Court of Appeal upholds challenge to Rwanda removals policy – an extended look

19 July 2023 by Jonathan Metzer

R ((AAA) Syria and Ors) v Secretary of State for the Home Department [2023] EWCA Civ 745

The Claimants in this case are 10 individual asylum-seekers from Syria, Iraq, Iran, Vietnam, Sudan and Albania who entered the UK irregularly by crossing the English Channel in small boats, together with one charity, Asylum Aid.

The Home Secretary (whose counsel included Neil Sheldon KC and Natasha Barnes of 1 Crown Office Row) is the Defendant, in circumstances where decisions were made on her behalf not to consider the individuals’ asylum claims but to remove them to Rwanda where their claims would be decided under the Rwandan asylum system, in accordance with arrangements between the two governments announced on 14^th April 2022 and contained in a Memorandum of Understanding and a number of diplomatic Notes Verbales, on the basis that Rwanda was a “safe third country” under the relevant provisions of the Immigration Rules.

There are also several interveners to the case, including the UN High Commissioner for Refugees (UNHCR) (whose counsel included Angus McCullough KC of 1 Crown Office Row).

In December 2022, the Divisional Court (Lewis LJ and Swift J) dismissed the general challenge brought by the Claimants to the policy of removing of asylum seekers to Rwanda for their claims to be determined there, finding it to be consistent with the Refugee Convention and other legal obligations on the Home Secretary, including those imposed by the Human Rights Act 1998. I discussed that decision here, including identifying what I consider to be weaknesses in that judgment.

The Court of Appeal, by a 2-1 majority (Sir Geoffrey Vos MR and Underhill LJ), has overturned this ruling on the issue of whether Rwanda is a safe third country (whilst unanimously dismissing the other grounds).

In particular, the majority held that the evidence which was before the Divisional Court indicates that there are deficiencies in the asylum system in Rwanda which give rise to substantial grounds for believing that there is a real risk that persons sent there will be returned to their home countries despite, in fact, having a good claim for asylum. Therefore, even though it was accepted that the assurances given by the Rwandan Government were made in good faith and were intended to address any defects in its asylum processes, it was held that the evidence did not establish, as a matter of substance, that the necessary changes had been reliably effected, or would have been at the time of the proposed removals. For these reasons, sending anyone to Rwanda would constitute a breach of Article 3 of the European Convention on Human Rights (EHCR), with which the Home Secretary must comply (under s. 6 of the Human Rights Act 1998).

In his dissenting judgment, Lord Burnett LCJ would have found for the Home Secretary.

This article will not be able deal with every issue which was considered and will mainly focus on the ground on which the Appellants succeeded.

The Judgment of the Master of the Rolls

The Master of the Rolls noted important parts of the relevant legal background, including a case which would prove to be central to the reasoning of the majority: Soering v United Kingdom (1989) 11 EHRR 439. In this case, the European Court of Human Rights identified, stated the Master of the Rolls, what is “widely taken as establishing the well-known test that the court should ask itself whether there were substantial grounds for believing that the persons being removed would face real risks of article 3 mistreatment”, which may be summarised as a question of whether there are “substantial grounds for believing” that the individual “would be faced in the receiving State by a real risk of exposure to inhuman or degrading treatment” (para 29; citing Soering, para 88).

The core reasoning in his judgment allowing the appeal was as follows:

1. The Divisional Court had asked itself the wrong question. Whilst it considered the Soering test in relation to conditions in Rwanda generally, in relation to other matters, such as the asylum system in Rwanda and the likelihood of the Rwandan Government’s assurances being realised, “it asked itself whether the Home Secretary had been entitled to reach the conclusions she did”. This “is a different question from whether the court assesses that there were in fact substantial grounds for thinking there was a real risk of article 3 mistreatment” (para 75). However, the Soering test “requires the court to reach its own conclusion” (para 76). As such, the Divisional Court did not ask the correct Soering question, “namely whether there were substantial grounds for thinking that asylum seekers sent to Rwanda under the MEDP but face a real risk of article 3 mistreatment” (para 78).
   
2. As to whether the evidence of the UNHCR, the events that occurred under the Israel/Rwanda agreement, the events that occurred at the Kiziba refugee camp in 2018 and Rwanda’s history of refoulement and of defects in its asylum system undermined the UK Government’s good faith opinion that Rwanda would honour its obligations, the Master of the Rolls held that the past and the present could not “be ignored or side-lined as the SSHD suggests”, as “the likelihood of promises being performed must, anyway in part, be judged by reference to what has happened in the past and the capacity and capability of the entity making the promises to keep them” (para 91) and that “the historical record described by the UNHCR, the significant concerns of the UNHCR itself, and the factual realities of the current asylum process in Rwanda” meant that there were substantial grounds for thinking that asylum seekers faced real risks of Article 3 mistreatment, as in practice “Rwanda can only deliver on its good faith assurances if it has control mechanisms and systems in place to enable it to do so”, but both history and the current situation “demonstrate that those mechanisms have not yet been delivered” (para 92).
   
3. In this regard, particular concern was expressed in relation to evidence provided by the UNHCR, including that: (1) Rwanda’s asylum process is marked by acute arbitrariness and unfairness, some of which is structurally inbuilt, and by serious safeguard and capacity shortfalls, some of which can be remedied only by structural changes and long-term capacity building; (2) the UNHCR’s evidence shows 100% rejection rates at Refugee Status Determination level for nationals of Afghanistan, Yemen and Syria, with an overall rejection rate for the 156 cases covered by the figures of 77%, together with further examples of recent expulsions, including of people arriving at Kigali airport to Libya, Afghanistan and Syria; (3) the detailed monitoring mechanisms are likely to come too late to affect the risk of the initial asylum seekers facing Article 3 ill-treatment and it was unclear whether they would account sufficiently for the approach taken to the granting of asylum thus far by the Rwandan Government (which had included stating that if an immigrant invokes an asylum claim as an alternative reason after failing to satisfy immigration entry requirements then deportation would continue whenever necessary, demonstrating a misunderstanding of the meaning of refoulement); (4) the training of those making asylum decisions in Rwanda would also come too late to affect the risk to the initial asylum seekers and “there was simply insufficient evidence to demonstrate that officials would be trained adequately to make sound, reasoned, decisions”; (5) as to the Rwandan legal system, in Government of Rwanda v Nteziryayo [2017] EWHC 1912 (Admin), the Divisional Court had concluded that “the evidence points to some risk, depending on the evidence before them and the safeguards in play, that judges might yield to pressure from the Rwandan authorities”, and a Foreign, Commonwealth and Development Office comment on a draft document before the Court had been that the Rwandan legal system “is not independent, is regularly interfered with and is politicised” and that “Opposition/political cases do not receive a fair trial or support”; (6) there was “some evidence” (indicated by an unofficial translation of a decision of the Israeli Supreme Court) that there were breaches of the agreement between Rwanda and Israel notwithstanding Rwandan Government assurances to the contrary; and (7) the events at the Kizibia camp in 2018 (where the UNHCR evidence was that the Rwandan police fired live ammunition on protesting refugees, killing at least 12 people, following which 66 refugees were arrested, many of whom were charged with offences) meant that “the UNHCR has grave concerns that asylum seekers under the MEDP would be at significant risk of harm and detention if they expressed dissatisfaction through protests in Rwanda” (paras 95-103).

For these reasons, the Master of the Rolls concluded that (at para 104):

… the problem with uncritical acceptance of the SSHD’s [Secretary of State for the Home Department’s] view that the unequivocal assurances in the MEDP [Migration and Economic Development Partnership] can wipe away all real risk of article 3 violations is that the structural institutions that gave rise to past violations remain in Rwanda today. The DGIE [Rwandan Directorate General of Immigration and Emigration] will still be responsible for asylum seekers arriving from the UK. It may have had some more training (though Mr Bottinick [the UNHCR’s Senior Legal Officer in the UK] describes that as being at “an extremely basic level”), but it is the same institution. The RSDC [Rwandan Refugee Status Determination Committee] will still decide asylum claims without the applicants being legally represented. The members of the RSDC may have learnt something since past violations, but it is impossible to be sure that they will be fair, when their processes are not attended by third parties. The appeals to the Minister and to the court are largely or, in the case of the court, completely untested. Rwanda is still, as the UK Government acknowledges, a one-party state which reacts unfavourably to dissent (see also Human Rights Watch’s public letter to the SSHD dated 11 June 2022, expressing its concerns about likely article 3 breaches). It is not an answer to say that Rwanda will have accepted the people sent under the MEDP, because the advanced information they will have about them will be limited and they may form adverse political opinions once there.

Therefore, “there were substantial grounds for thinking, bearing in mind the guarantees and assurances, that (a) Rwanda was not a safe third country, (b) there were real risks of refoulement or article 3 breaches, and (c) there were real risks that asylum claims would not be properly determined”, because the domestic asylum processes “will provide an insufficient safeguard, in particular for asylum seekers whose claims are rejected”, who “risk being returned either directly to their country of origin or indirectly through a third country”. In this regard, a “robust and effective asylum process in the receiving state is a necessary bulwark to mitigate against the risk of refoulement and related treatment” (paras 109-110).

The Judgment of Underhill LJ

In his judgment, Underhill LJ concurred in the result and provided a very detailed judgment, which included the following core reasoning:

1. The issue of whether the Divisional Court had asked itself the correct question “is not of crucial importance”, since “even if the Court asked itself the right question its answer was wrong”. However, if it were necessary to decide, he would, “albeit with hesitation”, accept the Claimants’ submission (para 129).
   
2. The three witness statements from Lawrence Bottinick (the UNHCR’s Senior Legal Officer in the UK), together with numerous exhibits, were considered to have been “evidently carefully prepared by reference to that experience, with as much detail and identification of sources as practicable”, such that there was “no good reason not to accept what they say on matters of fact except where there is cogent evidence to the contrary” (para 136).
   
3. Whilst the evidence of past instances of individuals seeking asylum in Rwanda being denied the opportunity to make claims for asylum (including in the context of the Israel/Rwanda arrangements in place between 2013 and 2018) did not, in itself, justify a conclusion that there was a real risk that individuals relocated under the MEDP would be subject to airport refoulement or otherwise denied access to the asylum process (para 153), Underhill LJ stated that “I do not accept that these episodes are of no relevance”, as these instances “are evidence of a culture of, at best, insufficient appreciation by DGIE officials of Rwanda’s obligations under the Refugee Convention, and at worst a deliberate disregard for those obligations, together with at least a suggestion of prejudice against asylum-seekers from the Middle East and Afghanistan and a willingness to take into account political considerations”, which then raised a question of whether this risk had been sufficiently mitigated (para 156).
   
4. Whilst there was nothing in the evidence that would justify a conclusion of bad faith on the part of the Government of Rwanda and there was no reason to suppose that it does not wish to ensure that individuals have their asylum claims determined fairly and effectively, “aspiration and reality do not necessarily coincide” (paras 261). In this regard, it was difficult in the absence of expert evidence and cross-examination (which would have been possible in the First-tier Tribunal had the Home Secretary not certified the human rights claims) to reach an answer, but it was concluded that “the Rwandan system for refugee status determination was not, as at the relevant date, reliably fair and effective”, in light of the evidence of the way that asylum interviews were conducted, the absence of any opportunity for a claimant to present their case to the RSDC through a lawyer, the evidence that the RSDC does not have sufficient skills and experience to make reliable decisions in claims of the relevant kind, the evidence that NGOs would be unlikely to have capacity to provide legal assistance during the administrative stage and the fact that the appeals process to the High Court was wholly untested and there were grounds for concern as to whether the culture of the Rwandan judiciary would mean that judges would be reluctant to reverse decisions that had been made (paras 263-264).

These overarching conclusions were based on findings which included the following:

1. The Government of Rwanda’s response did not indicate acceptance of a problem that had been identified in respect of asylum interviews being conducted in an unduly perfunctory and brief manner and it would not be a straightforward matter for officials to change the way that interviews were conducted (paras 166 and 169);
   
2. The evidence that claimants are not entitled to make representations to the RSDC through a lawyer had not been challenged by the Government of Rwanda, as confirmed by Sir James Eadie KC on behalf of the Home Secretary in oral submissions, which was “a serious defect in the process”, as “typically RIs will have neither the knowledge nor the articulacy to present their case to a non-specialist body, and still less where they will be having to do so through an interpreter” and “many RIs are likely to be especially vulnerable as a result of their experiences, which may include a history of torture”, such that the need to have a lawyer to present their claim “will be particularly important in such cases” (para 189);
   
3. Although it was now the intention of the RSDC to provide proper reasons for determination, it was “not possible to be confident that it will occur unless those drafting them receive proper training” (para 194);
   
4. The UNHCR’s evidence of claims from Syria, Yemen, Afghanistan and Eritrea being rejected indicated that there was a “surprisingly high rejection rate of claimants from known conflict zones, where UNHCR recommends against returns”, which “does indeed suggest a poor quality of decision-making”, and also constituted evidence, when taken with the evidence about the views expressed by senior Government of Rwanda officials that they should have sought asylum nearer to home that “RSDC members may hold such views and that they may influence its decision-making” (paras 199-200);
   
5. The main focus of UNHCR’s criticism appeared to relate to the absence of training, which was “fundamental” (para 206),as “Rwanda has comparatively little experience of assessing and determining individual asylum claims, and still less claims from claimants with the nationalities that are likely to be typical of those relocated under the MEDP”, and serious deficiencies had been identified which needed to be addressed by “particularly thorough and effective training”, in which regard the UNHCR evidence raised “a clear case to answer” as to whether the level of training would be sufficient, which had not been answered satisfactorily by the response (paras 246-248 and 259);
   
6. Whilst there was force in Sir James Eadie’s point that it was important that the evidence of the susceptibility of the Rwandan judiciary to pressure by the Government related to trials of political opponents rather than asylum appeals, this did not afford a complete answer, as the High Court would be tasked with deciding whether to overturn a decision of a Government Minister and a committee comprised of senior representatives of government bodies and agencies, such that “there must be a real risk that they will be generally reluctant to allow appeals against decisions of such bodies even in the absence of any specific pressure”, and further there may be cases in which the Government of Rwanda would have an interest in denying asylum to individuals of particular nationalities and there was no evidence as yet of how the appeal system has worked in practice as no appeal has yet been brought (paras 220-223);
   
7. It was unrealistic to suppose that individuals would be able to pay for legal assistance at the administrative stage of the process and the NGOs that the Government of Rwanda stated would provide pro bono assistance “would not with their present resources be able to provide legal assistance in the administrative stage of the RSD to any substantial further number of asylum-seekers relocated to Rwanda under the MEDP” (paras 234 and 238); and
   
8. Whilst those problems “could be resolved by making further changes to the process (e.g. allowing lawyers to make representations to the RSDC); by “capacity building” (e.g. as regards provision of legal assistance); and, importantly, by effective training of all those involved in the process (as noted at several points above)”, “the evidence is that those steps have not yet been taken or in any event not to the extent necessary to ensure the present fairness and reliability of the system” (para 265).

As to the assessment by the Home Office of the adequacy of the Rwandan asylum system, Underhill LJ accepted that “this is not a case where the Home Office was merely going through the motions of assessing the adequacy of the Rwandan asylum system” and there “were evidently dedicated civil servants genuinely trying to establish how the RSD process worked and to obtain assurances that addressed the perceived problems”, but stated that “perhaps as the result of the pressure of the timetable to which they were required to work, I believe that the officials in question were too ready to accept assurances which were unparticularised or unevidenced or the details of which were unexplored”, such as the late emergence of a problem about interpreters, noting that Independent Advisory Group on Country Information had criticised various aspects of the way in which the Asylum System CPIN was prepared, stating that there was limited critical information and fundamental gaps and unanswered questions (paras 267-268).

In light of the conclusions reached, Underhill LJ stated that it was not necessary to address the Claimants’ case that the repressive nature of the Rwandan regime meant that individuals would be at risk of inhuman and degrading treatment if they engaged in protests against or other criticisms of the Government of Rwanda, but it was doubted that the Divisional Court was correct that it would be a relevant factor if the Claimants themselves had not been shown to hold opinions of a kind likely to attract adverse attention (as this was a generic challenge), particularly where it was noted that it Home Office officials expressed concern about this very aspect (paras 289-292).

As to whether it was a breach of the Convention for a state to remove a person who has made an asylum claim to another country (even a safe country) without determining their claim and according them rights as a refugee if the claim is established, Underhill LJ held that “if the asylum-seeker will not face persecution or refoulement in the country to which they are returned they will have received the protection which the Convention is intended to afford them” and agreed with the conclusion of the Divisional Court that an obligation to determine every asylum claim on its merits had not been shown to be part of the obligations on states under the Refugee Convention and that to infer this “would go well beyond the limits of any notion of judicial construction of an international agreement” (paras 319-321). Further, he held that the Divisional Court was correct that removal of a person to a safe third country would not constitute an impermissible penalty within the meaning of Article 31 (1) of the Refugee Convention and that this would only be the case if removal amounted to a breach of Article 33 (the prohibition on refoulement) (paras 329-330).

As to procedural fairness, Underhill LJ did not agree with all aspects of the reasoning of the Divisional Court and also stated that it was important that the Home Office guidance makes it clear that the seven-day period for claimants to prepare representations “should not be treated as a norm” and “the grant of extensions is not necessarily exceptional”, but he concluded that the seven-day period did not render the decision-making process “structurally unfair and unjust”, as it would not be impossible for claimants to make effective representations in every case and the Home Office policy did countenance that extensions of time may be granted if reasonable grounds were shown (paras 434, 440-443 and 455).

The Dissenting Judgment

In his dissenting judgment, the Lord Chief Justice would have dismissed the appeal. As to the Rwandan asylum system, it was his view that “the terms of the agreement, the strong incentives on the Government of Rwanda to deliver its side of the bargain, the general scrutiny under which all decisions will be made and the strong monitoring arrangements in place lead to a conclusion that the risks of wrong or perverse decisions are also low” (para 517).

Whilst he shared the concerns identified by the UNHCR about whether those involved in the RSDC had sufficient training and expertise to deal appropriately with asylum claims and whether reported attitudes of scepticism towards claims made by Middle Eastern nationals would be influential, noting that there is “certainly evidence of poor practice”, the Lord Chief Justice would have held that the monitoring arrangements, both formal (by way of the levels of monitoring established under the Memorandum of Understanding) and informal (the latter stated to include “the reality that anyone removed to Rwanda, with their internet connected mobile phone, will be in a strong position to raise any personal concerns that they are not being treated in accordance with the agreement”), together with the reputation of Rwanda in the context of “this very high-profile public agreement” and what he considered to be “powerful financial incentives at work” in the form of substantial future aid support dependent on Rwanda’s compliance with its obligations, provided sufficient protection to drive good decision making (see paras 502-503, 505, 511 and 515).

Further, he agreed with Sir James Eadie’s proposed distinction between cases in which the Rwandan government had an interest in securing a conviction and asylum appeals, stating that the Government of Rwanda “had entered into a solemn agreement to abide by all its legal obligations regarding asylum claims, putting in place special features not hitherto available to other asylum applicants, and relies positively on the safeguard of an independent judicial process after the completion of the administrative determination of asylum claims”. As to this, he stated that there was “an obvious need for the judiciary of Rwanda in the High Court and Appeal Court to show its independence. One would expect it to do so.” (para 516)

However, he agreed with the majority that the claims should not have been certified and stated that the whole question of safety was “contestable” (para 480).

Comment

In my view, the reasoning of the majority is compelling in relation to the deficiencies in the Rwandan asylum system to date and the serious questions concerning whether these deficiencies have been properly remedied and whether the Rwandan judiciary is sufficiently independent to act as a safeguard.

Underneath the reasoning of the majority is the question of what is required for an asylum system which properly safeguards individuals against risks including the risks of refoulement and exposure to ill-treatment in breach of Article 3.

Part of what is suggested by the majority is that that there would need to be more thorough and effective training of decision-makers to address the deficiencies in the Rwandan asylum system which have been identified.

I would suggest that in view of the serious deficiencies identified in the Rwandan asylum system to date, any effective training must not merely teach decision-makers skills such as how to conduct interviews and draft decisions which address the evidence in front of them, but also incorporate the values which underlie the Refugee Convention – that, unless certain exceptions are met, if a person has a well-founded fear of persecution in their country of origin for reasons of race, religion, nationality, membership of a particular social group or political opinion, they deserve to receive protection.

Further, I would suggest that the training should include the principle that claims must be assessed through the prism of the potential difficulties that a person with a well-founded fear may have both in obtaining evidence to support their claim and in providing a cogent account of what has taken place. In the British legal context, these principles have been set out in authorities including the decision in Karanakaran v Secretary of State for the Home Department [2000] 3 All E.R. 449, which drew on a decision of the Federal Court of Australia in Minister for Immigration and Multicultural Affairs v Rajalingam [1999] FCA 719, where it was noted that “applicants with a genuine fear of persecution may not present as models of consistency or transparent veracity”.

It will be interesting to see how the Supreme Court decides this case if, as seems very likely (particularly in light of the dissenting judgment), permission to appeal is granted.

Jonathan Metzer is a barrister at 1 Crown Office Row and a member of the Editorial Team at the UK Human Rights Blog.

None of the barristers from 1 Crown Office Row instructed in this case were involved in writing this article.