Sydney Law Review
A Well-Founded Fear of Being Persecuted ... But When?
Adrienne Anderson, Michelle Foster,† Hélène Lambert‡ and Jane McAdam§
It is well established that the ‘well-founded fear’ test in refugee law requires a prospective assessment of potential future harm. Yet, the requisite timeframe for this test is rarely examined. Analysis of jurisprudence across a wide range of jurisdictions reveals that Australian courts have been unusually cognisant of the question of timing of harm. Indeed, they have been particularly insistent that a flexible and longer-range assessment is appropriate, encapsulated by the ‘reasonably foreseeable future’ test. This article provides an in-depth analysis of the principles set out by Australian courts and tribunals, and identifies particularly challenging contexts in which timing has played an important role. It also assesses the extent to which decision-makers at the tribunal level adhere to the flexible approach formulated by the judiciary. It is hoped that our analysis of Australian jurisprudence may prove helpful in other jurisdictions in which the issue of timing of harm is equally pertinent, but far less developed.
The question whether an individual qualifies for refugee status turns on an assessment of his or her risk of being persecuted for reasons of race, religion, nationality, political opinion, or membership of a particular social group. Assessing such risk is undertaken by reference to the ‘well-founded fear’ test contained in art 1A(2) of the Refugee Convention, which defines the term ‘refugee’. Given the broad and largely undefined terms in the definition, refugee status determination can be a highly contested undertaking, with credibility often at the core of decision-making. The absence of consistent ‘country of origin’ information and traditional evidentiary sources (especially witnesses), in conjunction with the forward-looking, speculative assessment of risk, means that refugee status determination constitutes a uniquely demanding fact-finding task. In this light, it is hardly surprising that the meaning of ‘well-founded fear of being persecuted’ has been extensively examined by courts and scholars alike. Around the world, courts have emphasised that the Refugee Convention ‘looks to the future’, and that ‘a well-founded fear of future persecution is the touchstone of asylum’. Likewise, scholars agree that the test requires a forward-looking assessment of risk: it ‘is essentially an essay in hypothesis.’ Yet, while it is acknowledged that ‘time is everywhere’ in refugee law, there is very little consideration in either the jurisprudence or the scholarship as to how far into the future the risk of persecution may extend for refugee protection to be forthcoming. On the one hand, it is impossible to develop any precise timeframe, since all refugee claims are necessarily contextual, with various factors weighted differently depending on the individual circumstances of the case. On the other hand, given how meticulously the concept of well-founded fear has been analysed, it is surprising that such limited attention has been paid to this question.
While caution ought to be exercised in circumscribing too closely an open-ended phrase such as ‘well-founded fear’, the lack of guidance on the question of timing has allowed a notion of imminence — or immediacy of risk — to infiltrate refugee status determination silently across a wide range of jurisdictions. In some cases, this has resulted in the denial of protection where harm is not deemed sufficiently imminent to warrant protection under international human rights law. In other cases, denial of protection is the result of a (mis)application of the Refugee Convention. Hence the concern to examine the question of timing of harm is not a mere academic exercise.
Moreover, there are certain ‘types’ of contemporary protection cases in which the nearness in time of harm seems to play a critical role — such as those relating to the (future) impacts of climate change and to deterioration of health over time. Although the feared harm is not felt acutely now, it may have deleterious consequences in the future. In New Zealand (‘NZ’), for instance, a series of cases has begun to delineate the scope of refugee and human rights law to protect Pacific Islanders at risk of the negative impacts of climate change, disasters and environmental degradation. These kinds of cases inevitably require analysis of how far forward in time the assessment of risk may extend. Of course, in relation to climate-related displacement, it is important to acknowledge that refugee law will not be a good fit in most cases. The impacts of climate change and/or disasters generally will not satisfy the meaning of ‘persecution’ because of the need for human agency, and even if such harm could be characterised as persecution, a further challenge would be linking the persecution to one of the five Refugee Convention grounds. However, human rights-based protections from refoulement may apply, which prevent states from sending people to places where they face a real risk of being arbitrarily deprived of life, or subjected to cruel, inhuman, or degrading treatment. Since jurisprudence in this area often draws on refugee law principles by analogy, the capacity of refugee law to protect people from future risks may be highly relevant. Indeed, in the NZ cases, the notion of ‘imminence’ was explicitly invoked, with the decision-maker observing that ‘[i]mminence should not be understood as imposing a test which requires the risk to life be something which is ... likely to occur’. 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’. That is why it is crucial to understand what the well-founded fear test entails.
Australian decision-makers have grappled with the idea of future risk in a relatively sophisticated and nuanced way, compared to other jurisdictions. Indeed, our analysis of decisions across a wide range of jurisdictions revealed a comparatively large number of Australian cases over the past 20 years — at both the Tribunal and court levels — in which the timing of harm was considered in some depth. The clear principle that emerges in Australian jurisprudence is that when examining whether a person has a ‘well-founded fear of being persecuted’, and thus requires international protection, the relevant timeframe for assessing risk is not the immediate future, but rather the ‘reasonably foreseeable future’. By contrast, our analysis of the implementation of the law of international protection in a range of international and regional jurisdictions has identified that the notion of ‘imminence’ — in the sense of immediacy of harm — has begun to infiltrate decision-making, sometimes explicitly, but, much more commonly, implicitly. Our concern is that because the role and significance of time is rarely explicitly acknowledged or assessed in other jurisdictions, including in comparative domestic jurisdictions, there is potential for an ‘imminence test’ to slip into refugee status determination without critical reflection or examination. In this context, a critical appraisal of one of the few jurisdictions to have examined the issue of timing of harm explicitly, namely, Australia, is instructive and timely, given that cases involving a risk of harm that may take longer to materialise are increasingly likely to arise in assessing international protection obligations.
There is no international tribunal or committee vested with jurisdiction to hear individual complaints or otherwise authoritatively dictate a common interpretation of the Refugee Convention, meaning that refugee law has evolved mostly through the interpretation of national courts and tribunals. As such, the Australian approach to these issues offers particular insights that may have broader relevance.
Part II of this article explores briefly the concept of well-founded fear and its relevance to the timing of harm. Next Part III examines the Australian doctrinal position on the question of timing in more detail. Part IV identifies particular circumstances in which imminence is still implicitly or explicitly invoked by Australian decision-makers, especially at the tribunal level, despite superior rulings that this is not the correct approach. Finally, Part V analyses the ‘reasonably foreseeable future’ standard in Australian law, drawing out relevant principles and guidance that may assist in future refugee status adjudication. It should be noted at the outset that while ‘time’ is central to many aspects of refugee protection, this article focuses specifically on the relevance and significance of the timing of future harm in determining whether a person is assessed as having a well-founded fear of being persecuted.
It is clear that the well-founded fear test in refugee law requires a prospective, forward-looking assessment. It ‘necessarily involve[s] a degree of speculation. No one knows with certainty what the future holds.’ As Goodwin-Gill and McAdam note, ‘a decision on the well-foundedness or not of a fear of persecution is ... an attempt to prophesy what might happen to the applicant in the future, if returned to his or her country of origin.’ The requirement that the fear be ‘well-founded’ speaks to the need for objective evidence that the fear is plausible and reasonable, thus permitting the finding that the individual, in his or her particular circumstances, faces a real chance of being persecuted. Past harm may be relevant to, but not determinative of, future persecution.
A person’s fear can be ‘well-founded’ even if he or she ‘only has a 10% chance of being shot, tortured, or otherwise persecuted’. In Australia, this approach is encapsulated by the ‘real chance’ test, which
clearly conveys the notion of a substantial, as distinct from a remote chance, of persecution occurring ... . If an applicant establishes that there is a real chance of persecution, then his fear, assuming that he has such a fear, is well-founded, notwithstanding that there is less than a 50 per cent chance of persecution occurring.
This is an appropriately liberal test that reflects the protective objective of the refugee definition and the inherent challenges involved in establishing a future risk of persecution with any certainty. In our view, it appears capable of accommodating a longer timeframe assessment of future risk, and is therefore able to encompass evolving, slower-onset harms that may present less immediate, but no less serious, threats. In this regard, it is interesting to consider a more general observation by the Federal Court of Australia on the relationship between time and harm:
All other factors remaining the same, the chance of harm will tend to multiply proportionally to the time spent exposed to it. In other words, the longer a person is exposed to a source of harm, the more likely it is that at some stage the person will encounter that harm. In that way, even a risk that on its face is remote or fanciful, may increase through prolonged exposure such that the level of risk becomes real.
Part III, below, examines the approach of Australian courts and tribunals with respect to how far into the future a risk may extend for protection to be warranted.
As early as 1996, the High Court of Australia endorsed the ‘reasonably foreseeable future’ test to assess a ‘real chance’ of harm. It did not engage in a detailed analysis of its meaning, but simply noted that the approach taken by the Tribunal was correct (namely, whether there was a real chance that the applicant ‘would be persecuted for a Convention reason were he to return at this time or within the reasonably foreseeable future’). The Federal Court of Australia has since held that a decision-maker who considers only ‘the present or immediate future’, as distinct from ‘the reasonably foreseeable future’, falls into jurisdictional error. The ‘question of harm in the reasonably foreseeable future is a mandatory relevant consideration’, and failure to consider it is a reviewable error of law. Thus, a decision-maker cannot simply consider ‘future possibilities over a very short, future time frame’, but must ‘prognosticate the situation into the reasonably foreseeable future’, which involves ‘an assessment of the period of time to look into the future’.
The most pertinent examination of the meaning of ‘reasonably foreseeable future’ was undertaken by Mortimer J in the 2017 Federal Court case of CPE15 v Minister for Immigration and Border Protection:
The ‘reasonably foreseeable future’ is something of an ambulatory period of time, but the use of reasonable foreseeability as the benchmark concept indicates that the assessment is intended to be one which can be made on the basis of probative material, without extending into guesswork. It is also intended to preclude predictions of the future that are so far removed in point of time from the life of the person concerned at the time the person is returned to her or his country of nationality as to bear insufficient connection to the reality of what that person may experience.
In essence, the assessment must be contextual. The ‘reasonably foreseeable future’ cannot be confined to any particular timeframe; it is a relative notion that will depend on the circumstances of each individual case. The further away in time the risk, the more probative the evidence needs to be that the particular individual is at risk. However, an applicant who fears harm in the more distant future does not have to provide a higher quality of evidence, and nor does the standard of proof change, regardless of the length of that period. There must also be a degree of flexibility, as suggested by Kirby J’s observations in Guo:
The places from which refugees normally flee rarely have legal or administrative systems that permit the rational and consistent application of logic which our courts like to boast of but sometimes themselves fail to provide. To say this is not to intrude into assessment of the merits or to impose a conclusion about likely events in any particular country. It is simply to accept the inherent unpredictability of the future and the special difficulties which arise in assessing accurately the possible course of political or other oppression in the kinds of countries from which refugees typically come.
In this regard, a decision-maker’s choice of, and willingness to engage with, country-of-origin information that speaks to future trends or trajectories of harm may be central to the outcome.
A decision-maker is not required to refer explicitly to the ‘reasonably foreseeable future’ since that language is not used in the legislation. Conversely, however, merely referring to the ‘reasonably foreseeable future’ is not sufficient to discharge the decision-maker’s duty to consider it. As the Federal Court held in BOT15 v Minister for Immigration and Border Protection, the Tribunal’s mention of ‘reasonably foreseeable future’ did
no more than set out the test. It is a bare assertion that is insufficiently explained and lacks logical connection to the material and analysis that precedes it. There is no consideration by the Tribunal of what may happen after the completion of the withdrawal of foreign troops and of how the country information demonstrates that the appellant does not face a real chance of serious harm or a real risk of significant harm in the reasonably foreseeable future. Its focus on the near completion of the withdrawal of foreign troops looks to the past and present and, possibly, to the near future, and not to the reasonably foreseeable future.
Likewise, it is not necessarily determinative if a decision-maker refers to ‘current risk’ and uses only the present tense (as opposed to referring to ‘future risk’ and/or phrasing the assessment in the future tense). The requirement is that a prospective assessment is made in substance, regardless of the ‘manner of expression’.
Thus, in all cases, a decision-maker must carefully assess the applicant’s claim in light of what might occur in the reasonably foreseeable future, regardless of what terminology he or she uses.
Notwithstanding that superior courts have been clear that decision-makers ought to consider a longer timeframe in assessing well-founded fear, extensive examination of the case law reveals ongoing challenges in practice. Indeed, in particular contexts, the most prominent of which are examined below, the notion of imminence or immediacy of harm still features in the reasoning of Australian decision-makers.
One of the most striking contexts in which imminence continues to be invoked is where the decision-maker undertakes a retrospective assessment of the applicant’s departure, often as a method of assessing credibility and/or the question of subjective fear.
This is particularly common at the tribunal level. There are dozens of decisions (including some very recent ones) in which the Tribunal’s reasoning identifies whether an applicant’s delay in leaving the country of origin, travel through a third country without claiming asylum, and/or delay in applying for protection in Australia was inconsistent with an imminent (risk of) harm. This tends to imply that an imminent risk at the point of departure is necessary in order to establish a well-founded fear of persecution. Applicants’ claims have been erroneously rejected where the decision-maker has based the decision on past events without substantively assessing the applicant’s fear into the reasonably foreseeable future.
There are a number of decisions refusing protection where the alleged delay (equated to a lack of immediacy of harm) has led to an inference that an applicant lacked genuine subjective fear or credibility. For example, in 1514886 (Refugee), the Tribunal stated that:
The applicant sought protection in Australia, after many years of visiting this country, because there is greater acceptance of LGBT people and more favourable lifestyle options compared with Fiji. His family and personal circumstances influenced the timing of this decision. He did not leave Fiji in response to any past persecution or significant harm, or any imminent fear that he would be subject to such harm.
Similarly, in 1101896, the Tribunal found, in relation to credibility,
that if the applicant was indeed under the threat of imminent harm, he would not have waited four months before applying for a protection visa. For [this and] all the above reasons, the Tribunal does not find the applicant to be a credible, truthful and reliable witness.
The Tribunal has also made numerous similar findings in relation to a lack of ‘urgent fear’, which may, in some cases, be considered a proxy for imminence. In only one such case invoking urgency was the applicant granted protection because it was accepted that she still had a well-founded fear of being persecuted, notwithstanding doubts as to her credibility. The Tribunal observed that the
notable delay in the applicant’s protection visa application and the weakly argued reasons for that delay ... indicate[d] that the applicant’s otherwise genuine fears were not as deeply and urgently held as exaggeratedly presented to the Tribunal.
Ultimately, however, it concluded that
the applicant was genuinely but not urgently motivated to leave her country of origin based on [a relevant] incident for reasons of safety. Therefore, the applicant did have a genuine personally-held fear of serious harm at the time of her departure and at the time she applied for a protection visa, as well as now and into the foreseeable future.
It is not possible to assess whether, and to what extent, this reasoning is invoked at the departmental stage of decision-making since such decisions are not published. However, some tribunal decisions indicate that the Department is also using imminence in assessing credibility and/or subjective fear. For example, in 1504740 (Refugee), the Tribunal noted that:
The applicant and his wife could neither afford not to work nor take their children to a childcare centre for they feared for their safety as their circumstances and the threats prevented them from continuing to attend school. The Department incorrectly interpreted the applicants’ actions as evidence the family did not face imminent danger and the children were safe at home. On the contrary the children could not go to school as there were threats to kidnap them.
In another matter, the applicant argued that a similar finding on delay by the departmental delegate was ‘inconsistent with the Convention which does not require that a person be in immediate trepidation’. The Tribunal did not address this argument in its reasoning, although the case was successful on other grounds.
The Tribunal’s reasoning on credibility and the absence of imminent fear/risk has received very little attention by the courts. Despite being set out in some decisions, it has not been the subject of review and it has only rarely been commented on.
References to imminent danger or fear at time of flight are problematic. These notions are not contained in the Refugee Convention, or in established jurisprudence on the meaning of well-founded fear. Yet, their invocation in the cases discussed above suggests that some applicants are being required to show that they left on account of an imminent risk of harm in order to establish at least one component of the well-founded fear test (namely, subjective fear). Although less frequent elsewhere, there are also instances in Canada, the United States (‘US’) and NZ where protection has been denied for this reason.
The fundamental problem with this approach is that refugee status determination is a prospective exercise in which an assessment must be made of future risk. As such, it should not matter whether the risk at the time of flight was urgent — or even in existence. Yet, the reasoning in this line of cases incorrectly implies that there is a right and a wrong time to leave, and that the Tribunal is in a position to determine whether the applicant chose the right time.
The language (and notion) of imminence should therefore be discarded in the context of reviewing the trajectory of an applicant’s experience prior to flight, as it may distract a decision-maker from the core function of assessing the well-founded fear of persecution in the future (which, in any event, is largely an assessment of objective risk of future harm).
The notion of imminence remains prevalent in assessing future harm even though it is clear that the well-established test of ‘reasonably foreseeable future’ for assessing well-founded fear in Australian refugee law requires a longer-frame assessment (see Part III). In some decisions, applicants themselves argued that they faced imminent risk or harm. However, more concerning is the fact that the Tribunal has rejected claims on the basis that there was no risk of harm arising from any ‘imminent’ event or circumstance.
The Australian Government’s Procedures Advice Manual (‘PAM3’) advises that ‘[h]ow far into the future a decision-maker should consider in terms of future harm (i.e., the real chance of persecution) will vary depending on the circumstances in the receiving country’, spanning from ‘a period as short as a week or month (possibly even days)’ to ‘more than a year’. It is, with respect, difficult to understand how the test of ‘reasonably foreseeable future’ could plausibly be satisfied by an assessment that considered a window confined to just one week. Indeed, it is difficult to imagine a scenario in which it would be appropriate to consider only a few days or a week into the future. In reality, what might vary is the period of time in which it is claimed (or the facts indicate) that harm may actually occur. For example, the evidence may point to a risk of immediate harm, such as arrest at the airport. However, that is an evidentiary, or factual, question, and in cases where the decision-maker finds that there is a real chance that this risk may eventuate immediately, it may be a simple decision. However, that may not always be the appropriate window of time for the decision-maker in similar cases. Were there to be a finding that risk would not eventuate in the short-term, further enquiry should not be foreclosed. Rather, in most cases, the decision-maker should go on to consider a period of time after return, in the event that the person is not arrested immediately or is released, but subsequently arrested (and that such treatment (cumulatively) amounts to persecution).
In situations where the Tribunal has improperly focused only on the current or present risk, the courts have been willing to intervene. There are numerous cases in which an appeal has been successful on this ground. However, despite lucid and coherent authority emanating from the superior courts, it has not always been followed by the Tribunal. Furthermore, the possibility of review decreases as one moves through the hierarchy, such that only a small percentage of cases reach judicial review stage, underlining the importance of the merits review stage.
In sum, the question of the timing of harm and the relevance of ‘imminence’ to assessing future harm remains problematic, at least at the tribunal level. This is despite the High Court of Australia signalling an expansive approach to assessing well-founded fear over 20 years ago. As a concern that has largely gone unnoticed in the scholarship to date, it is hoped that by drawing out the issues, this article can contribute to a more coherent and principled approach.
Even though there is some inconsistency in assessing risk prospectively, on the whole Australian courts and tribunals have engaged with the question of timing of harm more closely than any other jurisdiction we have examined, including by the United Nations (‘UN’) Committee against Torture, the UN Human Rights Committee, the European Court of Human Rights, the Court of Justice of the European Union, and in other countries.
A decision-maker’s willingness to engage with this question may be the difference between a positive and negative outcome for an applicant who fears persecution. A comparative analysis of claims in the United Kingdom (‘UK’) and Australia concerning possible political violence in upcoming elections illustrates the point. In EM, the UK Upper Tribunal (Immigration and Asylum Chamber) considered (and rejected) the reasoning in an Australian decision on ‘reasonably foreseeable future’ that gave credence to ‘reasonable speculation’, finding instead that ‘the further away the elections, the more uncertain are their consequences’. The UK Upper Tribunal explained: ‘[t]he combined effect of the evidential uncertainty of when elections may be called and what might happen when they are produces a picture that is too equivocal or obscure to amount to a real risk of future ill treatment.’
By contrast, the Federal Court of Australia found, in relation to an appeal by a Fijian applicant, that the Tribunal had erred in confining its enquiry to the immediate future when elections were due to occur sometime after the immediate period upon return. The Court considered that
there was a failure to address the reasonably foreseeable future in the context of the claims made. The dealing with the three incidents was based on immediate facts — no elections looming and the character of the present government. This reflected a focus on immediacy which was no real assessment of whether in the future, with elections looming, with the first appellant campaigning for the FLP, he would not face a similar beating for the same reasons, or threats from elements of the Taukei Movement who had already targeted him.
Parts V(A)–(E), below, identify key principles that emerge from decades of Australian jurisprudence on the meaning of ‘reasonably foreseeable future’ in the context of assessing risk. These principles have the potential to provide guidance to other jurisdictions, as well as in relation to emerging protection claims that raise issues of timing particularly acutely. Part V(A) analyses the different time periods that have been considered in the cases generally, while Parts V(B)–(E) briefly reflect on four contexts where a longer timeframe may be particularly appropriate: namely, claims concerning children, slow-onset environmental harm, armed conflict, and health deterioration.
Consistent with the notion that a well-founded fear of persecution involves an apprehension of future risk, the past may be a guide, but the assessment of risk must always be forward-looking from the date of the determination. In particular, this assessment must look beyond the ‘very short, future time frame’ into the ‘reasonably foreseeable future’. However, the ‘temporal limits embodied in the notion of reasonably foreseeable future are unclear’. For instance, in overturning a tribunal decision that had rejected a refugee claim because the applicant had remained safe in the two to three months prior to flight, the Federal Circuit Court of Australia stressed that ‘the future was not a short closed period but an indefinite period’.
This understanding of an open-ended notion of time when assessing ‘reasonably foreseeable future’ is present in other cases too. For instance, the Federal Circuit Court has suggested that ‘the future tense [should be used] in an unconfined way’ when making findings about what might happen to an individual on return, rather than focusing on a ‘looming event’.
But how indefinite or unconfined is this future period?
Very few cases circumscribe a concrete time period, although a number of cases have contemplated ‘a period of some years’. For instance, in AUK15 v Minister for Immigration and Border Protection, the decision-maker considered that the ‘reasonably foreseeable future’ could encompass the next two to three years (based on the anticipated withdrawal of troops from Afghanistan). In another case, the Full Federal Court found that the Tribunal should have considered the impact on the applicant of escalating political violence linked to possible elections in 12–18 months’ time (from when the applicant lodged his protection application in May 2010). The Court noted that ‘the Tribunal did not assess in any real or active way what the situation would be in mid to late 2011 or thereafter’.
The Full Federal Court has acknowledged that it may be difficult for an applicant to persuade a decision-maker that there is a real chance of persecution if it will not materialise ‘for some time after his or her return’. Even so, for a decision-maker to apply the correct test, ‘it may be necessary to consider whether the applicant’s fear of being persecuted in the more distant future (and not merely in the period shortly after his or her return) is well-founded’; ‘a month-by-month assessment’ is a reviewable error. Indeed, the Federal Court has repeatedly confirmed that a decision-maker may be in error where, in rejecting the applicant’s claim, he or she looks only to the near future and relies on a lack of evidence of a current threat. This is because the Tribunal ‘must not foreclose reasonable speculation about the chances of the hypothetical future event occurring’, which involves ‘speculation as to circumstances in the future on the basis of material in the present, and what has happened to the person in the past’.
In Mok v Minister for Immigration, Local Government and Ethnic Affairs (No 1), the Federal Court reviewed a decision in which the decision-maker had considered that a real chance of persecution just two months after the applicant’s return to Cambodia ‘wasn’t relevant’. The Court held that ‘the question of whether there was a real chance of persecution necessarily required the delegate to look at the future in so far as it was reasonably foreseeable at the time when he was making his decision.’ However, by way of explanation, it added:
On the one hand the delegate was not required to look at the possibility of something occurring in 50 years’ time — to take the example given by Mr Paterson to Mr Rose (transcript 2293). On the other hand a delegate errs in law, in my opinion, if he confines his attention to whether there is a real chance of persecution on the day after an applicant’s return. In my opinion on the delegate’s evidence of what he considered, he erred in law.
Thus, in this case, two months was considered too short a period; more broadly it was agreed that 50 years into the future was too long and one day was too confined.
There have been cases where decision-makers have considered a period of several years as the ‘reasonably foreseeable future’. In one case, the Tribunal found that an applicant aged in his 40s had a well-founded fear of persecution because of the risk he would be called up for reserve military service (mandatory to age 51) were he to return to Israel. There was evidence that men over 35 years of age often were not called up for reserve training as they were considered medically unfit, and were usually discharged at the age of 41 or 45. However, there was ‘no clear indication that the applicant would not be called up if he returned to Israel until the age of 51’; thus, although this was some years away, the Tribunal found that there was a real chance it could occur within the reasonably foreseeable future.
In sum, the cases discussed above support a non-prescriptive, flexible approach by the Australian courts (much less so by the Tribunal) to the notion of time in refugee status determination. This approach treats a ‘reasonably foreseeable future’ as open-ended, permits reasonable speculation and is attuned to the importance of context in refugee cases.
Parts V(B)–(E) below highlight four specific contexts where an open-ended approach to the future may be particularly appropriate.
In cases involving minor applicants, there is an especially strong case for taking a longer-range view of prospective harm. This is well accepted, particularly in cases involving the risk of female genital mutilation. This is also the case where the type of harm feared entails the denial of socio-economic rights, such as health care or education, or cumulative harm, which in itself is often assessed over a longer period (sometimes a lifetime). As the guidelines of the UN High Commissioner for Refugees (‘UNHCR’) on child asylum claims emphasise:
While it is clear that not all discriminatory acts leading to the deprivation of economic, social and cultural rights necessarily equate to persecution, it is important to assess the consequences of such acts for each child concerned, now and in the future.
A good example of this approach in Australian jurisprudence is found in a case concerning the potential risk to an Ethiopian child of being subjected to corporal punishment at school. At the time of the decision, the child was still a toddler, yet the Tribunal accepted a period of at least a decade as the ‘reasonably foreseeable future’:
I find that the reasonably foreseeable future does not extend beyond the applicant’s primary education [through to the age of 15 years], as anything beyond is dependent upon too many variables, including possible changes to government and the child’s own interests and abilities to pursue further studies.
Although, in this case, the Tribunal felt the need to impose an outer limit on the timeframe, it nonetheless adopted an appropriately longer-range view.
As noted above, the Federal Court of Australia in Mok (No 1) stated (in obiter dicta) that a decision-maker ‘was not required to look at the possibility of something occurring in 50 years’ time’. Presumably, this was on account of the potential for mitigating factors that might reduce or nullify the risk of harm. However, how should a decision-maker factor in scientific evidence that weighs very strongly in favour of certain risks manifesting over time — and certainly at a sufficient level to meet the well-founded fear threshold in refugee law?
This is particularly pertinent to cases involving the slower-onset impacts of climate change. As noted above, such cases may enliven states’ international protection obligations, which necessarily involve an assessment of whether or not a well-founded fear of harm (or real risk of harm, in human rights law) has been established. When it comes to climate change itself, the Intergovernmental Panel on Climate Change (‘IPCC’), which is the expert UN body for assessing the science related to climate change, has predicted with ‘very high confidence’ impacts from recent climate-related extremes, such as heat waves, droughts, floods, cyclones, and wildfires, and increased 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. The IPCC has noted with ‘high confidence’ that low-lying areas are at risk from sea-level rise, which will continue for centuries even if the global mean temperature is stabilised, and that it is virtually certain that global mean sea-level rise will continue for many centuries beyond 2100 (the amount will depend on future emissions). At the same time, the IPCC has acknowledged (with ‘very high confidence’ and ‘high confidence’, respectively) that ‘[i]nnovation and investments in environmentally sound infrastructure and technologies can reduce GHG [greenhouse gas] emissions and enhance resilience to climate change’ and ‘[t]ransformations in economic, social, technological and political decisions and actions can enhance adaptation and promote sustainable development’. That said, such mitigation and adaptation measures remain uncertain, and they do not detract from the current trajectory of adverse climate change impacts. Hence, a longer timeframe for assessing risk may be particularly suited to the context of slow-onset of climate change in order to capture both real risk and mitigating measures. Context is key, and the rules cannot be too prescriptive: in refugee law, the focus is not on certainty of harm, but whether there is a real risk of it. The competing possibilities therefore need to be assessed in light of this standard. A mere possibility of intervention or potential mitigating developments may not be sufficient to reduce an existing real risk (albeit one that will manifest in the distant future).
Armed conflict is another context in which arguments have been made for extending the timeframe. There are two lines of argument. First, the effects of armed conflict on civilians are indirect as well as direct, and the indirect effects (food insecurity, lack of essential healthcare, and so on) often take time to manifest at scale.
Second, armed conflicts are usually volatile in terms of severity and geographic scope, such that areas previously considered safe become conflict-affected. The point has been noted by the Tribunal in the case of Afghanistan, for instance, and appears to cause consternation for decision-makers applying the ‘reasonably foreseeable future’ test. There is a tendency in some cases to curtail the ‘reasonably foreseeable future’ to a relatively short period because of the fluidity of the situation and uncertainty over variables involved. For example:
Whilst the Tribunal accepts that Islamic State has made threats against Shias in Pakistan, particularly in Parachinar and Hangu, the Tribunal considers that the potential for Islamic State to infiltrate or increase its presence in Pakistan remains uncertain. The Tribunal considers that the applicant’s claims in relation to this issue are highly speculative given the variables involved such as actions that may be taken by the Pakistan state, neighbouring countries or Western countries.
An alternative approach would have been for the decision-maker to discuss the trajectory of Islamic State’s military gains against the already known actions of neighbouring or other countries, in order to assess the risk to the applicant based on these conflict trends.
The foreclosure of reasonable speculation about future events is reflected in PAM3, which states: ‘If a receiving country is subject to a violent political environment, events may change rapidly and the foreseeable future may be limited’. This general statement warrants scrutiny, as it does not make clear what is the implication of a violent political environment. While a rapidly changing environment might add to the complexity of a person’s experience and perception of risk, this statement from PAM3 should not be equated with an automatic shortening of the future window. Rather, existing authority supports the notion that volatility might increase the risk to the applicant in some cases and certainly requires a longer-frame assessment. For instance, in Minister for Immigration and Multicultural Affairs v Jama, a majority of the Full Federal Court of Australia noted the importance of taking into account the possibility of changing circumstances, particularly in a fluid situation, such as that in Somalia. Even though there may be no current risk of persecution, ‘a change in circumstances may readily be foreseen that would create a significant risk of persecution’. However, in cases involving the withdrawal of international troops from Afghanistan in 2014, some decision-makers took the withdrawal to suggest an increased risk based on predictions about the security situation post-2014, while others did not.
Signs that conditions might be improving should also be treated with caution since they may not be sustainable over the longer term. Thus, in 1703914 (Refugee), the Tribunal reasoned:
With such a rapidly changing political landscape it is difficult for the Tribunal to base a decision on probative material when so little is known and so many changes are occurring. It is possible that the political situation regresses. It is possible that the changes are cosmetic to the larger security apparatus and how it operates. It is possible that the country will reduce its heavy handedness. As such while acknowledging these changing circumstances I find that they cannot provide a basis upon which to ground a decision which projects into the reasonably foreseeable future.
In assessing the durability of changes in Afghanistan in the context of the cessation clause in art 1C(5) of the Refugee Convention, the Tribunal’s reasoning in V04/17240 provided an interesting insight into how future contingencies should be assessed in light of past and present circumstances:
(i) ... given the history of unrest in that country and apparent growing reluctance of the international community to make a long term and extensive commitment to securing some level of stability to the region it is premature to assume that anything has happened to date that would found a basis for confidence that the human rights situation in general in Afghanistan and the political dynamics which resulted in the persecution of the applicant in particular have materially changed so far as the short to mid term future of the nation is concerned. ...
(vii) Until effective new state structures are installed in Afghanistan, it would be premature to assert that the applicant no longer faces persecution in the reasonably foreseeable future ...
Logically, there is no reason why such analysis should be limited to cessation cases and not also apply to claims for refugee status: the two provisions, arts 1A(2) and 1C(5), are related. In V04/17240, the Tribunal (correctly in our view) considered the history of unrest and instability in Afghanistan, its human rights situation in general, and the absence of effective new state structures and permanent reform (that would indicate a safer future) over and above the many (volatile) changes in the political and military landscape in Afghanistan. Such a longer-frame assessment is appropriate in assessing refugee claims involving armed conflict.
Analogous issues have arisen in cases concerning the future health of applicants and the absence of appropriate medical treatment in the country of origin. In one matter, 1008364, involving an HIV-positive applicant from Indonesia, medical evidence suggested that the applicant was clinically stable and not ‘at imminent risk of death whilst taking the current medications’, but he would be ‘likely to endure a significant reduction in health status and a significantly increased risk of death if he were to return to Indonesia where access to the required treatment is unavailable’. The Tribunal found that while the Indonesian authorities were striving to provide appropriate care to people with HIV/AIDS, ‘their efforts [we]re hampered by, among other problems, a lack of resources, a recent decentralisation of the health sector, and by limited effectiveness in delivering basic health care services, especially to the poor’. Although a lack of appropriate medical treatment ‘would not occur as a result of discrimination but because of the lack of resources and endemic problems within the health service’, the end result would be that the applicant would have to rely on his family and community ‘to ensure his survival’. As such, the Tribunal found that
the hardship the applicant is likely to incur in the foreseeable future if he returns to Indonesia and is unable to access appropriate treatment, threatens his capacity to subsist, and therefore amounts to persecution in a Convention sense (s 91R(2) of the Act).
Another matter, 0903707, concerned a Vietnamese applicant with HIV and Hepatitis B who claimed that he would be unable to access necessary medical treatment on return to Vietnam and would die within three to five years. His family had disowned him because of his illness and he said that he would have no means of support. He also feared being ostracised and unable to find work or accommodation on account of the stigma associated with his illnesses. The Tribunal was provided with medical evidence that he ‘would be at risk of death in the short to medium term with the last months or years of his short life likely to be made more miserable because of discrimination.’ Even though there was anti-discrimination legislation in place in Vietnam, the Tribunal considered that ‘the Vietnamese government’s laws and policies regarding HIV/AIDS ha[d] not been effectively or uniformly implemented across the country, especially in rural areas such as Ca Mau’ and ‘there was little evidence that the social change required to minimize stigma and discrimination had occurred in remote rural areas like the applicant’s home area.’ The Tribunal found that the applicant had a well-founded fear of persecution given there was
a real chance that, if he returns to Vietnam now or in the reasonably foreseeable future, the applicant will face discrimination and stigmatisation which will deny him access to basic services and the capacity to earn a livelihood of any kind such that it threatens his capacity to subsist ...
These cases underline the importance of a contextual analysis, in which the intensity, severity and nature of future harm, based on its foreseeability in light of the individual’s circumstances, are the crucial factors. Although the temporal proximity of a potential event is not determinative of the likelihood of its occurrence, it is a contextual factor to be considered. Thus, 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. However, the decision-maker must weigh up the likelihood of the risk with any mitigation eventuating, rather than simply dismissing the risk as too speculative. Even so, if the risk of harm is distant but very real — as in each of the four specific contexts discussed in this Part — then protection may well be required.
It is widely understood that the well-founded fear test in refugee law requires some speculation in light of its forward-looking and predictive nature. Yet, the requisite timeframe involved in an assessment of well-founded fear is rarely examined. It is an error of law to suggest that protection is only forthcoming where (risk of) harm is imminent, yet this notion has started to creep into refugee status decision-making, usually without explicit examination or reflection.
Australian courts and tribunals are notable for being singularly attuned to the relevance of time in refugee law and, in particular, in identifying the need for a flexible, longer-frame, assessment of future harm. Yet, as this article has revealed, even in this jurisdiction, decision-makers still sometimes struggle with the appropriate frame of reference. This suggests that the issue of time requires explicit scholarly and judicial analysis, so as to safeguard against inappropriately narrow approaches that may result in a rejection of refugee status even where a person faces a well-founded fear of persecution, albeit not immediately on return.
This article constitutes the first analysis of the approach of Australian courts and tribunals in examining the relevant timeframe involved in a well-founded fear assessment. It identified that the accepted position is one that is forward-looking, from the date of determination, into the ‘reasonably foreseeable future’. That timeframe is open-ended (an ‘indefinite period’) and must not be confined (for instance, to impending events). Decisions (notably by the Tribunal) that conflict with this authoritative position are erroneous. The article highlighted four specific contexts in which there is a particularly strong case for a longer timeframe: claims in relation to children, slow-onset impacts of climate change, armed conflict, and health deterioration.
It is hoped that the considered and insightful Australian jurisprudence examined in this article may provide the basis for the formulation of an appropriately inclusive approach to assessing well-founded fear in other jurisdictions in which the issue of time is far less developed. In light of contemporary protection challenges, the need for an open-ended approach to the timing of harm in refugee law is more important than ever.
 Research Associate, Andrew & Renata Kaldor Centre for International Refugee Law, Faculty of Law, UNSW Sydney, New South Wales, Australia; PhD candidate, Melbourne Law School,
The University of Melbourne, Victoria, Australia. Email: firstname.lastname@example.org.
† Professor and Director, Peter McMullin Centre on Statelessness, Melbourne Law School,
The University of Melbourne, Victoria, Australia. Email: email@example.com; ORCID iD: 0000-0002-438-9484.
‡ Professor of Law, University of Technology Sydney, New South Wales, Australia. Email: firstname.lastname@example.org; ORCID iD:0000-0003-456-6219.
&# Scientia Professor and Director, Andrew & Renata Kaldor Centre for International Refugee Law, Faculty of Law, UNSW Sydney, New South Wales, Australia. Email: email@example.com; ORCID iD: 0000-0002-3999-7843.
This article was produced under the auspices of an Australian Research Council (‘ARC’) Discovery Grant on ‘The Concept of “Imminence” in the International Protection of Refugees’, DP160100079, and we acknowledge the generous support of the ARC. It builds on our foundational work in Adrienne Anderson, Michelle Foster, Hélène Lambert and Jane McAdam, ‘Imminence in Refugee and Human Rights Law: A Misplaced Notion for International Protection’ (2019) 68(1) International Comparative Law Quarterly 111. We are grateful to Hannah Gordon (JD student, Melbourne Law School) for her excellent editorial assistance, and to the anonymous referees for their thoughtful suggestions. We also appreciate the helpful feedback received at Harvard and McGill Law Schools when versions of this article were presented.
 Convention Relating to the Status of Refugees, opened for signature 28 July 1951, 189 UNTS 137 (entered into force 22 April 1954) read in conjunction with the Protocol Relating to the Status of Refugees, opened for signature 31 January 1967, 606 UNTS 267 (entered into force 4 October 1967) (together ‘Refugee Convention’). Article 1A(2) of the Refugee Convention defines a refugee as a person who,
owing to well-founded fear of persecution 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; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it.
 Hilary Evans Cameron, Refugee Law’s Fact-Finding Crisis: Truth, Risk, and the Wrong Mistake (Cambridge University Press, 2018) 33.
 Ibid 35–8.
 Minister for Immigration and Multicultural Affairs v Respondents S152/2003  HCA 18; (2004) 222 CLR 1,
27  (McHugh J) (‘S152/2003’).
 Camara v Attorney General (US), 580 F 3d 196, 202 (3rd Cir, 2009).
 Guy S Goodwin-Gill and Jane McAdam, The Refugee in International Law (Oxford University Press, 3rd ed, 2007) 56: ‘the decision-maker must then make a reasoned guess as to the future’; James C Hathaway and Michelle Foster, The Law of Refugee Status (Cambridge University Press, 2nd ed, 2014) 110: ‘the risk-oriented understanding of “fear” as forward-looking apprehension, and as mandating only a prospective appraisal of an applicant’s actual risk, is very much in accord with the underlying goals of the treaty’; Cornelis Wolfram Wouters, International Legal Standards for the Protection from Refoulement: A Legal Analysis of the Prohibitions on Refoulement contained in the Refugee Convention, the European Convention on Human Rights, the International Covenant on Civil and Political Rights and the Convention against Torture (Intersentia, 2009) 83: ‘The element of risk stipulated by the words “well-founded fear” is the backbone of the refugee definition ... . It implies that there needs to be a present or prospective risk of persecution which can objectively be established’; Andreas Zimmermann and Claudia Mahler, ‘Article 1 A, Para 2 1951 Convention’ in Andreas Zimmermann (ed), The 1951 Convention Relating to the Status of Refugees and Its 1967 Protocol: A Commentary (Oxford University Press, 2011) 281, 338: ‘The object and purpose of the 1951 Convention thus supports an interpretation of the notion of “well-founded fear” as forward-looking expectation of risk’.
 Goodwin-Gill and McAdam (n 6) 54.
 Bruce Burson, ‘The Concept of Time and the Assessment of Risk in Refugee Status Determination’ (Conference Paper, Kaldor Centre Annual Conference, 18 November 2016) 1 <http://www.kaldor
centre.unsw.edu.au/sites/default/files/B_Burson_2016_Kaldor_Centre_Annual_Conference.pdf>. See also at 2, 4.
 For the purposes of this article, the focus is refugee law. However, given the alignment of tests in the Australian context, the same analysis would apply to complementary protection cases.
 See Adrienne Anderson et al, ‘Imminence in Refugee and Human Rights Law: A Misplaced Notion for International Protection’ (2019) 68(1) International Comparative Law Quarterly 111. That article offers the first analysis of the notion of imminence in the jurisprudence on international protection from four supranational courts and international oversight bodies, namely, the United Nations (‘UN’) Human Rights Committee, the UN Committee against Torture, the European Court of Human Rights and the Court of Justice of the European Union. It argues that although the number of cases that have engaged explicitly with the notion remains quite small, ‘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’: Anderson et al at 124 (emphasis in original). It concludes with the urgent need for a greater understanding of the role of time in this area of law.
 In its full scope, our research project also considers jurisprudence from refugee status decision-making in Australia, Canada, France, Germany, New Zealand (‘NZ’), the United Kingdom (‘UK’) and the United States (‘US’). While important nuances exist between these jurisdictions, our general findings indicate imminent-like notions are used in the application of the well-founded fear test to varying degrees. The focus of this article, however, is the unusually sophisticated analysis of timing of harm in Australian case law compared to other jurisdictions.
 *AC (Tuvalu)  NZIPT 800517; AF (Kiribati)  NZIPT 800413; AF (Tuvalu)  NZIPT 800859; *BG (Fiji)  NZIPT 800091; Teitiota v Chief Executive of the Ministry of Business, Innovation and Employment  NZHC 3125;  NZAR 162; Teitiota v Chief Executive of the Ministry of Business, Innovation and Employment  NZCA 173;  NZAR 688; Teitiota v Chief Executive of the Ministry of Business, Innovation and Employment  NZSC 107. Note that an asterisked (*) NZ case means that it is a precedent. Australian and NZ authorities have considered cases on this issue at least as far back as 1995, as discussed in Jane McAdam, ‘The Emerging New Zealand Jurisprudence on Climate Change, Disasters and Displacement’ (2015) 3(1) Migration Studies 131. The UN Human Rights Committee followed the reasoning of the NZ tribunal and courts in Teitiota v New Zealand, UN doc CCPR/C/127/D/2728/2016 (24 October 2019).
 EW v Netherlands, UN Doc CCPR/C/47/D/429/1990 (8 April 1993); Aalbersberg v Netherlands, UN Doc CCPR/C/87/D/1440/2005 (12 July 2006).
 AF (Kiribati) (n 12) .
 Ibid. See also AC (Tuvalu) (n 12) . This was notwithstanding the Tribunal member’s recognition, in subsequent remarks given in a personal capacity, that in other areas of international law, such as self-defence, ‘imminence’ seems to envisage a very immediate timeframe for harm to materialise, and certainly more immediate than the real chance standard in refugee law: Burson (n 8) 7.
 This is the case in particular when compared with the jurisdictions mentioned above (nn 10–11).
 In this article, ‘Tribunal’ is used to describe decisions by both the Administrative Appeals Tribunal (Migration and Refugee Division) and its predecessor, the Refugee Review Tribunal.
 We identified 473 relevant Australian cases between 1996 and 2019 at either tribunal or court level as part of our broader ARC-funded imminence project. Approximately half of these cases concerned the question of timing of harm and were identified as relevant to this article because:
(1) ‘imminence’ was used explicitly in the case (by the applicant or the decision-maker);
(2) ‘imminence’ was used implicitly (through the use of synonyms such as ‘today’, ‘present’, ‘immediate’, ‘short-term’, ‘near future’) in framing the timing of harm; or (3) timing of harm was otherwise very relevant to the outcome of the claim as evidenced by the facts of the case and/or because the decision-maker considered whether the harm would occur in the ‘reasonably foreseeable future’. We footnoted only those that identified key principles, implemented those principles, or, on the contrary, where a rejection was based at least in part on the absence of a foreseeable risk, and risk of harm in the future was not (adequately) considered.
 See, eg, Minister for Immigration and Ethnic Affairs v Wu Shan Liang (1996) 185 CLR 259, 279 (Brennan CJ, Toohey, McHugh and Gummow JJ) (‘Wu Shan Liang’); SZQXE v Minister for Immigration and Citizenship  FCA 1292; (2012) 134 ALD 495; MZACU v Minister for Immigration and Border Protection  FCCA 486; AON15 v Minister for Immigration and Border Protection  FCAFC 48; (2019) 269 FCR 184; CSO15 v Minister for Immigration and Border Protection  FCAFC 14; (2018) 260 FCR 134; DUX16 v Minister for Immigration and Border Protection  FCA 1529 (‘DUX16’); DPM16 v Minister for Immigration and Border Protection  FCA 249 (‘DPM16’); ALV16 v Minister for Immigration and Border Protection  FCCA 626; CPF15 v Minister for Immigration and Border Protection (No 4)  FCCA 1169; EBZ17 v Minister for Immigration and Border Protection  FCCA 79.
 Anderson et al (n 10) 139.
 Our research reveals that the concept of imminence, in the sense of timing of future harm, has been invoked explicitly or implicitly in comparator jurisdictions, but with little or no analysis of the appropriateness or relevance of this factor. For example, in NZ decision-makers tend not to discuss future timeframes explicitly and, in many cases, the reasoning appears to indicate quite a circumscribed period of time as being the only period for assessing well-founded fear:
BA (Afghanistan)  NZIPT 801138; AH (Hungary)  NZIPT 801172; DT (India)  NZIPT 801159. In Canada, claims brought on the basis of a risk of female genital mutilation on return have sometimes been rejected because the harm was not assessed as likely to occur immediately on return: see, eg, Re X (Re) (Immigration and Refugee Board of Canada, Refugee Protection Division, RPD File Nos TA6-04120 TA6-11916, Daniel G McSweeney, 1 August 2007)  CanLII 80670; Re X (Immigration and Refugee Board of Canada, Refugee Appeal Division, RAD File Nos TB7-11035 TB7-11036 TB7-11037 TB7-11038, Christine Houde, 2 January 2019)  CanLII 64862. In the US, the language of imminent risk has been invoked to reject claims, see, eg, Hernandez-Jimenez v Sessions, 710 Fed Appx 257, 7 (7th Cir, 2018); Mejia-Ramos v Barr, 934 F 3d 789 (8th Cir, 2019). In the UK, these issues have arisen particularly in the context of health deterioration: Anderson et al (n 10) 131–2.
 Anderson et al (n 10) 140.
 Hélène Lambert, ‘Transnational Law, Judges and Refugees in the European Union’ in Guy S Goodwin-Gill and Hélène Lambert (eds), The Limits of Transnational Law: Refugee Law, Policy Harmonization and Judicial Dialogue in the European Union (Cambridge University Press, 2010) 1, 4.
 Hélène Lambert, Jane McAdam and Maryellen Fullerton (eds), The Global Reach of European Refugee Law (Cambridge University Press, 2013). Unlike many aspects of the refugee definition, which have been given a particular meaning under Australian statute, the well-founded fear test has not been domestically circumscribed.
 Anderson et al (n 10) 117–18. A practical example in the Australian context is the 2010 practice of the Australian Government Department of Immigration and Citizenship in imposing a ‘freeze’ on claims by Hazara asylum seekers from Afghanistan (for six months) and asylum seekers from Sri Lanka (for three months) in order ‘to ensure that decision-making was based on up-to-date, accurate realistic information about the country circumstances in those two places’: Joint Select Committee on Australia’s Immigration Detention Network, Final Report (Report, March 2012) [6.46]–[6.48] <https://www.aph.gov.au/Parliamentary_Business/Committees/Joint/Former_Committees/immigrationdetention/report/c06>. The Minister for Immigration and Citizenship, Chris Bowen, acknowledged at the time that
[a]s a result of the more exhaustive country information, there have [sic] been a decrease in the number of primary acceptances of claims from Afghans who are not subject to the processing pause. Even taking into consideration the possibility of some of these being overturned at review, the percentage of successful refugee claims is likely to be lower than in the past.
‘Govt Lifts Freeze on Refugee Processing’, ABC News (online, 30 September 2010) <https://www.abc.net.au/news/2010-09-30/govt-lifts-freeze-on-refugee-processing/2280450>.
Such a ‘freeze’ illustrates the broader relevance of time in refugee status determination (albeit not specifically in the application of the well-founded fear test).
 Wu Shan Liang (n 19) 288 (Kirby J).
 Goodwin-Gill and McAdam (n 6) 54.
 See ibid 64 and references there. As the US Board of Immigration Appeals has framed it, an individual’s fear of persecution ‘must have a solid basis in objective facts or events’: Matter of Acosta, 19 I&N Dec 211, 225–6 (BIA, 1985). See also Hathaway and Foster (n 6) 105: ‘The existence of a “well-founded fear” of being persecuted requires only that there be a forward-looking apprehension of risk, thus mandating a purely objective inquiry.’
 The requisite degree of probability must take into account the intensity of the fear, the nature of the projected harm (death, imprisonment, torture, detention, serious discrimination, etc.), the general history of persecution in the home country, the applicant’s personal experience and that of his or her family, and all other surrounding circumstances.
UN High Commissioner for Refugees (‘UNHCR’), ‘Brief of the Office of the United Nations High Commissioner for Refugees as Amicus Curiae in support of Respondent’, Submission in Immigration and Naturalization Service v Cardoza-Fonseca 85-782, 14 July 1986, 29 (United States) <https://www.refworld.org/docid/4b03c5818.html>.
 In Minister for Immigration and Ethnic Affairs v Guo (1997) 191 CLR 559, 574–5 (‘Guo’), the majority stated that:
Past events are not a certain guide to the future, but in many areas of life proof that events have occurred often provides a reliable basis for determining the probability — high or low — of their recurrence. The extent to which past events are a guide to the future depends on the degree of probability that they have occurred, the regularity with which and the conditions under which they have or probably have occurred and the likelihood that the introduction of new or other events may distort the cycle of regularity.
See also Minister for Immigration and Multicultural Affairs v Ibrahim  HCA 55; (2000) 204 CLR 1, 26  (McHugh J): ‘Past acts of persecution are usually strong evidence that the applicant will again be persecuted if returned to the country of his or her nationality. But the relevance of past acts of persecution depends upon the degree of likelihood that they or similar acts will occur in the future’. In S152/2003 (n 4) 27  McHugh J noted:
The Convention looks to the future. What has occurred in the past does not determine whether a person is a refugee for the purpose of the Convention. In determining whether that person has a well-founded fear that he or she will be persecuted if returned to the country of nationality, the past is a guide — a very important guide — as to what may happen. But that is all.
In N00/35501  RRTA 582, the Tribunal stated: ‘My task is not, however, to dwell on the past. Rather, it is to make an assessment about whether the applicants face a real chance of persecution in the future. In making this assessment I must necessarily draw on the events in the past’.
 Immigration and Naturalization Service v Cardoza-Fonseca,  USSC 32; 480 US 421, 440 (Stevens J) (1987). See also Minister for Immigration, Local Government and Ethnic Affairs v MOK  FCA 1035; (1994) 55 FCR 375, 407 (Sheppard J) (‘MOK’): ‘The chance spoken of is a chance that is less than 50 per cent and one which may be as low as 10 per cent.’
 Chan v Minister for Immigration and Ethnic Affairs (1989) 169 CLR 379, 389 (Mason CJ), 398 (Dawson J), 406–7 (Toohey J), 427–29 (McHugh J) (‘Chan’). Like in Australia, a well-founded fear test is required in other jurisdictions and is expressed as a ‘real chance’ test (such as in NZ and the UK), ‘reasonable possibility’ test (such as in Canada and the US) or ‘crainte avec raison’ test (such as in France), all requiring a forward-looking assessment. For an overview of this standard in different jurisdictions, see Hathaway and Foster (n 6) 113. See also Anderson et al (n 10).
 Chan (n 32) 389 (Mason CJ), referring also to Boughey v The Queen (1986) 161 CLR 10, 21 (Mason, Wilson and Deane JJ). See 1404760  RRTA 769, : ‘while “anything could happen in the future” and that “bombs and explosions could occur anywhere at any time”... [the Tribunal must] base its decision on the information before it and assess whether the chance of an event occurring in the reasonably foreseeable future is real or remote’. See also 1214267  RRTA 168, .
 Anderson et al (n 10) 119.
 AEN15 v Minister for Immigration and Border Protection  FCA 509, . See also Matthew Scott, Climate Change, Disasters, and the Refugee Convention (Cambridge University Press, 2020) 107: ‘the predicament of being persecuted is best understood as a condition of existence, permeated by risk and potentially punctuated by acts of persecution or other serious denials of human rights that reflect and reinforce that predicament’.
 AOX16 v Minister for Immigration and Border Protection  FCCA 132,  (‘AOX16’).
 Wu Shan Liang (n 19) 279 (Brennan CJ, Toohey, McHugh and Gummow JJ).
 MZYXR v Minister for Immigration and Citizenship  FCA 252; (2013) 141 ALD 276, 283  (‘MZYXR’).
 DUX16 (n 19) .
 QAAH v Minister for Immigration and Multicultural and Indigenous Affairs  FCAFC 136; (2005) 145 FCR 363, 391  (‘QAAH’). Although this case concerned cessation of refugee status under art 1C(5) of the Refugee Convention, rather than consideration of the refugee definition per se, similar principles apply in both contexts.
 AON15 v Minister for Immigration and Border Protection  FCAFC 48; (2019) 269 FCR 184, 196 .
 CPE15 v Minister for Immigration and Border Protection  FCA 591,  (Mortimer J) (‘CPE15’).
 By way of analogy, the Court of Justice of the European Union has explained that in situations of generalised violence, ‘the more the applicant is able to show that he is specifically affected by reason of factors particular to his personal circumstances, the lower the level of indiscriminate violence required for him to be eligible for subsidiary protection’: Elgafaji v Staatssecretaris van Justitie (C-465/07)  ECR I-00921, I-955 .
 Guo (n 30) 596 (Kirby J). See also Kirby J in Applicant NABD of 2002 v Minister for Immigration and Multicultural and Indigenous Affairs  HCA 29; (2005) 79 ALJR 1142, 1158 :
That risk would need to be judged by reference not only to current political and religious conditions in Iran but also to possible future conditions. Those conditions might change; not necessarily for the better. These questions were not considered explicitly by the second Tribunal, although clearly raised by the appellant’s reference to the dangers of return to Iran for an apostate Muslim like himself. They are crucial in judging whether his ‘fear’ of persecution is ‘well-founded’.
 Ministerial Direction No 84 — Consideration of Protection Visa Applications (24 June 2019), made under s 499 of the Migration Act 1958 (Cth), requires the Tribunal to take account of available relevant country information assessments prepared by the Department of Foreign Affairs and Trade (Cth) ‘expressly for protection status determination purposes’ (at ). See also BOR15 v Minister for Immigration and Border Protection  FCCA 152, ; 1213438  RRTA 1043; 1216090  RRTA 213, where the Tribunal preferred information that spoke to the future, rather than just the current situation. In 1215874  RRTA 58 the decision-maker relied on predicted future harm, giving weight to independent evidence and finding that ‘the situation regarding the Palestinian refugees in Lebanon is volatile and fragile and can be described metaphorically as a “time bomb.” I find this is not mere fanciful speculation but is an assessment based on factual evidence.’ (at ), and that the security situation, ‘rather than diminishing, is on the rise’ (at ). Contrast this with 1517729 (Refugee)  AATA 3034 , where the Tribunal found that a claim by a Chinese applicant was too speculative, where the applicant was separated from his wife, but feared violating the family planning policy in the future if he remarried and had more children. See also 1703287 (Refugee)  AATA 409, .
 CDW18 v Minister for Home Affairs  FCA 270,  (‘CDW18’). See also at :
‘The reasons read as a whole indicate that the Authority was assessing what might happen on return, including into the future.’ See also CDW18 v Minister for Home Affairs  FCCA 2334, –.
 CDW18 (n 47) . See also BOT15 v Minister for Immigration and Border Protection  FCA 654,  (‘BOT15’).
 BOT15 (n 48) . See also AIE15 v Minister for Immigration and Border Protection  FCA 610,  (‘AIE15’).
 See AOX16 (n 36) , , , in which the Court was not persuaded that the Tribunal’s exclusive use of the present tense indicated that the reasonably foreseeable future had not been considered. See also BYH16 v Minister for Home Affairs  FCCA 2051; cf cases where the Court found that the Tribunal’s present-tense expression pointed to a focus only on the past and present: BKD17 v Minister for Immigration and Border Protection  FCCA 3182 (‘BKD17’); DRO17 v Minister for Immigration and Border Protection  FCCA 3547 (‘DRO17’). But in some cases, the use of the future tense enabled the Court to find that the Tribunal had carried out its assessment prospectively: BJU15 v Minister for Immigration and Border Protection  FCCA 1296; AQK17 v Minister for Immigration and Border Protection  FCCA 3584.
 AOX16 (n 36) .
 SZGHS v Minister for Immigration and Citizenship  FCA 1572,  (‘SZGHS’); AIE15 (n 49) .
 By way of example, in DRO17 (n 50)  the Court found the Tribunal to have erred where it ‘assessed the future prospects of harm simply on the basis of what view the authorities had at [the time of his departure]. There was, in other words, no reasonable speculation as to what might occur in the future’. Cf 1005911  RRTA 923, where the Tribunal acknowledged, and avoided, the potential for error. The Tribunal found, at  and , that the applicant’s actions in returning to his country of former habitual residence and his delay in applying for protection in Australia were inconsistent with him facing ‘imminent harm’ at that time, but went on to state at : ‘However the Tribunal must look, not merely to the immediate, but to the reasonably foreseeable future.’ The Tribunal ultimately concluded that the applicant would face persecution on cumulative (largely socio-economic) grounds, including some forms of harm that would manifest further into the future such as difficulty finding work which might compromise his ability to subsist.
 In some cases, lack of credibility or subjective fear on this basis formed at least part of the reasons for rejection. While in three of these cases it appears that the applicant claimed to fear imminent harm, in the rest the Tribunal itself introduced the concept of imminence or immediacy: see, eg, V97/08193  RRTA 196; N01/39734  RRTA 137; N02/42403  RRTA 272; 0800904  RRTA 99; 071959605  RRTA 256; 0808521  RRTA 270; 1000095  RRTA 371; 1009089  RRTA 17; 1103103  RRTA 592; 1101896  RRTA 401; 1101401  RRTA 620; 1205487  RRTA 710; 1103936  RRTA 857; 1218579  RRTA 741; 1403634  RRTA 230; 1305041  RRTA 378; 1406956  RRTA 894; 1301484  RRTA 98; 1600083 (Refugee)  AATA 3066; 1611485 (Refugee)  AATA 4213; 1514886 (Refugee)  AATA 3849; 1514689 (Refugee)  AATA 3629.
 1514886 (Refugee) (n 54)  (emphasis added). In this case, while noting that it was not necessarily determinative, the Tribunal considered that the applicant’s ‘migration history significantly weaken[ed] his claims to have experienced persecution or significant harm in Fiji ... ; and to fear such harm if he returns there in the future’: at .
 1101896 (n 54) –.
 See, eg, 1617430 (Refugee)  AATA 4977; 1602952 (Refugee)  AATA 4556; 1608643 (Refugee)  AATA 3630; 1500744 (Refugee)  AATA 3094; 1501478 (Refugee)  AATA 2732; 1616021 (Refugee)  AATA 2204; 1606474 (Refugee)  AATA 2216; 1612947 (Refugee)  AATA 2681; 1606330 (Refugee)  AATA 3187; 1507725 (Refugee)  AATA 3775. All but one of these cases was decided by the same decision-maker.
 1507725 (Refugee) (n 57).
 Ibid .
 Ibid .
 In at least 23 cases, the Tribunal’s reasoning reveals that the departmental delegate had raised an issue relating to lack of immediacy of the applicant’s fears. Of these 23, the Tribunal nevertheless went on to make a positive decision in 12 cases. In four cases it rejected the claim, at least in part, on the basis of delay or return to country of origin; in others the rejection related to other reasons. See also AYT15 v Minister for Immigration and Border Protection  FCCA 688; BPC16 v Minister for Immigration and Border Protection  FCCA 1140.
 1504740 (Refugee)  AATA 868,  (emphasis added).
 1407181, 1407173  RRTA 140, .
 See, eg, BDT16 v Minister for Immigration and Border Protection  FCA 249; SZVSE v Minister for Immigration and Border Protection  FCA 1435; DTJ16 v Minister for Immigration and Border Protection  FCCA 2049; DFE16 v Minister for Immigration and Border Protection  FCCA 308; (2017) 317 FLR 215; FIG17 v Minister for Immigration and Border Protection  FCCA 3751. But see Applicant MZKAO v Minister for Immigration and Multicultural and Indigenous Affairs  FCA 1484, , where the Federal Court did expressly consider and approve of the Tribunal’s reasoning:
This point can be dealt with shortly. It is clear that the appellant’s voluntary return to India on three occasions is relevant to assessing whether and to what extent the appellant fears persecution or ‘serious harm’ in India. The fact that the appellant was prepared to return there more than once strongly suggests that the fear he held, if any, was either not a fear of immediate harm or not a fear of treatment that would amount to ‘serious harm’.
 See generally Hathaway and Foster (n 6) 100–2, 150–4.
 See, eg, Re X (Immigration and Refugee Board of Canada, Refugee Appeal Division, RAD File No TB7-24034, G Erauw, 6 November 2018)  CanLII 143642, ; Pang v Bureau of Citizenship and Immigration Services,  USCA2 179; 448 F3d 102 (2nd Cir, 3 May 2006) 104 . In NZ, see
CX (China)  NZIPT 801141 ; Refugee Appeal No. 75402  NZRSAA 40; Refugee Appeal No. 73504  NZRSAA 90.
 See, eg, 0908758  RRTA 227; N01/41067  RRTA 1127; N06/53069  RRTA 88; N99/30512  RRTA 404; 1213207  RRTA 1078; 1610107 (Refugee)  AATA 1429; 1514117 (Refugee)  AATA 1362; 1511167 (Refugee)  AATA 1524; 1504584 (Refugee)  AATA 650; 1606403 (Refugee)  AATA 2615; SZTBD v Minister for Immigration and Border Protection  FCCA 2182; BDF15 v Minister for Immigration and Border Protection  FCCA 3014; CCX15 v Minister for Immigration and Border Protection  FCCA 1307. See also, in four cases, the applicant referred to the imminence of a harm-causing event (war), rather than the harm itself: V00/11087  RRTA 92; N00/33037  RRTA 879; 0908758 (n 67); 1512717 (Refugee)  AATA 212.
 See, eg, N06/53217  RRTA 58; N03/45607  RRTA 961; N99/30760  RRTA 399; N01/36945  RRTA 536; N98/21019  RRTA 730; 1215644  RRTA 377; 1213201  RRTA 113; N02/45124  RRTA 1054; 1722737 (Refugee)  AATA 643.
 Department of Immigration and Border Protection, Procedures Advice Manual 3: Refugee and Humanitarian — Refugee Law Guides (1 July 2017) [7.2] (‘PAM 3’). These guidelines are not formally binding on decision-makers, but are generally followed unless inconsistent with the Migration Act or Regulations. In 1215016  RRTA 169,  the Tribunal regarded the following year as the ‘reasonably foreseeable future’, noting that such a timeframe was ‘not qualitatively any more speculative than an assessment of the situation at present.’ (citations omitted). See also 1214267  RRTA 168, . The Tribunal has its own, brief, guidance on the meaning and application of ‘reasonably foreseeable future’: Administrative Appeals Tribunal, Migration and Refugee Division (Cth), A Guide to Refugee Law in Australia (2020) 3-8–3-9.
 See, eg, BKD17 (n 50); DRO17 (n 50); AOO15 v Minister for Immigration and Border Protection  FCCA 2871; MZZXD v Minister for Immigration and Border Protection  FCCA 104 (‘MZZXD’); SZTFI v Minister for Immigration and Border Protection  FCA 322; (2015) 231 FCR 222; SZTUI v Minister for Immigration and Border Protection  FCCA 1667 (‘SZTUI’); Minister for Immigration and Border Protection v MZYTS (2013) 230 FCR 431 (‘MZYTS’); MZYXR (n 39); Minister for Immigration and Citizenship v SZQKB  FCA 1189; (2012) 133 ALD 495; MZYAY v Minister for Immigration and Citizenship  FCA 644; (2009) 109 ALD 498; SZGHS (n 52); NACZ v Minister for Immigration and Multicultural and Indigenous Affairs  FCA 457 (‘NACZ’).
 We found only one decision that referred to Mortimer J’s discussion of the ‘reasonably foreseeable future’ test in CPE15 (n 43), namely 1703914 (Refugee)  AATA 3088, .
 See Anderson et al (n 10).
 At the time of the UK decision, the election date had not yet been set. This was a crucial factor in the decision, even though the elections were predicted to take place the same year or the year following the decision:
If, after promulgation of this determination, evidence emerges that elections will be held at a particular time, without any of the safeguards and other countervailing features we have described, then the structures underpinning the country guidance system ensure that judicial fact-finders will be required to have regard to the new state of affairs, in reaching determinations on Zimbabwe cases.
EM and Others (Returnees) Zimbabwe CG  UKUT 98 (IAC), .
 Australian case law was raised by counsel at ibid :
Mr Henderson also relied upon the Court of Appeal’s judgment in Karanakaran  EWCA Civ 11. In particular, he drew attention to what the Court had to say in that case about the Australian decision of Sackville J in Minister for Immigration and Multicultural Affairs v [Rajalingam]  FCA 719 [(1999)  FCA 719; 93 FCR 220]. The thrust of the Australian decision, according to Mr Henderson, was that the decision maker in a case involving a claim to international protection must not foreclose reasonable speculation about the chances of a future hypothetical event occurring.
 Ibid .
 Ibid .
 SZGHS (n 52) . See also 1404682  RRTA 893, , where the Tribunal took into account risk attendant on a future election:
The Tribunal is mindful too that the applicants avoided harm between August 2010 and their departure to Australia in March 2013. It notes though Mr S remains a politician. Although the applicants did not make this claim and the migration agent did not submit it, the past harm the applicants experience[d] were temporal to election cycles in Sri Lanka in the beginning of (presidential) and mid (parliamentary) 2010. The Tribunal considers there is some chance and that is more than speculative or remote chance during any future election cycle, Mr S will again take steps to silence any perceived opponents, including the first named applicant. The Tribunal finds the first named applicant will continue to hold his pro-opposition, anti-government and anti-corruption political opinion. The Tribunal finds in the reasonably foreseeable future, the threat to the first named applicant from Mr S is ongoing if the applicants return to Sri Lanka, as may be presidential elections in early 2015.
 Guo (n 30) 574–5.
 QAAH (n 41) 391 . See also discussion above in Part III.
 V03/16047  RRTA 11. See also V03/16046  RRTA 15; V03/16249  RRTA 31; V03/16433  RRTA 349; V04/16763  RRTA 530; V03/16303  RRTA 33; V03/16260  RRTA 32.
 SZSZO v Minister for Immigration and Border Protection  FCCA 242, .
 AIE15 v Minister for Immigration  FCCA 451,  (emphasis in original). See also the approach in SZGHS (n 52) .
 QAAH (n 41) 391 : it was ‘necessary to estimate how confidently any non-Taliban settlement can be predicted to endure, on a widespread basis, for a period of some years.’ See also CDC15 v Minister for Immigration and Border Protection  FCA 18 (‘CDC15’).
 AUK15 v Minister for Immigration and Border Protection  FCCA 1872, . See also MZZXD (n 70)  and CDC15 (n 83) , in which the Court stated that decision-makers should have assessed protection claims by Hazaras from Afghanistan in light of the likely withdrawal of international troops in one to two years’ time, which would have an adverse effect on their security.
 MZYTS (n 70).
 Ibid 444  (emphasis added).
 NAGT of 2002 v Minister for Immigration and Multicultural and Indigenous Affairs  FCAFC 319, .
 NACZ (n 70) .
 See, eg, SZGHS (n 52) .
 Minister for Immigration and Multicultural Affairs v Rajalingam  FCA 719; (1999) 93 FCR 220, 239 . See also MZXSA v Minister for Immigration and Citizenship  FCAFC 123; (2010) 117 ALD 441, 460–61 ; EMX17 v Minister for Immigration and Border Protection  FCCA 284, . But this was extended to not foreclosing ‘reasonable speculation about the chances of the hypothetical future event not occurring’ in SZVKA v Minister for Immigration and Border Protection (2017) 320 FLR 453, 461  (emphasis in original).
 MZYTS (n 70) 443  (citations omitted). In that case, it was noted that a decision-maker could rely on information that was ‘several years old ... as part of a weighing process’ helping them decide ‘which information best and most reliably supports the prediction of future risk’: at 452 .
 Mok v Minister for Immigration, Local Government and Ethnic Affairs (No 1)  FCA 545; (1993) 47 FCR 1, 66 (‘Mok (No 1)’).
 Ibid. See also MOK (n 31).
 Mok (No 1) (n 93) 66. See also MOK (n 31).
 1001683  RRTA 506,  (emphasis added).
 We are grateful to Kate Jastram for this point. See also for more detailed discussion of these issues: Michelle Foster, International Refugee Law and Socio-Economic Rights: Refuge from Deprivation (Cambridge University Press, 2007).
 UNHCR, Guidelines on International Protection: Child Asylum Claims under Article 1(A)2 and 1(F) of the 1951 Convention and/or 1967 Protocol relating to the Status of Refugees, UN Doc HCR/GIP/09/08 (22 December 2009) 15  (emphasis added).
 1703914 (Refugee) (n 71) .
 Mok (No 1) (n 93) 66. See also MOK (n 31).
 1703914 (Refugee) (n 71) . See also 1319201  RRTA 835,  (‘1319201’); 1314106  RRTA 796,  (‘1314106’).
 This was, in effect, acknowledged in AF (Kiribati) (n 12) , where the Tribunal observed that
the concept of an ‘imminent’ risk to life ... requires no more than sufficient evidence to establish substantial grounds for believing the appellant would be in danger. In other words, these standards should be seen as largely synonymous requiring something akin to the refugee ‘real chance’ standard. That is to say, something which is more than above mere speculation and conjecture, but sitting below the civil balance of probability standard.
 A combination of the IPCC’s evaluation of the underlying scientific evidence, and the degree of expert agreement about it, determines the level of ‘confidence’ ascribed, which ranges from ‘very low’ to ‘very high’. Confidence in the ‘validity of a finding [is] based on the type, amount, quality and consistency of evidence (e.g., mechanistic understanding, theory, data, models, expert judgment) and on the degree of agreement’: Intergovernmental Panel on Climate Change (‘IPCC’), Climate Change 2014: Synthesis Report (Report, 2014) 121. The term ‘virtually certain’ means that the probability is 99–100%: at 2.
 Ibid 8.
 Ibid 15.
 Ibid 13.
 Ibid 16.
 Ibid 26 (emphasis omitted).
 Ibid 20 (emphasis omitted).
 While, in some cases, uncertainty about mitigating factors has led decision-makers to deny the protection claim, in others, it has worked in the applicant’s favour — while conditions may improve, they also may not (and the applicant should, in effect, be given the benefit of the doubt): see 1515485 (Refugee)  AATA 724; 1605348 (Refugee)  AATA 785.
 We are reminded of Lord Diplock’s cautionary statement in the 1971 extradition case of Fernandez v Government of Singapore  2 All ER 691, 696–7, in which his Lordship explained that:
There is no general rule of English law that when a court is required, either by statute or at common law, to take account of what may happen in the future and to base legal consequences on the likelihood of its happening, it must ignore any possibility of something happening merely because the odds on its happening are fractionally less than evens. ... The degree of risk should be an important factor in the court’s decision, whether it is more or less than 50 per cent.
 Hélène Lambert and Theo Farrell, ‘The Changing Character of Armed Conflict and the Implications for Refugee Protection Jurisprudence’ (2010) 22(2) International Journal of Refugee Law 237, highlighting that the indirect effects often are more deadly on the civilian population (ie, refugees) than the direct effects of armed conflict. See also UNHCR, Guidelines on International Protection No. 12: Claims for Refugee Status related to Situations of Armed Conflict and Violence under Article 1A(2) of the 1951 Convention and/or 1967 Protocol relating to the Status of Refugees and the Regional Refugee Definitions, UN Doc HCR/GIP/16/12 (2 December 2016), 5  (‘Guidelines on Armed Conflict’).
 Guidelines on Armed Conflict, UN Doc HCR/GIP/16/12 (n 112) 6 .
 Theo Farrell and Olivier Schmitt, ‘The Causes, Character and Conduct of Internal Armed Conflict and the Effects on Civilian Populations, 1990–2010’ in Volker Türk, Alice Edwards and Cornelius Wouters (eds), In Flight from Conflict and Violence: UNHCR’s Consultations on Refugee Status and Other Forms of International Protection (Cambridge University Press, 2017) 25. See also Paul Collier, The Bottom Billion: Why the Poorest Countries Are Failing and What Can Be Done About It (Oxford University Press, 2007).
 See, eg, 1215343  RRTA 240, ; 1213298  RRTA 43, ; 1217298  RRTA 81, .
 1703914 (Refugee) (n 71) . See also DPM16 (n 19) : ‘The fluidity of the political situation was directly relevant to that chance [of harm].’ This approach is also sometimes applied in contexts other than armed conflict: see 1601317 (Refugee)  AATA 2740.
 1314106 (n 101) . See also 1319201 (n 101)  containing the same reasoning by the same decision-maker in relation to the risk from Islamic State in Pakistan.
 PAM3 (n 69) [7.2].
 Guidelines on Armed Conflict, UN Doc HCR/GIP/16/12 (n 112) 6 .
 Minister for Immigration and Multicultural Affairs v Jama  FCA 1680, , .
 Ibid  (Branson and Sackville JJ). See also DUX16 (n 19) .
 In CDC15 (n 83) , Charlesworth J noted that:
The predictive and speculative function of the Tribunal could not be properly discharged without assessing and dealing with the appellant’s assertions that the troops would withdraw as predicted, that the Taliban would assume increased control as predicted, and that the chance of him being persecuted, or the risk that he would suffer serious harm, would be elevated as a result.
See also MZZXD (n 70) ; 1215016 (n 69) .
 SZTUI (n 70) ; 1220489  RRTA 292, .
 See Linda Kirk, ‘The Concept of Imminence in Refugee Law: Administrative Appeals Tribunal (Migration and Refugee Division)’ (Conference Paper, Expert Workshop on the Concept of ‘Imminence’ in the International Protection of Refugees and Other Forced Migrants, 20 August 2018) . See also 1703914 (Refugee) (n 71) . In the context of art 1C(5) of the Refugee Convention, it has been stated that the contemplated change in circumstances must ‘have some degree of permanency, at least in the context of future foreseeable events’: V04/17240  RRTA 52.
 See 1703914 (Refugee) (n 71) . See also other cases where the future possibility of improving circumstances did not affect the finding of risk based on the existing circumstances: 1515485 (Refugee) (n 110); 1605348 (Refugee) (n 110).
 V04/17240 (n 124) (emphasis added). It could be argued that referring to the future of a ‘nation’ may imply a year-long timeframe, given how long it can take for systemic changes such as those outlined here to be realised. See also V03/16047 (n 80), which involved the same decision-maker and same reasoning.
 1008364  RRTA 1135, .
 Ibid .
 0903707  RRTA 758.
 Ibid .
 Ibid .
 Ibid . Cf, 1200203 concerning an applicant from Japan who feared future nuclear disasters, which would exacerbate his existing physical and psychological medical conditions. The Tribunal was not satisfied that he faced serious harm for a Refugee Convention reason were he to return to Japan in the reasonably foreseeable future: 1200203  RRTA 145.
 See Foster (n 97) 192–3.
 See 1314106 (n 101) ; 1319201 (n 101) .
 See generally Anderson et al (n 10).