as necessity are introduced, then the longer the time frame considered, the less likely it is that the long-term harm can be said to amount to a necessary consequence of removal. As such, if extraneous concepts are incorporated into the real risk test in human rights law, it may be less able to accommodate slower-onset forms of harm than the well-founded fear test in refugee law. In other words, the well-founded fear test may have a greater capacity and openness to accommodate longer-range harm than the more stringent real risk test applied in some of the jurisdictions considered in this article. In order to test this hypothesis, a comprehensive examination of the extent to which imminence has been introduced—explicitly or implicitly—into national refugee decision-making is ripe, and forms the next stage of our research project.

VI. CONCLUSION

This article provides the first analysis of the notion of imminence in international refugee and human rights law in assessing the scope of international protection. It examines how acute or ‘real’ a risk of harm must be before supranational courts and international oversight bodies consider that international protection is warranted. The comprehensive analysis of cases from our selected jurisdictions suggests that the notion of imminence has been introduced into the case law—sometimes explicitly, but most commonly only implicitly. Its application is inconsistent, problematic, and difficult to justify by reference to standard principles of refugee and human rights law. Given that a number of contemporary threats are more likely to raise these issues than ever before, further analysis is warranted, including a comprehensive examination of domestic jurisprudence.

While this article represents part of a larger project, some interim conclusions can be drawn. The role of time as a factor in risk assessment in refugee law and international protection more broadly is radically underdeveloped, both in the jurisprudence and the scholarship. To be clear, we do not believe that the notion of imminence should be used to limit international protection in any way: it should not. However, the point in time at which an event may occur is a contextual factor in assessing risk because it can affect how likely something is to occur. In other words, if a threat is likely to manifest very soon, it may be easier for an individual to argue that he or she faces a real chance of being harmed, since there may be less opportunity for mitigating factors to intervene. By contrast, a threat in the more distant future might be able to be allayed or diminished through intervening measures. Nevertheless, if the risk of harm is distant but very real—for instance, the death penalty has been imposed on a person but it may be years before it is carried out—then protection may be required. The core question in each case is whether the risk is sufficiently real to meet the well-founded fear or real risk tests; any further gloss is a distraction.¹⁵² Our critique is that because the role and significance of time is simply not acknowledged by most decision-makers, there is the potential for imminence to creep erroneously into substantive assessments, as has been seen in the examination of jurisprudence above. In our view, both the well-founded fear and real risk tests are sufficiently open-ended for decision-makers to be guided by these alone; they do not foreclose a risk of harm simply because it may eventuate further into the future.

It is indisputable that a protection-limiting criterion should not infiltrate decision-making without an explicit consideration of its legal validity and purpose. It is our hope that this initial attempt to parse out the relevant issues in this context will mark the first of many serious analyses.