AAA (Syria) & Ors, R (on the application of) v The Secretary of State for the Home Department

The Divisional Court (Lewis LJ and Swift J) decided, in essence, to reject all the generic challenges made in these proceedings to the policy of the Secretary of State for the Home Department (SSHD) to relocate certain asylum seekers to Rwanda. The 10 individual appellants and Asylum Aid have been given permission to argue some 22 grounds of appeal in this court, and seek permission to argue one more. We have been provided with thousands of pages of documents and authorities and heard 4 days of concentrated argument.

Yet, at its foundation, the issue we have to decide is short. It is, at its most basic, whether the Divisional Court was right to decide, if that is what it did decide, one fairly straightforward issue, bearing in mind the guarantees and assurances that the Government of Rwanda had given to the UK Government. The SSHD submits that the Divisional Court decided, in effect, that there were no substantial grounds for thinking that: (a) Rwanda was not a safe third country, (b) there were real risks of refoulement (asylum seekers being sent back to their home countries) or breaches of article 3 (article 3) of the European Convention on Human Rights (ECHR), and (c) there were real risks that asylum claims would not be properly and fairly determined in Rwanda. The question is whether that was right. There are, of course, other issues but that is the central one. Article 3 provides that “[n]o one shall be subjected to torture or to inhuman or degrading treatment or punishment”.

On 14 April 2022, the then Prime Minister announced the Migration and Economic Development Partnership (MEDP) with Rwanda, which he said would mean that anyone entering the UK illegally might be relocated to Rwanda. He said that “[t]he deal we have done is uncapped and Rwanda will have the capacity to resettle tens of thousands of people in the years ahead”. The MEDP comprises a Memorandum of Understanding of 13 April 2022 (MoU) and three Notes Verbales. The MoU has an initial term of 5 years. The Notes Verbales disclosed in these proceedings provide the guarantees of the Government of Rwanda regarding “the asylum process of transferred individuals”, and “the reception and accommodation of transferred individuals”.

The MEDP was developed to deter people from risking their lives in making dangerous journeys to the UK to claim asylum. These journeys are typically made by crossing the English Channel in small boats. They are often facilitated by people smugglers and criminal gangs, to whom asylum seekers pay considerable sums of money. The policy is a politically sensitive one which has attracted significant public and media attention. Notwithstanding that position, the case must be determined on the basis of the evidence and of accepted and familiar principles of public law. Nothing in this judgment should be construed as supporting or opposing any political view of the issues.

The appellants complain that the Divisional Court adopted the wrong tests, concentrating too much on whether the SSHD was entitled to reach the conclusions she did about the safety of Rwanda on the basis of four assessment documents which she published on 9 May 2022. On the same day, the SSHD had published her Inadmissibility Guidance to Home Office case workers outlining the powers available and the procedures to be followed to declare asylum claims inadmissible and remove asylum claimants to safe third countries. In this context, the appellants contend that the SSHD failed (a) to undertake a “thorough examination” of “all relevant generally available information” as required by the principles explained by the European Court of Human Rights (ECtHR) in Ilias v. Hungary (2020) 71 E.H.R.R. 6 (Ilias) at [137]-[141], and (b) to ask herself the right question and take reasonable steps to acquaint herself with the relevant information to enable her to answer it correctly as explained in Secretary of State for Education and Science v. Tameside Metropolitan Borough Council [1977] AC 1014 at 1064-1065 (Tameside). Specifically, the appellants submit that the SSHD took no or no proper account of (i) the fact that there was no independent judiciary in Rwanda, (ii) the collapse of a similar scheme entered into between the State of Israel and Rwanda, (iii) the Rwandan Government’s misunderstanding of the meaning of refoulement, and (iv) 15 areas of inadequacy in Rwanda’s current asylum process identified by the United Nations High Commission for Refugees (UNHCR). The appellants also contend that the Divisional Court failed properly to apply the test adumbrated in Soering v. United Kingdom (1989) 11 E.H.R.R. 439 (Soering) at [88] by asking, as it should have done, whether there were substantial grounds for believing that the asylum seekers sent to Rwanda would face real risks of article 3 mistreatment. In doing so, they submit that the Divisional Court failed to consider the guarantees and assurances given by the Rwandan Government in accordance with the principles established in Othman v. United Kingdom (2012) 55 E.H.R.R. 1 (Othman) at [186]-[189]. It failed, it is said, to assess the quality of the assurances, and whether, in the light of Rwanda’s existing practices those assurances could be relied upon, having regard to the 11 factors listed at [189] in Othman, or equivalent factors applicable in this case.

The UNHCR submitted that it had issued a rare unequivocal warning that there should be no transfers of asylum seekers to Rwanda, because of its clear view that the arrangement was incompatible with the UK’s obligations under the 1951 Convention on the Status of Refugees (the Refugee Convention). A Home Office memorandum of 22 February 2022 recognised the UNHCR’s expertise in relation to Rwanda. An 8 March 2022 Home Office email reported that the UNHCR was a critical part of assessing Rwanda’s safety, and that Rwanda depended heavily on the UNHCR for delivering its domestic asylum and refugee processes. Yet, the UNHCR submits that it was not consulted on the final terms of the MEDP and not involved in the formulation of the assurances given to the UK Government. The UNHCR submitted that MSS v. Belgium and Greece (2011) 53 EHRR 2 (MSS) at [349], and Lord Kerr in R (EM (Eritrea)) v. SSHD [2014] UKSC 12, [2014] AC 1321 (EM (Eritrea)) at [71]-[74], confirmed that special regard should be paid to the views of the UNHCR where it has special expertise and the subject matter is within its remit (see also R (Tabrizagh) v. SSHD [2014] EWCA Civ 1398 (Tabrizagh) at [18]-[20]). That principle applied in this case.

The UNHCR’s institutional conclusion was as follows:

I believe that Rwanda’s RSD [Refugee Status Determination] process is marked by acute unfairness and arbitrariness, 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. I believe that asylum seekers transferred to Rwanda are at serious risk of both direct and indirect refoulement and will not have access to fair and efficient asylum procedures, adequate standards of treatment or durable solutions, in line with the requirements set out in international refugee law.

The UNHCR submitted that, unless a current evaluation of Rwanda’s asylum system was irrelevant, the Divisional Court’s judgment could not stand.

The SSHD submitted, in effect, that that was precisely the position. She said that the MoU and the Notes Verbales did provide all the assurances that were required. There was no basis to doubt the good faith of the Rwandan Government’s assurances. The Divisional Court had been right to conclude that ex facie those assurances would be sufficient to avoid any risk of a breach of article 3. The MEDP was an entirely new arrangement. The existing systems and past bad practices were irrelevant to an evaluation of the reliability of these new assurances. The UK and Rwandan Governments had very strong vested interests in making the MEDP work in a way that was lawful and complied with the Refugee Convention and with the ECHR. Rwanda was a sovereign state signatory to the Refugee Convention, the 1984 United Nations Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (UNCAT) and was a member of the Commonwealth. There was no risk of refoulement where the Rwandan Government had specifically to consent in advance to each asylum seeker being sent to Rwanda. Moreover, the scheme would be carefully and independently monitored by the Joint Committee of the UK and Rwandan Governments and a Monitoring Committee comprised of people independent of the two Governments and the UNHCR itself. If things went wrong, they would come to light. It was a scheme that had been considered with unparalleled care and thoroughness by the UK Government. The fact finding by the Divisional Court was to be respected by this court (see DB v. Chief Constable of Police Service of Northern Ireland [2017] UKSC 7, [2017] 3 LRC 252 (DB) at [78]-[80] per Lord Kerr and R (Z) v. Hackney LBC [2020] UKSC 40, [2020] 1 WLR 4327 (Hackney) per Lord Sales at [56], [67] and [74]).

The appellants also made a connected submission that the SSHD had unlawfully certified Rwanda as safe for individuals under [17(c)] (paragraph 17(c)) of Part 5 of Schedule 3 to the Asylum and Immigration (Treatment of Claimants etc.) Act 2004 (the 2004 Act). The SSHD had, it was argued, created a presumption that Rwanda was safe in her assessment documents, and had, by certifying the asylum seeker’s claim to be clearly unfounded under [19(c)] of Part 5 of Schedule 3 to the 2004 Act, prevented them from appealing to the First-tier Tribunal (Immigration and Asylum), circumventing the statutory scheme. To make that certification, the SSHD had to form the opinion under [17(c)] that Rwanda was a place where the person’s life and liberty would not be threatened by reason of his race, religion, nationality, membership of a particular social group or political opinion, and from which the person would not be refouled. In the course of oral argument, the court asked whether that was a stricter test than the one mentioned at [2] above, namely that there were no substantial grounds for thinking there were real risks of ECHR breaches in Rwanda. I will return to that point. I note at this stage that section 94 of the Nationality, Immigration and Asylum Act 2002 allows the SSHD also to make that certification if she concludes under section 94(7)(b) that there is no reason to believe that the person’s rights under the ECHR would be breached in a third country.

The appellants also made four further main submissions that can be summarised as follows:-

The SSHD submitted that each of these four submissions was unfounded and the Divisional Court had been right on each of them for the reasons it gave.

On the crucial safety of Rwanda issues, I have determined that:

On the other issues, I have concluded in broad terms that the Divisional Court was right for the reasons given by Lord Justice Underhill.

I will now deal with the issues in the following order: essential factual background, the essential legal background, the reasoning of the Divisional Court, the issues, the discussion of those issues, and my conclusions.

This summary of the factual background is a brief summary of [6]-[11], and [15]-[35] of the Divisional Court’s judgment. Reference to those paragraphs should be made for the detail.

The SSHD declared the claims of some 47 asylum seekers to be inadmissible in May and June 2022, intending that they should be removed to Rwanda by charter flight on 14 June 2022. Some 32 claims for judicial review had been issued by the time of the Divisional Court’s judgment. On 10 June 2022, the Administrative Court refused an application for an interim injunction to prevent removal. The Court of Appeal dismissed an appeal, and the Supreme Court dismissed an application for permission to appeal. On 14 June 2022, three Claimants made applications to the ECtHR for interim measures. NSK, RM and HTN were granted that relief. The practical consequence was that no removals to Rwanda have yet taken place.

On 5 July 2022, the SSHD re-took all the inadmissibility, removal and some ECHR claims decisions affecting these appellants.

The UNHCR, as intervener, filed three witness statements of Mr Lawrence Bottinick, the High Commissioner’s Senior Legal Officer in the UK (Mr Bottinick). Both the SSHD and the Government of Rwanda were able to respond to that evidence.

It was agreed below that version 6.0 of the SSHD’s Inadmissibility Guidance of 9 May 2022 was the operative policy document. The purpose pursued is to encourage “… asylum seekers to claim asylum in the first safe country they reach and [to deter] them from making unnecessary and dangerous onward journeys to the UK”. The policy explained in the Inadmissibility Guidance excludes claims made by unaccompanied children, families and EU nationals. Its material part provides as follows:

If a case assessed as suitable for inadmissibility action appears to stand a greater chance of being promptly removed if referred to Rwanda (a country with which the UK has a [MEDP]), rather than to the country to which they have a connection, [the Third Country Unit] should consider referring the case to Rwanda. An asylum claimant may be eligible for removal to Rwanda if their claim is inadmissible under this policy and (a) that claimant’s journey to the UK can be described as having been dangerous and (b) was made on or after 1 January 2022. A dangerous journey is one able or likely to cause harm or injury.

The SSHD considers that the MoU and the Notes Verbales underpin her conclusion that Rwanda is a safe third country for the purposes of [345B]. [2] of the MoU sets out the objectives as follows:

The objective … is to create a mechanism for the relocation of asylum seekers whose claims are not being considered by the United Kingdom, to Rwanda, which will process their claims and settle or remove (as appropriate) individuals after their claim is decided, in accordance with Rwanda domestic law, the Refugee Convention, current international standards, including in accordance with international human rights law and including the assurances given under this Arrangement.

The MoU provides that a person may only be transferred to Rwanda with the agreement of the Government of Rwanda. Account will be taken of Rwanda’s capacity to receive persons and the administrative needs associated with their transfer. The UK provides the Rwandan Government with information on the persons it proposes to transfer. If it agrees, the Rwandan Government gives “access to its territory … in accordance with its international commitments and asylum and immigration laws”. Persons transferred are to be provided with accommodation and support “… adequate to ensure [their] health, security and wellbeing …”. Moreover, the Rwandan Government will have regard to information concerning (and accommodate) the special needs of a person transferred as a victim of modern slavery and human trafficking.

The MoU provides that the Rwandan Government will ensure that each person transferred will be treated and each asylum claim will be processed “in accordance with the Refugee Convention, Rwandan immigration laws and international and Rwandan standards, including under international and Rwandan human rights law, and including but not limited to ensuring their protection from inhuman and degrading treatment and refoulement”. Specific provisions are made in the MoU for access to interpreters, procedural or legal assistance at every stage of their asylum claim including on appeals, and access to an “independent and impartial due process of appeal in accordance with Rwandan laws”. The MoU makes specific provision as to the treatment of those granted and refused asylum, including the prevention of refoulement. Adequate support and accommodation are to be provided “until such time as their status is regularised or they leave or are removed from Rwanda”. The obligations under the MoU will survive its termination as regards those transferred under it.

The MoU provides for the Joint Committee to monitor and review the MEDP and to make non-binding recommendations, and for the Monitoring Committee to monitor the entire relocation process, and to report on conditions in Rwanda, the processing of asylum claims and the treatment and support provided to those transferred.

As regards the financial arrangements, the UK paid £20 million to the Rwandan Government on 29 April 2022 in respect of preparations to receive the first group of asylum claimants. Also in April 2022, the UK paid a further £120 million as an initial contribution to a fund intended to promote economic development in Rwanda. The UK will make further payments for the costs of processing claims, to ensure the safety and wellbeing of claimants, and for the costs for 5 years of welfare and integration for those who stay, and for 3 years for those who do not qualify for refugee status or humanitarian protection. The Note Verbale concerning these financial issues has not been disclosed. Finally, the MoU makes provision for management and protection of personal data transferred between the governments.

The four assessment documents published by the SSHD on 9 May 2022 contain her assessment of the safety of Rwanda as a safe third country for the purposes of [345B].

Each asylum seeker in these cases was detained upon arrival in the UK. Their English language skills and health were assessed. They were issued with a mobile phone, and given IT, welfare and legal representation information, including information on the free duty solicitor scheme. Each asylum seeker made an asylum claim, attended an asylum screening interview conducted by reference to a standard script and recorded on a standard form, a copy of which record was provided to the asylum seeker. The Home Office National Asylum Allocations Unit suspected in each case that the asylum seeker had spent time in a safe third country on their way to the UK, and referred their case to the Third Country Unit. That Unit considered an inadmissibility decision under [345A] and [345B], and issued each asylum seeker with a Notice of Intent. The Notice of Intent explained that they were being considered for removal to Rwanda, and sought their representations within 7 days for those detained.

After that period expired, the Third Country Unit (in Glasgow) and the Detained Barrier Casework Team (in Croydon) took the decisions in respect of the asylum seekers on behalf of the SSHD. They each issued a decision letter. The first dealt with inadmissibility, removal and certification under [17(c)]. The second dealt with ECHR claims. Removal directions also provided that, unless the asylum seeker left the UK voluntarily, they would be removed by plane to Kigali Airport.

In Soering, the ECtHR identified, in an extradition case, the test that the court should apply where it was removing a person to another state where it was alleged their article 3 rights would be violated. Soering 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. The ECtHR said this at [88]:

It would hardly be compatible with the underlying values of the Convention, that ‘common heritage of political traditions, ideals, freedom and the rule of law’ to which the Preamble refers, were a Contracting State knowingly to surrender a fugitive to another State where there were substantial grounds for believing that he would be in danger of being subjected to torture, however heinous the crime allegedly committed. Extradition in such circumstances, while not explicitly referred to in the brief and general wording of Article 3, would plainly be contrary to the spirit and intendment of the Article, and in the Court’s view this inherent obligation not to extradite also extends to cases in which the fugitive would be faced in the receiving State by a real risk of exposure to inhuman or degrading treatment or punishment proscribed by that Article [emphasis added].

In Ilias, the ECtHR explained the procedural duty of states considering the removal of asylum seekers to third countries without considering the merits of their asylum application. The case concerned the decision by Hungary to return to Serbia Bangladeshi asylum seekers who had arrived in Hungary from Serbia. Ilias has been widely understood as establishing that the removing state should undertake a thorough examination of all relevant generally available information before deciding whether to remove such a person. The ECtHR said this at [137]-[141]:

137.

Where a Contracting State removes asylum seekers to a third country without examining the merits of their asylum applications, however, it is important not to lose sight of the fact that in such a situation it cannot be known whether the persons to be expelled risk treatment contrary to art. 3 in their country of origin or are simply economic migrants. It is only by means of a legal procedure resulting in a legal decision that a finding on this issue can be made and relied upon. In the absence of such a finding, removal to a third country must be preceded by thorough examination of the question whether the receiving third country’s asylum procedure affords sufficient guarantees to avoid an asylum-seeker being removed, directly or indirectly, to his country of origin without a proper evaluation of the risks he faces from the standpoint of art. 3 of the Convention. Contrary to the position of the respondent Government, a post-factum finding that the asylum seeker did not run a risk in his or her country of origin, if made in national or international proceedings, cannot serve to absolve the state retrospectively of the procedural duty described above. If it were otherwise, asylum-seekers facing deadly danger in their country of origin could be lawfully and summarily removed to “unsafe” third countries. Such an approach would in practice render meaningless the prohibition of ill-treatment in cases of expulsion of asylum seekers.

138.

While the Court acknowledges the respondent Government’s contention that there are cases of abuse by persons who are not in need of protection in their country of origin, it considers that states can deal with this problem without dismantling the guarantees against ill-treatment enshrined in art. 3. It suffices in that regard, if they opt for removal to a third safe country without examination of the asylum claims on the merits, to examine thoroughly whether that country’s asylum system could deal adequately with those claims. In the alternative, as stated above, the authorities can also opt for dismissing unfounded asylum requests after examination on the merits, where no relevant risks in the country of origin are established.

(c)

Nature and content of the duty to ensure that the third country is “safe”

139.

On the basis of the well-established principles underlying its case-law under art. 3 of the Convention in relation to expulsion of asylum-seekers, the Court considers that the above-mentioned duty requires from the national authorities applying the “safe third country” concept to conduct a thorough examination of the relevant conditions in the third country concerned and, in particular, the accessibility and reliability of its asylum system. …

140.

Furthermore, a number of the principles developed in the Court’s case-law regarding the assessment of risks in the asylum-seeker’s country of origin also apply, mutatis mutandis, to the national authorities’ examination of the question whether a third country from which the asylum seeker came is “safe”.

141.

In particular, while it is for the persons seeking asylum to rely on and to substantiate their individual circumstances that the national authorities cannot be aware of, those authorities must carry out of their own motion an up-to-date assessment, notably, of the accessibility and functioning of the receiving country’s asylum system and the safeguards it affords in practice. The assessment must be conducted primarily with reference to the facts which were known to the national authorities at the time of expulsion but it is the duty of those authorities to seek all relevant generally available information to that effect. General deficiencies well documented in authoritative reports, notably of the UNHCR, Council of Europe and EU bodies are in principle considered to have been known. The expelling state cannot merely assume that the asylum-seeker will be treated in the receiving third country in conformity with the Convention standards but, on the contrary, must first verify how the authorities of that country apply their legislation on asylum in practice [emphasis added].

In Tameside at pages 1064-1065, the House of Lords established that a public body has a duty to carry out a sufficient inquiry prior to making its decision. Lord Diplock said at pages 1065A-B that “the question for the court is did the Secretary of State ask himself the right question and take reasonable steps to acquaint himself with the relevant information to enable him to answer it correctly?” (see the Divisional Court in R (Plantagenet Alliance Ltd) v. Secretary of State for Justice [2014] EWHC 1662 (Admin), [2015] 3 All ER 261 at [100] and [139], and the Court of Appeal in R (CAAT) v. International Trade Secretary [2019] EWCA Civ 1020, [2019] 1 WLR 5765 at [35] and in R (Friends of the Earth Ltd) v Secretary of State for International Trade/UK Export Finance (UKEF) [2023] EWCA Civ 14 at [57]).

Finally, the ECtHR considered in Othman how the court should deal with assurances received from a foreign Government as to the article 3 rights of someone considered for deportation to that foreign state. The case concerned the intended removal of Abu Qatada to Jordan, the circumstances of which were completely different from the present case. It is nonetheless useful to consider the guidance provided as follows at [186]-[189] of the ECtHR’s judgment:

186 … Before turning to the facts of the applicant’s case, it is therefore convenient to set out the approach the Court has taken to assurances in art. 3 expulsion cases.

187 In any examination of whether an applicant faces a real risk of ill-treatment in the country to which he is to be removed, the Court will consider both the general human-rights situation in that country and the particular characteristics of the applicant. In a case where assurances have been provided by the receiving state, those assurances constitute a further relevant factor which the Court will consider. However, assurances are not in themselves sufficient to ensure adequate protection against the risk of ill-treatment. There is an obligation to examine whether assurances provide, in their practical application, a sufficient guarantee that the applicant will be protected against the risk of ill-treatment. The weight to be given to assurances from the receiving state depends, in each case, on the circumstances prevailing at the material time.

188 In assessing the practical application of assurances and determining what weight is to be given to them, the preliminary question is whether the general human-rights situation in the receiving state excludes accepting any assurances whatsoever. However, it will only be in rare cases that the general situation in a country will mean that no weight at all can be given to assurances.

189 More usually, the Court will assess first, the quality of assurances given and, second, whether, in light of the receiving state’s practices they can be relied upon. In doing so, the Court will have regard, inter alia, to the following factors: (1) whether the terms of the assurances have been disclosed to the Court; (2) whether the assurances are specific or are general and vague; (3) who has given the assurances and whether that person can bind the receiving State; (4) if the assurances have been issued by the central government of the receiving state, whether local authorities can be expected to abide by them; (5) whether the assurances concerns treatment which is legal or illegal in the receiving state; (6) whether they have been given by a Contracting State; (7) the length and strength of bilateral relations between the sending and receiving states, including the receiving state’s record in abiding by similar assurances; (8) whether compliance with the assurances can be objectively verified through diplomatic or other monitoring mechanisms, including providing unfettered access to the applicant’s lawyers; (9) whether there is an effective system of protection against torture in the receiving state, including whether it is willing to co-operate with international monitoring mechanisms (including international human-rights NGOs), and whether it is willing to investigate allegations of torture and to punish those responsible; (10) whether the applicant has previously been ill-treated in the receiving state; and (11) whether the reliability of the assurances has been examined by the domestic courts of the sending/Contracting State [emphasis added].

The inadmissibility and removal decisions were made in the exercise of the powers in [345A] to [345D]:

Inadmissibility of non-EU applications for asylum

345A. An asylum application may be treated as inadmissible and not substantively considered if the Secretary of State determines that:

(i)

the applicant has been recognised as a refugee in a safe third country and they can still avail themselves of that protection; or

(ii)

the applicant otherwise enjoys sufficient protection in a safe third country, including benefiting from the principle of non-refoulement; or

(iii)

the applicant could enjoy sufficient protection in a safe third country, including benefiting from the principle of non-refoulement because:

(a)

they have already made an application for protection to that country; or

(b)

they could have made an application for protection to that country but did not do so and there were no exceptional circumstances preventing such an application being made, or

(c)

they have a connection to that country, such that it would be reasonable for them to go there to obtain protection.

Safe Third Country of Asylum

345B. A country is a safe third country for a particular applicant, if:

(i)

the applicant’s life and liberty will not be threatened on account of race, religion, nationality, membership of a particular social group or political opinion in that country;

(ii)

the principle of non-refoulement will be respected in that country in accordance with the Refugee Convention;

(iii)

the prohibition of removal, in violation of the right to freedom from torture and cruel, inhuman, or degrading treatment as laid down in international law, is respected in that country; and

(iv)

the possibility exists to request refugee status and, if found to be a refugee, to receive protection in accordance with the Refugee Convention in that country.

345C. When an application is treated as inadmissible, the Secretary of State will attempt to remove the applicant to the safe third country in which they were previously present or to which they have a connection, or to any other safe third country which may agree to their entry.

Exceptions for admission of inadmissible claims to UK asylum process

345D. When an application has been treated as inadmissible and either

(i)

removal to a safe third country within a reasonable period of time is unlikely; or

(ii)

upon consideration of a claimant’s particular circumstances the Secretary of State determines that removal to a safe third country is inappropriate the Secretary of State will admit the applicant for consideration of the claim in the UK.

[17] and [19] of Part 5 of Schedule 3 the 2004 Act provide as follows:

17 This Part applies to a person who has made an asylum claim if the Secretary of State certifies that—

(a)

it is proposed to remove the person to a specified State,

(b)

in the Secretary of State’s opinion the person is not a national or citizen of the specified State, and

(c)

in the Secretary of State’s opinion the specified State is a place—

(i)

where the person’s life and liberty will not be threatened by reason of his race, religion, nationality, membership of a particular social group or political opinion, and

(ii)

from which the person will not be sent to another State otherwise than in accordance with the Refugee Convention.

19 Where this Part applies to a person — …

(b)

he may not bring an immigration appeal from within the United Kingdom in reliance on an asylum claim which asserts that to remove the person to the State specified under paragraph 17 would breach the United Kingdom’s obligations under the Refugee Convention,

(c)

he may not bring an immigration appeal from within the United Kingdom in reliance on a human rights claim if the Secretary of State certifies that the claim is clearly unfounded, and

(d)

he may not while outside the United Kingdom bring an immigration appeal on any ground that is inconsistent with the opinion certified under paragraph 17(c).

I should record first our thanks to the Divisional Court for producing an impressive, comprehensive and carefully organised judgment at great speed. At [39], the Divisional Court summarised the generic issues that it had to decide. In the light of the submission that the Divisional Court applied the wrong test to its consideration of the safety of Rwanda, it is important to look at what it said about those issues. It described the appellants’ submission at [39(1)] as being that the SSHD’s conclusion that Rwanda was a safe third country was legally flawed. Their primary contention was that the SSHD’s assessment was contrary to article 3, based on (a) Ilias, (b) the fact that the asylum seekers would face a real risk of article 3 ill-treatment in breach of the Soering principle, and (c) the fact that it was inevitable that the policy would lead to occasional article 3 ill-treatment. Put another way, the Divisional Court described the submission as being that the conclusion that Rwanda was a safe third country had not taken account of relevant matters, was the result of insufficient enquiry, and rested on material errors of fact and was irrational. At [39(2)], the Divisional Court said that it was central to the appellants’ case that the asylum claims would not be determined effectively in Rwanda, running the risk that they would be refouled. The SSHD was not entitled to have confidence that the Rwandan Government would honour the MoU and the Notes Verbales.

At [41]-[42], the Divisional Court held (i) that any issue that went to the legality of decisions contained in the SSHD’s four assessments and the Inadmissibility Guidance was to be assessed as at the date of the inadmissibility decisions, and (ii) the correct focus was on the SSHD’s replacement decisions issued on 5 July 2022 rather than the earlier decisions made in May and June 2022. These decisions are not challenged in this court.

At [43]-[61], the Divisional Court dealt with the question of “whether the assessment that Rwanda [was] a safe third country [was] legally flawed”. At [43], the primary submission was described as being that the SSHD’s removal decisions under [345C] were unlawful because the conclusion that Rwanda was a safe third country under [345B] was legally flawed. The primary submission was said to be put in different ways on the basis that “the conclusion that Rwanda [met] the criteria at 345B”: (a) amounted to a breach of article 3 for the reasons explained in Ilias, (b) rested on material errors of fact or a failure to comply with Tameside obligations, (c) was an irrational conclusion, and (d) was part of a policy which was unlawful in the sense explained in Gillick v. West Norfolk and Wisbech AHA [1986] AC 112 (Gillick) in that it authorised removals in breach of article 3. It will be observed at once that this description of the submissions places the emphasis on the challenge to the conclusion reached by the SSHD as to the safety of Rwanda.

At [44], however, the Divisional Court recorded that it was also submitted that removal to Rwanda would be in breach of article 3 “in the sense of the Soering principle because there are reasonable grounds for believing that if a person is removed to Rwanda that will expose him to a real risk of article 3 ill-treatment because of the conditions in Rwanda”. It then said, importantly, at [45] that the legal arguments converged on two issues as follows:-

The Divisional Court gave its answers to these questions under four headings: (i) thorough examination and reasonable inquiries, (ii) adequacy of asylum system, (iii) the Gillick issue, and (iv) conditions in Rwanda generally.

The Divisional Court dealt first at [46]-[47] with the sources of information available to the SSHD including the UNHCR. It then dealt with Ilias, describing it as “an example of the application of the principle in Soering”, and gave a “relatively brief description of the Rwandan asylum procedure”. The description is not controversial and was as follows:

Rwanda has a significant history of providing asylum to refugees fleeing local conflict. In July 2020, the UNHCR reported that since 1990, Rwanda had maintained “an open door policy” to refugees from neighbouring countries, and that there were nearly 149,000 refugees in Rwanda. The overwhelming majority were from the Democratic Republic of Congo and the Republic of Burundi. Rwanda has also supported the UNHCR “emergency transport mechanism” which, since 2019, has assisted a little over 1,000 asylum seekers to be removed from Libya to Rwanda. Once in Rwanda, their claims are processed by the UNHCR and claimants have, to date, been resettled by the UNHCR in third countries. Mr Bottinick’s evidence was that at present, some 440 asylum claimants are in Rwanda under this scheme.

Persons who have fled to Rwanda from neighbouring countries have been permitted to remain in Rwanda without going through any formal asylum determination process. The Rwandan system for determining asylum claims has only been used to determine claims made by those coming from further afield. This is a small number of cases. The UNHCR estimated that in the last 3 years there have been approximately 300 cases. Asylum claims must be registered with the Directorate General of Immigration and Emigration (DGIE). The DGIE will interview the claimant, issue him with a residence permit and forward the case to the Refugee Status Determination Committee (RSDC). The RSDC comprises 11 members drawn from 11 ministries and government departments. Each holds his position ex-officio; membership of the RSDC will be only one part of the person’s overall responsibilities. The RSDC determines the asylum claim. There is a right of appeal to the Minister for the Ministry in Charge of Emergency Management, the government department with responsibility for, among other matters, refugee affairs. There is a further appeal from the Minister to the High Court of Rwanda. That is an appeal in the way of re-hearing.

The Divisional Court then summarised at [53]-[54] the appellants’ submission that the Rwandan asylum system was not adequate to prevent the risk of refoulement, derived from Mr Bottinick’s evidence as follows. The summary is controversial before us, but it is nonetheless important to recite it:

(1)

There are instances where the Rwandan authorities have refused to register claims for asylum. To the UNHCR’s knowledge there have been 5 occasions (involving claimants from Libya, Syria and Afghanistan) where a person has made an asylum claim to the DGIE, but the DGIE refused to accept the claim as a valid claim. Those claims were made at Kigali Airport in Rwanda and the asylum claimants were refused entry to and, ultimately were removed from Rwanda. Generally, Mr Bottinick is critical of the DGIE not just in terms of its approach to registering asylum claims but also when it comes to interviewing asylum claimants. He says the airport cases are an indication that the DGIE discriminates against those who are not nationals of neighbouring states and, especially, against persons from Middle Eastern countries. He says the DGIE has on other occasions refused to interview asylum claimants. He suggests the DGIE may discriminate against asylum claimants who are lesbian, gay, bisexual, trans-sexual or inter-sex. He says the UNHCR is aware of two such cases. Mr Bottinick also says that when the DGIE refuses to refer a claim to the RSDC it does not give reasons for its decision. When interviews do occur, he says no record of the interview is provided to the asylum claimant.

(2)

Mr Bottinick also considers the process before the RSDC is inadequate. The members of the RSDC are not expert or trained in asylum law. He gives examples of three occasions when the RSDC refused to see the asylum claimant. When hearings have taken place, they are too short to give claimants a fair chance to make their case, and hearings tend to lack focus because of the size of the RSDC. There are no interpreters at RSDC hearings which significantly prejudices claimants who speak neither French nor English. The RSDC does not allow claimants to be represented by lawyers. The RSDC does not provide proper reasons for decisions; decisions tend to be all in a standard form that simply informs the claimant of the outcome.

(3)

Mr Bottinick is sceptical about the value of the appeal to the Minister. He says the UNHCR is not aware of any case where the Minister has reversed a decision of the RSDC. He also points out that legal representatives are not available for appeals to the Minister. Ministerial decisions are also in standard form and are not properly reasoned.

(4)

Mr Bottinick also says that the lack of reasoned decisions from the RSDC and the Minister impedes effective use of the right of appeal to the High Court. This right of appeal was introduced in 2018. There is no evidence that any such appeals have been filed with or heard by the High Court.

(5)

Rwandan asylum law is said to be defective. Mr Bottinick refers to a “protection gap”. He says that the definition of “political opinion” in article 7 of Rwanda’s 2014 Law on Asylum does not cover the possibility of protection against persecution on grounds of imputed political opinion or from the risk of ill treatment by non-state actors.

(6)

Mr Bottinick’s opinion is that the Rwandan asylum system lacks the capacity and expertise necessary to deal effectively with asylum claims. This is material in two ways. Important aspects of asylum law may not be properly understood and properly applied. As an example, Mr Bottinick says that “it can be difficult for decision-makers to understand” that asylum claims should not be denied on the premise that the claimant could hide a characteristic protected under the Refugee Convention, such as his political opinion or sexual orientation. Further, the Rwandan system will not be able to cope with the volume of claims generated by the MEDP. Mr Bottinick comments that claimants in the Rwandan asylum system have insufficient access to legal assistance and interpretation services are not available. He also raises a concern that details of asylum claimants and their claims may not have been treated as confidential and information may have been passed to the asylum claimants’ countries of origin.

The Divisional Court then recorded at [55]-[56] that the SSHD and the Rwandan Government disputed much of Mr Bottinick’s evidence, noting only that: (i) the UNHCR had described the 2014 Law relating to Refugees in its July 2020 Universal Periodic Review as “fully compliant with international standards”, (ii) article 7 of the 2014 Law exactly followed the language of article 1 of the Refugee Convention, and (iii) as to whether Rwandan authorities had maintained the confidentiality of asylum seekers, it was satisfied that the RSDC had sought information from Rwandan embassies abroad, rather than countries of origin.

The Divisional Court went on to record at [57]-[58] that the SSHD’s primary reliance was on the detailed contents and assurances in the MoU and the Notes Verbales (summarised at [21]-[25] above and in [18]-[27] of the Divisional Court). Those, together with the steps she had taken to investigate the matters in the assessment documents, were sufficient to satisfy her Ilias and Tameside duties, “and permitted her rationally to conclude that Rwanda does meet the criteria at [345B(ii) to (iv)] to be a safe third country”.

At [59]-[60], the Divisional Court concluded that the SSHD had complied with her Ilias obligations. The assessment documents were a thorough examination of all relevant generally available information, including that emanating from the UNHCR. At [61], the Divisional Court concluded that the SSHD had complied with her Tameside obligations.

The Divisional Court described the next question at [62] as being “whether the [SSHD] was entitled to conclude that there were sufficient guarantees” to ensure proper determination of asylum claims, the absence of a risk of refoulement, and that Rwanda was a safe third country within [345B(ii)-(iv)]. That, it said, raised “the question of whether she was entitled” to rely on the assurances provided by the Rwandan Government.

The Divisional Court began by noting that, on their face, the assurances addressed all the UNHCR’s significant concerns. It then dealt with Othman, noting that the ECtHR’s list of criteria at [189] was intended to be neither prescriptive nor exhaustive. Rather, when the risk of article 3 ill-treatment was in issue, “the court’s approach must be rigorous and pragmatic notwithstanding that ultimately it is an assessment to be undertaken recognising that the court must afford weight to the [SSHD’s] evaluation of the matter. That approach will rest on a recognition of the expertise that resides in the executive to evaluate the worth of promises made by a friendly foreign state”.

The Divisional Court then said at [64] that it had concluded that the SSHD “was entitled to rely on the assurances contained in the MoU and Notes Verbales” for the reasons given at [64]-[71], which I can summarise as follows:

The Divisional Court relied on R (A) v. SSHD [2021] UKSC 37, [2021] 1 WLR 3931 for the proposition that the relevant question was whether the Inadmissibility Guidance positively authorised or approved unlawful conduct. It did not, because the SSHD could lawfully conclude that Rwanda was a safe third country and that asylum claims made in Rwanda would be effectively determined.

The Divisional Court said at [73] that it would consider “the wider Soering submission, that persons removed to Rwanda under the terms of the MEDP [were] exposed to a real risk of article 3 ill-treatment … by reason of conditions in Rwanda, generally”. It dealt with the arguments that the Rwandan authorities were intolerant of criticism, and that there might be an extreme response to criticism of the asylum seekers’ conditions or treatment, or to opposing political opinion.

The Divisional Court did not think that any direct inference could be drawn from the events at Kiziba refugee camp in 2018, where protesters were killed, because the circumstances were unlikely to be repeated for persons transferred under the MEDP, and the Rwandan authorities would abide by the terms of the MEDP. There was no evidence that any individual appellant held political opinions opposed to the Rwandan Government. If they did, that would be considered under [345B(i)].

At [77], the Divisional Court said that there was no force at all in the submission that there might be a breach of article 15 of the Refugee Convention regarding rights of association. The appellants’ reliance on various facts as to the repressive nature of the Rwandan Government was speculative, when the terms of the MEDP had been agreed and could be expected to be complied with. This overbore evidence that: (i) opportunities for political opposition in Rwanda were very limited and closely regulated, (ii) there were restrictions on the right of peaceful assembly, freedom of the press and freedom of speech, (iii) from a US State Department report of 2020, political opponents have been detained in “unofficial” detention centres and that persons so detained have been subjected to torture and article 3 ill-treatment short of torture, and (iv) prisons in Rwanda are over-crowded and the conditions are very poor. The terms of the MEDP meant that there was no real risk of these things affecting the individuals sent to Rwanda under the partnership.

I refer to the judgment of Underhill LJ for a summary of the Divisional Court’s conclusions on the four further submissions identified at [11] above.

It was common ground between the parties to the appeal that the central issue as to the safety of Rwanda was whether, bearing in mind the guarantees and assurances that the Government of Rwanda had given to the UK Government, there were no substantial grounds for thinking that: (a) Rwanda was not a safe third country, (b) there were real risks of refoulement or breaches of article 3, and (c) there were real risks that asylum claims would not be properly and fairly determined in Rwanda.

Before I answer that question, however, it is necessary to decide whether the Divisional Court actually determined that question or a different one. On one analysis, the Divisional Court seems, from some of the language it used, to have asked itself whether the SSHD was entitled to accept the assurances from the Rwandan Government or was entitled to conclude on the evidence that there were no article 3 risks for asylum seekers sent to Rwanda.

Accordingly, in my judgment, the first issue for this court is whether the Divisional Court asked itself the correct question as to the safety of Rwanda. If it did, as the SSHD submits it did, this court would only be entitled to interfere if it made some other error of law (see DB and Hackney referred to at [9] above).

I propose therefore to identify the following issues in two parts: (i) as to the safety of Rwanda and (ii) as to the remaining issues.

Issue 1: Whether the Divisional Court addressed the right question as to the safety of Rwanda.

Issue 2: Whether the Divisional Court was right to decide at [66] that it could only go behind the SSHD’s opinion that Rwanda would honour its obligations if there were compelling evidence to the contrary. If not, how should the court evaluate the reliability of guarantees and assurances given by one sovereign government to another?

Issue 3: Whether the Divisional Court was right to decide at [67]-[71] and [73]-[74] that (a) the evidence of the UNHCR generally, (b) what occurred when Israel and Rwanda made a similar agreement in 2013, (c) what occurred at the Kiziba refugee camp in 2018, and (d) Rwanda’s history of refoulement and of defects in its asylum system, did not undermine the SSHD’s opinion that Rwanda would honour its obligations.

Issue 4: Whether the Divisional Court was right to conclude at [73]-[77] that there was no real risk that the response of the Rwandan authorities to hostile political opinions expressed by asylum seekers in the future might subject them to article 3 ill-treatment?

Issue 5: Bearing in mind the correct test, whether the Divisional Court was right to decide at [64] that the SSHD was entitled to rely on the assurances contained in the MoU and Notes Verbales.

Issue 6: Whether there were in fact 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.

Issue 7: Whether the Divisional Court was right to conclude at [71] that the SSHD’s decision that, for the purposes of the criteria at [345B(ii) to (iv)], Rwanda was a safe third country, was neither irrational, nor a breach of article 3.

Issue 8: Whether the Divisional Court was right to accept at [59] that the SSHD complied with the obligations identified in Ilias, namely to undertake a “thorough examination” of “all relevant generally available information”.

Issue 9: Whether the Divisional Court was right to accept at [61] that the SSHD complied with her Tameside duty to ask herself the right question and take reasonable steps to acquaint herself with the relevant information to enable her to answer it correctly.

Issue 10: Whether the Divisional Court was right at [72] to decide that the Inadmissibility Policy, including the possibility of removal to a safe third country, was not Gillick unlawful.

Issue 11: Whether the SSHD’s certification under [17(c)] was unlawful, because the asylum seekers’ ECHR claims were not clearly unfounded under [19(c)].

Issue 12: Whether the Divisional Court ought to have concluded that removal of asylum seekers to Rwanda under the MEDP was inconsistent with article 33 and/or constituted the imposition of a penalty contrary to article 31 and was, therefore, a breach of section 2 of the 1993 Act.

Issue 13: Whether the Divisional Court ought to have concluded that articles 25 and 27 of the Procedures Directive made the MEDP unlawful because they were still retained EU law.

Issue 14: Whether the Divisional Court ought to have decided that the SSHD had created a presumption that Rwanda was safe in her assessment documents, thereby circumventing the statutory scheme under Schedule 3 to the 2004 Act.

Issue 15: Whether the Divisional Court ought to have decided that the SSHD’s alleged breaches of the UK GDPR would, if established, invalidate the SSHD’s decisions under the MEDP.

Issue 16: Whether the Divisional Court ought to have decided that the decisions to treat the asylum application as inadmissible under [345A], to remove the asylum seeker to Rwanda under [345C] (having decided that Rwanda was a safe third country under [345B]), and to make a certification decision under [17(c)] to the effect that Rwanda was a safe country for the asylum seeker, were rendered unlawful by procedural unfairness.

It was submitted by the SSHD that it would be remarkable if such an experienced Divisional Court had, in fact, addressed the wrong question. It is also clear from the material parts of the Divisional Court’s judgment that it understood that the Soering test required the court to decide whether there were substantial grounds for believing that the asylum seekers sent to Rwanda would face real risks of article 3 mistreatment (see [39(1)], [44] and [73]). The issue is not, in my judgment, whether we might think that the Divisional Court understood the test it had to apply, but whether an objective reading of its judgment shows that it did.

Moreover, in considering the process adopted by the Divisional Court, I bear closely in mind the SSHD’s powerful submission that the Divisional Court had itself said, when it refused permission to appeal, that the ground of appeal was “based on a misreading of the judgment and a reference to one sentence not read in context”. The point that it and the SSHD made was that one of the two central questions posed by the Divisional Court at [45] was whether the SSHD could lawfully reach the conclusion that the arrangements governing relocation to Rwanda would not give rise to a real risk of refoulement or other article 3 ill-treatment. The Divisional Court said in its permission judgment, and the SSHD now submits, that the SSHD could only lawfully do that if there would be no risk of refoulement, and that that was the issue the court then considered at [46]-[71].

On analysis, however, it seems to me that the Divisional Court only considered the Soering test in relation to conditions in Rwanda generally at [73]-[77]. 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 SSHD had been entitled to reach the conclusions she did. It is obvious, I think, that the question of whether the SSHD was entitled (within a margin of appreciation) to reach a particular conclusion 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.

It is also not appropriate to express the Soering question as being whether the SSHD could lawfully reach the conclusion that there was a real risk of article 3 mistreatment, even if it is true that she could only lawfully reach that conclusion if she assessed there was no such real risk. The difference is between the court’s assessment of the SSHD’s decision (which is a normal judicial review question) and the Soering question, which requires the court to reach its own conclusion.

I can explain my reasoning as to the detail by reference to the text of the Divisional Court’s judgment, which I have already summarised at [35]-[51] above:

Accordingly, I conclude that as to the asylum system and everything except the general situation in Rwanda at [73]-[77], 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 wouldface a real risk of article 3 mistreatment. I note at this stage that I have reached the same substantive conclusion as Underhill LJ at [129] of his judgment on this point.

This conclusion means that this court must look again at the Soering question. Whilst proper respect must be given to [73]-[77], the question is not a compartmentalised one. Accordingly, this court must, as it seems to me, look at the situation in Rwanda for asylum seekers sent there in the round, against the backdrop, of course, of the governmental assurances received and the terms of the MEDP itself. I turn then to deal with the approach that the court should adopt to such assurances.

I have already summarised the approach that the Divisional Court adopted to these questions at [45]-[47] above. Since the court is going to have to consider the Soering question for itself, it may not be necessary to deal with this issue at length. The approach that the court must adopt towards assurances is set out in detail in the passage I have already cited at [32] above from Othman. In summary, the ECtHR made clear that whilst “assurances are not in themselves sufficient to ensure adequate protection against the risk of ill-treatment … [t]here is an obligation to examine whether assurances provide, in their practical application, a sufficient guarantee”. It is critical to note, in my opinion, that the ECtHR emphasised that “[t]he weight to be given to assurances from the receiving state depends, in each case, on the circumstances prevailing at the material time”. It will, as the ECtHR also said, “only be in rare cases that the general situation in a country will mean that no weight at all can be given to assurances”. This, in my judgment, is one of the more usual cases referred to by the ECtHR where the court should “assess first, the quality of assurances given and, second, whether, in light of the receiving state’s practices they can be relied upon”.

I do not find any great assistance from all of the 11 factors listed out in Othman and cited at [32] above. As the SSHD submitted and the Divisional Court said the factors were neither exhaustive nor prescriptive. Moreover they were directed specifically to the unusual circumstances of Othman, which are different to those in this case. That being said, some of the factors, particularly 7-11, have been useful to consider, by analogy with the present case, as a form of cross-check with my assessment.

It is, however, important, as the SSHD submitted, for the court to bear closely in mind the injunction of Lord Bingham in R v. SSHD ex p Yogathas [2002] UKHL 36, [2003] 1 AC 920 at [9] where he said “The first [important consideration] is that the Home Secretary and the courts should not readily infer that a friendly sovereign state which is party to the Geneva Convention will not perform the obligations it has solemnly undertaken”.

Considering that this court is deciding the Soering question afresh, it is not strictly necessary for us to consider whether the Divisional Court was right to decide at [66] that it could only go behind the SSHD’s opinion that Rwanda would honour its obligations if there were compelling evidence to the contrary: we have taken into account all the relevant material in reaching our judgments. Since the Divisional Court was right to regard the UK Government’s evaluation of the reliability of assurances as an important factor, it would certainly have needed clear evidence to “go behind” that opinion. The UK Government has, of course, huge experience of diplomatic relations with the Government of Rwanda. I will, in the light of my earlier conclusions, however, have to “assess first, the quality of assurances given and, second, whether, in light of the receiving state’s practices they can be relied upon” (Othman at [189]). In the first stage of that process, due weight must be given to the UK Government’s view of the diplomatic assurances it was receiving.

In this case, it is not suggested that the Rwandan Government’s detailed guarantees and assurances, contained in the MEDP, were not given in good faith. In those circumstances, it can be assumed that the Rwandan Government intended to comply with them. The question is only whether, in the light of the events on the ground, it was or was not likely that the assurances would be complied with. Put another way, that is the heart of the Soering question: assuming that the UK Government had received detailed and fulsome guarantees and assurances given in good faith by a foreign sovereign Government with whom relations were long-standing and friendly, were there substantial grounds for thinking there was a real risk that asylum seekers would face article 3 ill-treatment?

In many ways, this question encapsulates the core of the appeal. I shall deal first with the question of what weight the court should give to evidence from the UNHCR in answering the Soering question. The SSHD submits that the Divisional Court was right to say at [71] that the UNHCR’s opinion as expressed in Mr Bottinick’s statements and on instructions carried no overriding or pre-eminent weight.

I shall consider each of the main authorities that has been referred to in chronological order:

In my view, these authorities demonstrate that particular importance should be attached to the evidence and opinions from the UNHCR, even though that evidence is not necessarily decisive nor pre-eminent. It is obvious that the UNHCR’s evidence will be of greater weight when it relates to matters within its particular remit or where it has special expertise in the subject matter.

The Divisional Court said at [71] that the UNHCR’s opinion carried no overriding weight and had to be considered together with all the other evidence to decide whether, on the totality of that evidence, the SSHD’s opinion was undermined to the extent it can be said to be legally flawed. In my judgment, this view was in error insofar as the Soering test was concerned, since the focus there was not on any legal flaws in the SSHD’s decision, but on whether, on the basis of all available evidence, there were substantial grounds for thinking that asylum seekers would face real risks of article 3 ill-treatment in Rwanda. As I have explained, the two things are not the same.

Against that background, I turn to the question of whether (a) the evidence of the UNHCR, (b) what occurred under the Israel/Rwanda agreement, (c) what occurred at the Kiziba refugee camp in 2018, and (d) 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 SSHD submitted to us, in effect, that, in the light of the detailed guarantees and assurances in the MEDP and the UK’s longstanding relationship with Rwanda and its financial and other incentives to perform on its obligations, what happened in the past was of limited, if any, real significance. This was a key submission, because the UNHCR’s evidence was all directed to what had happened in the past and what was happening at the current time. The SSHD said that a predictive evaluation was needed, and that in such an exercise, it was not significant that things had been less than ideal in the past.

I do not accept that the past and the present can either be ignored or side-lined as the SSHD suggests. Of course, a predictive evaluation is required, and of course great weight will be given to the guarantees and assurances of the Rwandan Government. But 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. The SSHD acknowledges, in effect, that the Rwandan asylum system has some way to go if it is to perform according to the MoU and the Notes Verbales. But she submits that its capacity can and will be built and that Rwanda will not accept more asylum seekers than it can handle in strict accordance with the MEDP.

The balancing exercise that the court needs to undertake in making the Soering assessment I have described is multi-factorial and complex. I confess to having vacillated within the decision-making process. Ultimately, however, I have concluded that there were substantial grounds for thinking that asylum seekers sent to Rwanda under the MEDP, as at the date of the SSHD’s decision-making in these cases in July 2022, faced real risks of article 3 mistreatment. That is the consequence of 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. 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. Both history and the current situation demonstrate that those mechanisms have not yet been delivered. They may in the future be delivered but they are not, on the evidence, there now. I have read Underhill LJ’s impressive and comprehensive judgment on these points and broadly agree with it. As it seems to me, it supports the conclusions I have reached.

In identifying some of the most important features of the evidence in the following paragraphs, I should not be taken as having excluded any parts of the massive body of evidence from my evaluation. It is, as I have said, a multi-factorial exercise giving due weight to the Rwandan Government’s assurances, coming as they do from a party to Refugee Convention, the SSHD’s opinion that the assurances will be realised, and the evidence from other quarters, notably the UNHCR.

So far as the UNHCR is concerned, it is not disputed that it was entrusted in 1950 by the UN with the mandate to supervise the application of the Refugee Convention. The UNHCR has been present in Rwanda since 1993 and had at the time of its evidence 332 staff there. It plays no official role in Rwanda’s asylum system, and has been denied observer status at RSDC sessions. It does, however, fund and train non-governmental organisations working with the Rwandan asylum system. As the appellants submitted, a Home Office memorandum of 22 February 2022 recognised the UNHCR’s expertise in relation to Rwanda, and an 8 March 2022 Home Office email said that the UNHCR was a critical part of assessing Rwanda’s safety.

The UNHCR’s evidence was that 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”. The DGIE plays the role of gatekeeper to the entire system. Although the DGIE is not authorised in law to reject asylum claims, and despite Rwandan Government denials, it summarily rejected without reasons 8% of the 319 asylum claims of which the UNHCR was aware between 2020 and June 2022. There are serious deficiencies in the rights of an asylum seeker to be heard after a perfunctory 20-30 minute interview with the DGIE, from which lawyers and representatives are excluded. The process at all levels, before the DGIE, the RSDC and the Minister, is marked by an absence of representation, interpretation and written reasons. Where reasons are provided, they are often perfunctory and inadequate: the Court was shown evidence of RSDC decision letters, including several issued after the conclusion of the MEDP, which rejected asylum claims either without reasons, or with very slim reasons. For example, a standard response from the RSDC was that “Refugee Status requested was not granted because you don’t meet the eligibility criteria, and the reasons you provided during the interview were not pertinent”. There has not yet been an appeal to the court and there are concerns as to the willingness of the judiciary to find against the Rwandan Government. The same processes of DGIE, RSDC, and the Minister, as have proved defective in the past, are envisaged under the MEDP. In this respect, I agree with, and adopt, Underhill LJ’s analysis at [158]-[223], where he breaks down the stages of the Rwandan RSD process and explains why the UNHCR’s criticisms of that process merit more consideration than was given to them by the Divisional Court.

The UNHCR’s evidence shows 100% rejection rates at RSDC level for nationals of Afghanistan, Yemen and Syria, from which asylum seekers under the MEDP may well emanate (I note here that my conclusion is the same as Underhill LJ at [200]). The overall rejection rate for the 156 cases covered by the figures is 77%. Of the 34 recent asylum seekers from a country with close bilateral relations with Rwanda, three were peremptorily rejected by the DGIE, three were forcibly expelled to the Tanzanian border, two were instructed to leave Rwanda within days, and two more were threatened with direct expulsion to their country of origin. In addition, there have been 5 recent cases of expulsion of those arriving at Kigali airport: two Libyans removed from Kigali airport in February 2021, two Afghans chain refouled to Afghanistan on 24 March 2022, and one Syrian chain refouled to Syria on 19 April 2022. The cases of airport refoulement do not themselves demonstrate that there would be a real risk of identical methods of refoulement under the MEDP, since those asylum seekers arriving in Rwanda would have been pre-approved by the Government of Rwanda and would arrive on planned flights. These instances nonetheless illustrate, as described by Underhill LJ at [156], “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”. The Divisional Court was provided with a table of instances of at least 100 allegations of refoulement and threatened refoulement cited in the UNHCR’s evidence and in notes of meetings with the SSHD. The UNHCR does not accept that the guarantees will be sufficient to prevent the risk of these events occurring to those accepted by Rwanda under the MEDP.

The Rwandan Government’s responses to these contentions demonstrate its misunderstanding of the meaning of the concept of refoulement. The UNHCR says that its responses are “indicative of fundamental misunderstandings … of its obligations under the Refugee Convention and give no reason to believe that such practices will change”. In essence, the Rwandan Government suggests that it can reject asylum claims and expel people where they seek asylum using forged documents (as many do), or have not been granted an entry visa, or where domestic immigration law allows. The UNHCR informed the UK Government in the weeks prior to a meeting in Kigali on 25 April 2022 about three recent cases of the refoulement of two Afghans and one Syrian who were denied asylum and put on flights out of Rwanda. The Rwandan Government’s response to this claim was a demonstration of its misunderstanding of the meaning of refoulement. It said that deceitful travellers attempting to abuse its border openness are routinely intercepted, but if the immigrant invokes an asylum claim “as an alternative reason after failing to satisfy immigration entry requirements”, deportation will continue whenever necessary. Whilst it also says that such things will not happen under the MEDP, the concern of the UNHCR is as to what happens to those sent to Rwanda under that scheme, whose asylum claims are rejected.

The detailed monitoring mechanisms are likely to come too late to affect the risk of these initial asylum seekers facing article 3 mistreatment. Moreover, it is unclear that these monitoring mechanisms would account sufficiently for the approach to the granting of asylum taken up to now by the Rwandan Government, which I have outlined in the previous paragraph.

The same can be said for the training of those making asylum decisions in Rwanda. At the time of the Divisional Court hearing, and on the evidence before this court, there was simply insufficient evidence to demonstrate that officials would be trained adequately to make sound, reasoned, decisions. In this respect, I agree with Underhill LJ at [245]-[260].

The ultimate reliability of the safeguards in the Rwandan asylum system will depend on the promised ability of asylum seekers to appeal to the court. That is not to ignore the fact that the bulk of the claims will be determined by non-judicial means, but to reassert that access to the courts is a core component of the right of access to justice. The Divisional Court in Government of Rwanda v. Nteziryayo [2017] EWHC 1912 (Admin) considered the independence of the Rwandan judiciary in an admittedly different context, but at some considerable length. It concluded at [234]-[240] and [372]-[374] 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” (see also the findings below in that case recorded at [149]). An FCDO (Foreign, Commonwealth and Development Office) comment on an FCDO draft document entitled “review of asylum processing: Rwanda” from April 2022 said that Rwanda was: “… the country of contradictions. For these cases I agree the legal support is likely to be independent … unless it gets political. Which may be farther down the road when refugees are in a process [of] settlement and make demands on certain things. The Rwandan legal system is not independent, is regularly interfered with and is politicised. Opposition/political cases do not receive a fair trial or support”.

The SSHD submitted that the UNHCR’s evidence as to the Israel/Rwanda agreement was irrelevant. I do not agree. Mr Bottinick explained that it was illustrative of the “danger and suffering that are, in UNHCR’s view, liable to arise from the UK’s externalisation plan”. The UNHCR gathered the information it provided from interviews between 2015 and 2017 with those who had been sent to Rwanda. Arrivals under that arrangement were “routinely moved clandestinely to Uganda even if they were willing to stay in Rwanda”. Mr Bottinick concluded that the “UNHCR considers that the UK-Rwanda Agreement creates serious risks of (a) increased people smuggling, and (b) an increase in asylum seekers being exposed to dangerous journeys and life-threatening conditions”.

The UNHCR drew our attention to the decision of the Israeli Supreme Court in Sagitta v. Ministry of Interior Administrative Appeal 8101/15. We were provided with an unofficial translation. It is clear from [87] of that judgment that the court held that the agreement between Israel and Rwanda (which we have not seen) was “specific” insofar as it related to the rights granted to the deportees and included “an explicit undertaking of [Rwanda] according to which the deportees will enjoy human rights and freedoms and that the principle of non-refoulement shall be complied with”. Accordingly, this constituted some evidence that the breaches mentioned by the UNHCR occurred notwithstanding Rwandan Government assurances to the contrary. The Rwandan Government’s response to the allegations lacked substance. It merely said that Rwanda had entered into other transfer arrangements for the international protection of refugees that differed from the Israel/Rwanda scheme.

As regards the events at Kiziba camp in Rwanda, the UNHCR’s evidence was that in February 2018, about 700 Congolese refugees resident there had marched towards Karongi town and camped outside the UNHCR Karongi Field Office. The refugees were protesting against a 25% cut in food rations. Two days later, the Rwandan police fired live ammunition on protesting refugees killing at least 12 people. Between February and May 2018, 66 refugees were arrested, and many were charged with a range of offences including “spreading false information with intent to create a hostile international opinion against the Rwandan state” (article 451 of the Penal Code), “inciting insurrection or trouble amongst the population” (article 463 Penal Code), and participating in an illegal demonstration or public gathering (article 685 Penal Code). Human Rights Watch’s investigation found that the refugees were unarmed and that the Rwandan police had used excessive force. Although Rwanda’s National Commission for Human Rights expressed the view that the police responded as a last resort to a violent attack, the UNHCR has grave concerns that asylum seekers relocated under the MEDP would be at significant risk of harm and detention if they expressed dissatisfaction through protests in Rwanda.

In my judgment, the problem with uncritical acceptance of the SSHD’s view that the unequivocal assurances in the MEDP 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 will still be responsible for asylum seekers arriving from the UK. It may have had some more training (though Mr Bottinick describes that as being at “an extremely basic level”), but it is the same institution. The RSDC 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.

The application of the Soering test requires the court to find there are “substantial grounds” for thinking there is a real risk of article 3 breaches. I have concluded that the matters I have mentioned, weighed against the Rwandan assurances and the SSHD’s view, giving appropriate but not overriding weight to the evidence of the UNHCR, means that such substantial grounds existed as at July 2022. I am conscious that Underhill LJ has expressed the view at [132] that the correct date for our consideration of the generic issues was September and October 2022 when the hearing took place before the Divisional Court. I understand his reasoning, but it seems to me that this is an appeal from the generic decisions made by the Divisional Court which were taken in respect of the position as at July 2022. That said, I do not think that there was any material difference between the position as at the two dates, and I would make the same decision whichever of those dates was being considered.

In effect, this question has already been answered under the previous issue. I do not think that the Divisional Court was right to compartmentalise the Soering test. It would have been better if it had asked itself whether, having regard to all the Rwandan Government’s guarantees and assurances, the SSHD’s view that they would be complied with, and all the evidence including the evidence of the UNHCR, there were substantial grounds for thinking that asylum seekers sent to Rwanda under the MEDP would face any real risk of article 3 mistreatment. I also adopt Underhill LJ’s comments at [290]-[291].

Accordingly, in my view, the events at Kiziba, to which the Divisional Court referred at [74], were something that ought to have been taken into account generally in relation to the Soering test, rather than specifically in relation to the conditions in Rwanda generally.

I can fully understand why the SSHD might have thought in good faith that she could rely on the Rwandan Government’s assurances in the MEDP. The court has, however, now made an objective evaluation of the Soering test on the basis of evidence that either was or ought to have been available to her. On that basis, the SSHD ought not reasonably to have relied on the assurances in the MEDP.

The answer to this issue is now also clear. 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. I refer specifically to [273]-[286] of the judgment of Underhill LJ for his analysis of how the risk of refoulement may eventuate.

It is important to understand how the conditions on the ground justify the conclusion that Rwanda was not a safe third country for the purposes of refoulement and article 3. In effect, that is because where a country’s domestic asylum processes are deficient to the extent set out above, and in Underhill LJ’s judgment, that will provide an insufficient safeguard, in particular for asylum seekers whose claims are rejected. Those asylum seekers risk being returned either directly to their country of origin or indirectly through a third country. They will thereby face real risks, in circumstances where they should not have been returned at all. A robust and effective asylum process in the receiving state is a necessary bulwark to mitigate against the risk of refoulement and related ill-treatment.

It does not necessarily follow from what I have already said that the SSHD’s decision to accept the Rwandan Government’s assurances was either irrational or itself a breach of article 3, although it gives rise at least to a real risk of such a breach. In the light of the answer I have given to the Soering question, it is not necessary for the court to reach a conclusion on this issue and I would prefer not to do so.

The Divisional Court spent much time examining the Ilias question of whether the SSHD made a thorough examination of all relevant generally available information. As explained in Ilias itself, the duty to undertake a thorough examination is primarily a procedural one. Having determined the substantive question, I do not think it is necessary to decide the procedural one, which looks in detail at the process that the SSHD undertook, rather than the substantive issue that had to be resolved. Underhill LJ takes the same approach at [266]. The SSHD undoubtedly considered much of the material available to the court.

The Divisional Court also spent much time examining the Tameside question of whether the SSHD asked herself the right question and took reasonable steps to acquaint herself with the relevant information to enable her to answer it correctly. Again, like Underhill LJ (at [266]), I do not think that, in the light of the conclusions I have reached on the substantive question, it is necessary to answer this question.

The Divisional Court asked itself whether the Inadmissibility Guidance positively authorised or approved unlawful conduct. It held it did not, because the SSHD could have lawfully concluded that Rwanda was a safe third country, and that asylum claims would be effectively determined. In the light of the answer to the Soering question, it is not necessary to determine whether the Inadmissibility Guidance itself was also unlawful. If I had thought it necessary to decide the Gillick issue, I would have agreed with [296]-[301] of the judgment of Underhill LJ.

The SSHD, in fact, certified (i) under [17(c)] that in her opinion Rwanda was a place where the asylum seekers’ “life and liberty will not be threatened by reason of [their] race, religion, nationality, membership of a particular social group or political opinion”, and “from which the person will not be sent to another State otherwise than in accordance with the Refugee Convention”, and (ii) under [19(c)] that the asylum seekers’ human rights claims were clearly unfounded, meaning that they could not bring an immigration appeal from within the United Kingdom in reliance on those claims.

It follows from what I have already said that the court could not endorse the SSHD’s opinion under [17(c)] that Rwanda was a place where the asylum seekers would not be subjected to the risk of ECHR breaches or refoulement. In that respect, I agree with Underhill LJ at [302]. It also follows that I do not agree with the SSHD’s certification under [19(c)] that the asylum seekers’ ECHR claims were clearly unfounded.

As suggested at [10] above, the “clearly unfounded” test is a stricter one than the Soering test which asks whether there are substantial grounds for thinking there are real risks of article 3 mistreatment. In these circumstances, the SSHD ought not to have certified the appellants’ ECHR claims under [19(c)]. The result is that the claims ought to have been permitted to be pursued in the usual way through the Tribunals process.

I agree with the judgment of Underhill LJ as to the answers to the remaining five issues, namely:

I would, as I have said, allow the appeal on the Soering issue. The Lord Chief Justice, for the reasons he gives, would dismiss the appeal. Since Underhill LJ agrees with me on the Soering issue, the appeal will be allowed. We will invite the parties to agree an appropriate order to reflect this judgment and that of Underhill LJ.

I have read the judgment of the Master of the Rolls. I gratefully adopt his summary of the background facts and issues, although I will occasionally repeat matters to save the need for cross-reference. I will use the term “Claimants” to refer to the individual Appellants (i.e. apart from Asylum Aid) collectively. Like the Master of the Rolls, I will take first the issues relating to the safety of Rwanda, on which the Claimants’ submissions were advanced by Mr Raza Husain KC and the Secretary of State’s by Sir James Eadie KC.

SAFETY OF RWANDA

It has been established since the decision of the European Court of Human Rights (“the ECtHR”) in Soering v United Kingdom 14038/88 [1989] ECHR 14 that it is a breach of article 3 of the European Convention on Human Rights (“the ECHR”) for a contracting state to remove an individual to a third country (whether or not itself a contracting state) where there are substantial grounds for believing that they will be at real risk of being treated contrary to the standards required by that article – this is the so-called “Soering test”. The position is helpfully summarised at para. 126 of the judgment of the Grand Chamber of the ECtHR in Ilias and Ahmed v Hungary 47287/15 (2020) 71 EHRR 6, as follows:

“Deportation, extradition or any other measure to remove an alien may give rise to an issue under Article 3, however, and hence engage the responsibility of the Contracting State under the Convention, where substantial grounds have been shown for believing that the person in question, if removed, would face a real risk of being subjected to treatment contrary to Article 3 in the receiving country. In such circumstances, Article 3 implies an obligation not to remove the individual to that country … .”

That principle was applied in the context of asylum-seekers in MSS v Belgium and Greece 30696/09 [2011] ECHR 108, in which the ECtHR held that Belgium was in breach of article 3 by returning asylum-seekers under the Dublin regime to Greece where (as it held separately) there were substantial grounds for believing that there was a real risk that their article 3 rights would be breached. In such a case the risk may take the form of serious ill-treatment in the country to which they are returned, but it may also take the form of the risk of refoulement (direct or indirect) to their country of origin where they have a well-founded fear of persecution: in MSS both risks were in fact found to be present. References in this context to whether the third country is “safe” cover both aspects.

As regards the latter risk – that is to say, the risk of direct or indirect refoulement – the Grand Chamber in Ilias observes, at para. 131, that “the main issue … is whether or not the individual will have access to an adequate asylum procedure in the receiving third country”. The reason is obvious: if the asylum-seeker does not have access to such a procedure there will prima facie be a real risk of their being refouled, either because their claim is not entertained at all or because it is not determined properly and fairly. As it says at para. 137:

“Where a Contracting State removes asylum seekers to a third country without examining the merits of their asylum applications, however, it is important not to lose sight of the fact that in such a situation it cannot be known whether the persons to be expelled risk treatment contrary to Article 3 in their country of origin or are simply economic migrants. It is only by means of a legal procedure resulting in a legal decision that a finding on this issue can be made and relied upon. In the absence of such a finding, removal to a third country must be preceded by thorough examination of the question whether the receiving third country’s asylum procedure affords sufficient guarantees to avoid an asylum-seeker being removed, directly or indirectly, to his country of origin without a proper evaluation of the risks he faces from the standpoint of Article 3 of the Convention.”

It follows, as it goes on to say at para. 139, that:

“On the basis of the well-established principles underlying its case-law under Article 3 of the Convention in relation to expulsion of asylum-seekers, the Court considers that the above-mentioned duty requires from the national authorities applying the ‘safe third country’ concept to conduct a thorough examination of the relevant conditions in the third country concerned and, in particular, the accessibility and reliability of its asylum system.”

Where the safety of the third country depends on assurances given by its government about the treatment which the individual will receive on return, it is necessary to consider the reliability of those assurances in accordance with the guidance given by the ECtHR in Othman v United Kingdom 8139/09 [2012] ECHR 56, as set out at para. 32 of the judgment of the Master of the Rolls. But the underlying question remains whether there are substantial grounds for believing that there is a real risk of a breach of article 3.

If the relocation of asylum-seekers to Rwanda under the MEDP involves a breach of their article 3 rights it would of course be unlawful as a result of section 6 of the Human Rights Act 1998.

The Claimants contend that Rwanda is “unsafe” in both the respects identified above – that is to say both because of the risk of serious ill-treatment in Rwanda itself and because of the risk of refoulement as a result of the inadequacies of its asylum system. However, in their submissions before us – as indeed, as I understand it, they did in the Divisional Court – they have concentrated mainly on the latter risk. I will accordingly focus on the question identified in Ilias, namely “whether or not the individual will have access to an adequate asylum procedure in the receiving third country”. I consider the question of other potential ill-treatment in Rwanda at paras. 287-291 below.

Before proceeding further I need to deal with three preliminary matters.

First, the Claimants contend that in the part of its judgment in which the Divisional Court considered the adequacy of the asylum system, at paras. 62-71, it proceeded on the basis that its task was not to decide that question for itself but only to decide whether the Secretary of State was entitled to conclude that the system was adequate. The Master of the Rolls considers that contention as his Issue 1 and concludes that the criticism is well-founded. The Lord Chief Justice has reached the opposite conclusion.

For me the issue is not of crucial importance, since, for reasons which will appear, I believe that even if the Court asked itself the right question its answer was wrong. However, if it were necessary I would, albeit with hesitation, accept the Claimants’ submission. Like the Lord Chief Justice, I do not accept for a moment that this experienced Court failed to appreciate the difference between the approaches required to determining a claim for judicial review and a claim of a breach of the ECHR, and I agree with him that its initial statement of the article 3 issue, at para. 39 of its judgment, was unimpeachable. But I think, with respect, that in its attempt to bring order to the confusing way in which the issues were presented it created a problem for itself by addressing the Claimants’ case under article 3 and their conventional judicial review claim under a single heading. That made it difficult to keep the different approaches distinct. At a number of points in the relevant passage, it uses what is unambiguously the language of judicial review when addressing the particular issues that are determinative of its overall conclusion on the adequacy of the Rwandan asylum system. In summary:

Second, Mr Husain sought at one stage in his submissions to argue that because the Secretary of State had certified the human rights claims made by (most of) the individual Claimants under paragraph 19 of Schedule 3 to the Asylum and Immigration (Treatment of Claimants etc) Act 2004 (with which I deal more fully in connection with Issue 14 below) the Divisional Court should have been concerned only with whether their claims based on article 3 were “clearly unfounded” and not sought itself to determine those claims: see in particular para. 43 of the AAA Claimants’ skeleton argument. With respect, that is a confusion. Although the human rights claims in question were indeed certified, the Claimants’ challenge to that certification is not before us because it succeeded in the Divisional Court (albeit, as the Court made clear at para. 179 of its judgment, on procedural grounds rather than on the substance of the article 3 claims); and the Secretary of State has not appealed. Accordingly, we are concerned only with the separate “generic” challenge advanced by the Claimants to the lawfulness of (in shorthand) “the Rwanda policy”, and the challenges to the lawfulness of the individual certification decisions are an irrelevance.

Having said that, I cannot help noting that the fact that the safety of Rwanda issue arises in the context both of the generic challenge to the Rwanda policy and of the challenges to the certification of the individual claims potentially creates a procedural conundrum about which challenge should be prioritised. If the only challenge had been to the certification of the human rights claims in the individual cases, the question for the Divisional Court would have been whether an appeal to the First-tier Tribunal (“the FTT”) against the refusal of those claims had a realistic chance of success. If the Court had held that it did, the judicial review proceedings would have achieved their aim and an appeal to the FTT would have proceeded. It is not clear to me that the position should necessarily be different where, as here, claimants plead a distinct challenge to the policy underlying the individual decisions. It is at least arguable that in such a case the better course is for the court to determine the challenge to the certification first, and, if it succeeds, to decline to consider the generic challenge on the basis that the FTT was the more appropriate forum, both because of its specialist expertise and because it would normally hear oral evidence. (Footnote: ¹) However, that question no longer arises in this case, if it ever did, and I mention it only in case it is relevant to future cases.

Third, the fact that we are on this appeal concerned with a challenge to the lawfulness of the policy means that we are not required to focus on the dates of the decisions in the individual cases (i.e. 5 July 2022). In substance, we are concerned with the lawfulness of the policy as at the date of the hearings before the Divisional Court in September and October: I note that the Master of the Rolls takes a different view (see para. 105 above) but I agree with him that the point makes no difference in practice. Another consequence is that we are not concerned with the risk of refoulement in the case of all or any of the individual claimants. Rather, we are concerned with the risk to the group as a whole to whom the asylum policy is intended to be applied.

Section 9 of the MoU reads:

“Asylum processing arrangement

9.1

Rwanda will ensure that:

9.1.1

at all times it will treat each Relocated Individual, and process their claim for asylum, in accordance with the Refugee Convention, Rwandan immigration laws and international and Rwandan standards, including under international and Rwandan human rights law, and including, but not limited to ensuring their protection from inhuman and degrading treatment and refoulement;

9.1.2

each Relocated Individual will have access to an interpreter and to procedural or legal assistance, at every stage of their asylum claim, including if they wish to appeal a decision made on their case; and

9.1.3

if a Relocated Individual’s claim for asylum is refused, that Relocated Individual will have access to independent and impartial due process of appeal in accordance with Rwandan laws.

9.1.4

If a Relocated Individual does not apply for asylum, Rwanda will assess the individual’s residence status on other grounds in accordance with Rwandan immigration laws.”

(The term “Relocated Individuals” – or RIs, as it appears in some Home Office documents – has a rather de-personalising tone, but I will sometimes use it for convenience.) More detailed guarantees are given in the Asylum Process Note Verbale (“the APNV”). I will refer to its particular provisions later as relevant.

I should summarise in outline the Rwandan process for determining individual asylum claims, described as the “refugee status determination”, or “RSD”, process. The basis of the process is a law passed in 2014, in some respects supplemented by subsequent Prime Ministerial Orders. Unfortunately, not all its details are clear, but I will defer consideration of the points of difficulty. The relevant stages are as follows:

The Claimants’ case that that system is seriously defective is supported by UNHCR as intervener and is largely based on the evidence adduced by it (Footnote: ²). The evidence in question consists of three witness statements from Lawrence Bottinick, the UNHCR Senior Legal Officer in the United Kingdom, dated 9 June, 26 June and 27 July 2022, to which I refer as “LB 1-3”. LB 2 is the most substantial: it runs to some 148 paragraphs, with numerous exhibits. Those exhibits include previous statements by UNHCR about its concerns about the asylum system in Rwanda, notably a seven-page Note dated 8 June 2022 headed “UNHCR Analysis of the Legality and Appropriateness of the Transfer of Asylum-Seekers under the UK-Rwanda Arrangement” (“the UNHCR Note”). Mr Bottinick’s evidence represents the institutional view of UNHCR, and I will sometimes refer to it as the UNHCR evidence.

The Master of the Rolls has already quoted the conclusion to LB 2, in which Mr Bottinick expresses UNHCR’s conclusion about the deficiencies in the Rwandan asylum system: see para. 7 above. I have read what he says at paras. 85-88 about the weight to be given to UNHCR’s opinion, and I respectfully agree with it. As he notes at para. 94, UNHCR has a large and active presence in Rwanda. It is true that the main focus of its activities is in the large camps housing refugees from neighbouring countries, and on the “ETM” project under which vulnerable migrants stranded in Libya are relocated to Rwanda on a temporary basis pending resettlement elsewhere; and Mr Bottinick complains about the limited extent to which the Rwandan government has allowed it to be engaged in the RSD process. Nevertheless it is clear that UNHCR’s staff in Kigali, and its partner organisations, have had extensive dealings with asylum-seekers who have been through, or sought to go through, that process. Mr Bottinick’s witness statements have been evidently carefully prepared by reference to that experience, with as much detail and identification of sources as practicable. There is no good reason not to accept what they say on matters of fact except where there is cogent evidence to the contrary.

The Secretary of State responded to LB 2 in witness statements from Kristian Armstrong, the Head of the Third Country Asylum Partnerships unit in the Home Office (“TCAP”), dated 5, 7 and 22 July 2022. The responses are, inevitably, dependent on information supplied by the GoR. That information is mainly contained in four exhibits to Mr Armstrong’s witness statements, as follows:

The Secretary of State also relies on the witness statement of Chris Williams dated 5 July 2022. Mr Williams is an Assistant Director Country Returns and Projects Immigration Enforcement, who was deployed to Rwanda in mid-May 2022 in order to support the operational aspects of the first relocations, then anticipated for 14 June. That involved ensuring that the various assurances in the MoU and the NVs could be fulfilled. His witness statement gives details of what he and his colleagues were told in a number of meetings and discussions with GoR officials, including a presentation on 25 May from, among others, someone he describes as “the DGIE Director”. Similar, but in some respects fuller, information deriving from the same exercise appears in a Home Office document of the same date called “MEDP Pre-Departure Assurance” (“the PDA”) exhibited by Mr Armstrong.

I should also mention the suite of “Country Policy and Information Notes” (“CPINs”) published by the Home Office in May 2022, all with the general title “Review of asylum processing”. These comprise a general CPIN headed “Rwanda: assessment” (“the Assessment CPIN”) and two CPINs dealing more fully with specific aspects – “Rwanda: country information on the asylum system” (“the Asylum System CPIN”) and “Rwanda: country information on general human rights” (“the Human Rights CPIN”) – to which the Assessment CPIN regularly cross-refers. The Asylum System CPIN contains an account of the system based on what the CPIT team were told by Rwandan officials on visits in January and March 2022. Full notes of the interviews which they conducted on those visits are published separately as “Annex A”. The CPINs are important for the purpose of Issue 13, with which I deal below, but on the issue of the safety of Rwanda it is necessary to proceed by reference to the evidence in the proceedings, as identified above.

The Claimants’ criticisms of the Rwandan system are summarised under fourteen headings in paras. 92-170 of the AAA Claimants’ skeleton argument below. (The skeleton argument before us refers back to those submissions but is less directly helpful because it is structured by reference to the criticisms which the Claimants make of the reasoning of the Divisional Court.) UNHCR makes substantially the same criticisms under fifteen heads in para. 18 of its Written Observations in the Divisional Court.

I propose to consider the Claimants’ and UNHCR’s criticisms of the Rwandan asylum system in two parts. First, I will consider the criticisms directed at particular stages, including the question of whether asylum-seekers can reliably gain access to the system at all. I will then address three issues which are common to the process as a whole, namely access to legal assistance/representation; availability of interpreter services; and training. I should make three points by way of preliminary.

First, UNHCR’s evidence is, necessarily, based on its experience of the system as it operated in the period up to the date of LB 2, i.e. 26 June 2022 (with the exception of a few matters added in LB 3). It is a central part of the Secretary of State’s case that that past experience is not a reliable guide to how those relocated to Rwanda under the MEDP will be treated, both because the GoR’s assurances involve specific improvements in the system as it may have operated previously and because it will operate in a wholly different context than before. Accordingly, in what follows I will have to consider not only how the system has operated up to now but the nature of any proposed changes (including, but not limited to, assurances given in the MoU and the APNV) and the likelihood that they will be implemented.

Second, it is important to appreciate that, despite having now been in place for some time, the RSD process has in practice been comparatively little used: UNHCR has described it as “nascent”. Rwanda has for many years had a very creditable record of granting asylum to large numbers of refugees from neighbouring countries, most of whom are accommodated in camps where UNHCR has an active role; but until August 2020 asylum was granted on a “prima facie basis”, i.e. without individual assessment of the claimants. The first individual assessments made by the RSDC were made in 2018 and the numbers remain small: see paras. 195-197 below. There have been comparatively few appeals to MINEMA and none to the High Court. Although the Divisional Court at paras. 55 and 70 of its judgment drew attention to a review published by UNHCR in July 2020 which was complimentary about Rwanda’s compliance with the Refugee Convention (albeit also expressing some criticisms), that statement was not directed to the RSD process.

Third, as will appear, the GoR disputes the factual basis of many of UNHCR’s criticisms. The Divisional Court says at para. 55 of its judgment that it did not believe that it was necessary to resolve most of those disputes, and in fact it considered only two of them: see paras. 202 and 203 below. I agree that in order to reach a conclusion on the adequacy of the Rwandan asylum system it is unnecessary, even if it were possible, to make definitive findings on each of the disputed matters of fact. The ultimate question is always whether there are substantial grounds to believe that there is a real risk of a breach of article 3 rights; and for that purpose what is required is an overview of the totality of the evidence (cf. the observations of Lord Hoffmann in an analogous context at paras. 45-49 of his speech in Secretary of State for the Home Department v Rehman [2001] UKHL 47, [2003] AC 153). But it remains necessary at least to review the evidence relating to the more important of UNHCR’s criticisms and so far as possible to form a view as to the reliability of their factual basis.

UNHCR asserts that there have in the recent past been a number of instances where individuals seeking asylum in Rwanda have been denied the opportunity to make asylum claims at all. The evidence is of three kinds.

First, UNHCR relies on four instances of “airport refoulement” (Footnote: ³) between February 2021 and April 2022, in which six people – two Libyans, two Afghans, one Syrian and one Yemeni – who had claimed asylum at Kigali airport on arrival were denied entry by DGIE staff and returned to the countries from which they had flown. The two Afghans and the Syrian are believed by UNHCR to have been returned from those countries to their countries of origin, i.e. refouled; it was able by rapid intervention to prevent a similar outcome for the two Libyans and the Yemeni. Evidence about these incidents is given in LB 2 and LB 3, but they are also referred to in the UNHCR Note, and were the subject of two Notes Verbales from UNHCR to the GoR – the first, dated 3 February 2021, relating to the Libyans and the second, dated 21 April 2022, relating to the Afghans and the Syrian. DGIE’s conduct is said not only to show an arbitrary disregard for the requirements of the Refugee Convention but also to be evidence of a prejudice against claimants from the Middle East and Afghanistan.

The GoR accepts that the individuals in question were denied entry and sent back as alleged, but it disputes UNHCR’s account of the circumstances. Its response appears not only in the primary and the tabular responses but also in a MINEMA “Feedback” document dated 11 May 2022 responding to UNHCR’s Note Verbale of 21 April. These are not entirely consistent as regards details (and, confusingly, the primary response refers to five Syrians rather than the single case raised by UNHCR), but the GoR’s essential case is that each of the individuals in question initially claimed entry on other grounds which proved false (e.g. possession of forged passports or inability to substantiate their claim to be travelling for business) and only at that point claimed asylum.

There are two points made by the GoR in this context which the Claimants say show serious misunderstandings of refugee law:

The Divisional Court adverts to these cases (except for that of the asylum-seeker from Yemen) at paras. 53 (1) and 55 of its judgment but it makes no findings about what occurred or what conclusions can be drawn from them. In my view the GoR’s responses do not satisfactorily answer the allegation that it acted in breach of the Refugee Convention by declining to consider the asylum claims made and that its expulsion of the claimants led to their (indirect) refoulement in some of the cases and risked it in the others; and they do indeed show an imperfect understanding of the requirements of the Convention.

It is UNHCR’s case that these episodes are very unlikely to be the only instances of airport refoulement. It has no presence at the airport, and it will only get to hear of such episodes if the claimants themselves, or other interested persons, are able to contact it, which would not always be the case.

Second, paras. 112-113 of LB 2 refer to three incidents, involving two families and an individual, where access to the asylum process was denied to persons who were already in the country. The claimants were all nationals of a non-African country with which Rwanda has particularly close relations (“country X”). They made asylum claims, but DGIE did not refer them to the RSDC. Instead, it gave the claimants a short period of notice to leave the country, and in fact thereafter simply transported them to the border (in one case Uganda and in the other Tanzania). It is Mr Bottinick’s evidence that they would have been indirectly refouled to country X if UNHCR had not been able to intervene and find other countries willing to accept them. He relies on these episodes not simply as instances of DGIE disregarding claimants’ Convention rights but of their doing so in the interests of their relationship with a foreign government. These incidents are not addressed in the GoR’s responses or therefore in the Secretary of State’s evidence.

Third, it relies on the experience of individuals who had sought asylum in Israel but had agreed to be relocated to Rwanda under the terms of an agreement between Israel and the GoR. The agreement was in place between 2013 and 2018. There is no dispute that those who accepted relocation under the agreement suffered serious breaches of their rights under the Refugee Convention. I can adopt the summary of Mr Bottinick’s evidence on this from para. 15 of the AAA Claimants’ skeleton argument:

“(i)

Many of those relocated were ‘not permitted to lodge their [asylum] claims’ and reported ‘arrests for lack of documentation’, ‘threats of deportation from unknown agents, following which eight disappeared’ and ‘continuous, random overnight visits by unknown agents at their accommodations’. All ‘feared for their personal safety, and feared refoulement to their country of nationality’. Further asylum-seekers later became known to UNHCR, of whom another seven ‘went missing’. Relocated asylum-seekers were ‘routinely moved clandestinely to Uganda even if they were willing to stay in Rwanda’. Dozens of asylum-seekers reported that ‘their documents were confiscated’ on arrival, ‘and they were taken to a house in Kigali where they were kept under guard’, before being ‘smuggled to Uganda’.

(ii)

UNHCR’s interviews with 80 Eritrean and Sudanese asylum-seekers in Italy who had been relocated under this arrangement revealed that, ‘feeling they had no other choice, they travelled many hundreds of kilometres through conflict zones’ and had ‘suffered abuse, torture and extortion before risking their lives once again by crossing the Mediterranean to Italy’. Some reported that those travelling with them had died en route to Libya. UNHCR is aware of two individuals who were transferred from Israel to Rwanda and who (as of 2022) still have no formal status in Rwanda despite claiming asylum several years ago. UNHCR has identified at least 50 cases of refoulement or threatened refoulement under the Israel-Rwanda arrangement.”

That evidence goes beyond the particular question of denial of access to the asylum process and also illustrates arbitrary and oppressive behaviour by Rwandan state agents, presumably DGIE.

Although the Secretary of State was aware of the Israel-Rwanda agreement and of the problems about it, she did not as part of the process leading to the MEDP seek to investigate why it had failed, whether by enquiring with the GoR or otherwise. Among other things, she did not attempt to obtain information about the terms of the agreement or what assurances the GoR had given about the treatment of asylum-seekers relocated under it. (Footnote: ⁴) As the Master of the Rolls explains at para. 101 of his judgment, her position in the Divisional Court, and before us, was that what had happened under a different agreement, made under different circumstances some years previously, could shed no light on whether the GoR could be relied on to comply with its assurances under the MEDP. She has accordingly not sought to adduce evidence contradicting or qualifying Mr Bottinick’s account of a wholesale failure by the GoR to respect the Convention rights of those returned under the agreement.

I would thus accept UNHCR’s evidence about all three episodes. But it is necessary to be clear what conclusions do and do not follow.

On the one hand, I do not accept that the evidence in question justifies the conclusion that there is a real risk that individuals relocated under the MEDP will be subject to airport refoulement or otherwise denied access to the asylum process. Their situation is clearly different from that of unheralded individual asylum-seekers, and also from that of the in-country nationals of country X. The GoR will have been supplied with their details in advance and will have expressly agreed to accept them under the terms of the MEDP. Their flight will be expected and arrangements put in place for their reception and accommodation. In those circumstances it is in my view inconceivable that DGIE would deny them entry to the country or would refuse them the opportunity to make an asylum claim as expressly promised in the MoU and the APNV. The case of those returned under the Israel-Rwanda agreement may seem rather closer to the present, but the available evidence does not establish that the circumstances in which they were unable to access the asylum system in Rwanda are comparable to those which will obtain under the MEDP.

On the other hand, I do not accept that these episodes are of no relevance. They 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. Those factors are capable of impacting on the treatment of asylum-seekers even if they are admitted to the process, particularly given the responsibility of DGIE for the first stage of that process. I will return in due course to the question whether the MEDP eliminates, or sufficiently mitigates, that risk.

I should mention one other issue. Para. 41 (i) of LB 2 records that UNHCR has since 2017 received reports that DGIE has refused to register claims made by claimants claiming fear of persecution on the basis of sexual orientation or gender identity. Mr Bottinick accepts that more recently two such claims have been permitted to progress, although he also reports a very recent episode where a LGBTQI+ asylum-seeker was submitted to very hostile questioning in his asylum interview. Similar concerns were raised by UNHCR at its meeting with Home Office officials in March 2022. The primary GoR response says:

“LGBTIQ+ not able to register. This is demonstrably untrue. The RSDC has received applications from LGBTIQ+ and has granted refugee status to those who have been determined as having a well-founded fear of persecution on the basis of their sexual orientation.”

The tabular response instances a particular case where an asylum-seeker of Egyptian origin was found to have a well-founded fear of persecution on the basis of his transgender status: see the answer to qu. 28. In LB 3 Mr Bottinick points out that that response is not inconsistent with his original evidence. However, even if there have been cases of the kind alleged in LB 2, for similar reasons to those given at para. 155 above I do not believe that there is a real risk that RIs claiming to fear persecution on the grounds of their sexual orientation will be denied access to the RSD process.

The first stage of the process is the responsibility of DGIE. As a matter of Rwandan law, its role is simply to do the necessary preparatory work, primarily by conducting the asylum interview. But UNHCR says that DGIE also performs a substantive screening or gatekeeper role which means that it may refuse to process claims, or delay them indefinitely, so that they never proceed to the RSDC (and a residence permit is not issued): see LB 2 paras. 38-39. Where this occurs there is no written decision, still less written reasons; but claimants have sometimes been given reasons orally which are inconsistent with the Convention. Those decisions are not appealable, and although the RSDC has power to consider a claim which DGIE has failed to refer timeously (under article 8 of a Prime Ministerial order) UNHCR has no experience of this ever occurring. The evidence relied on in support of the allegation that DGIE plays this gatekeeper role is identified at para. 39 of LB 2 as being the airport refoulement and in-country refusal cases discussed above, together with

“… several cases in 2021 and 2022 where individuals were told by the DGIE that their cases would not be referred to the RSDC solely because they had come to Rwanda on a work permit or tourist visas ... [and] ... were told by the DGIE to make an application only once their permit expires”.

The GoR denies that DGIE acts as a gatekeeper in the way described: see the primary response to para. 38 of LB 2. But the denial goes no further than re-stating the statutory position that DGIE is obliged to refer claims to the RSDC. There is a suggestion that UNHCR’s evidence is based on a misunderstanding of article 8 of the Prime Ministerial Order, but it is in my view clear that Mr Bottinick is reporting the actual experience of UNHCR staff in Rwanda.

I have already expressed my view about the airport refoulement and in-country denial episodes. As for the cases where claimants were told to re-present their claims only when their current residence permits expire, I do not believe that the GoR’s general denial justifies the rejection of Mr Bottinick’s evidence. I accept, however, that that conduct falls short of expulsion or a definitive refusal to entertain the claim.

Para. 4.7.3 of the Asylum System CPIN records that DGIE provides “a preliminary analysis of the application”. Mr Bottinick says that UNHCR staff have on several occasions been told by DGIE officials that following the asylum interview DGIE makes a recommendation to the RSDC as to the outcome of the claim (LB 2, para. 40), but the claimant is not given a copy of that recommendation. A “recommendation” is not quite the same as a “preliminary analysis” but both would involve DGIE giving the RSDC the benefit of its views on the substance of the application.

The GoR’s response to LB 2 (primary response at para. 40) reads:

“DGIE’s role in the RSD process consists of preparing files to be submitted to the RSDC for consideration. These files include information gathered about the asylum seeker. DGIE does not make any recommendation that may influence outcome of the RSDC decision. The RSDC takes decisions based on the information contained in the file and the additional information provided by the Eligibility Officer.”

Mr Bottinick confirms in LB 3 (para. 24 (d)) that that is inconsistent with what UNHCR staff have repeatedly been told.

I note that what the response denies is that DGIE makes “any recommendation that may influence outcome of the RSDC decision”. That is not inconsistent with “the information contained in the file” giving a statement of the views of the DGIE officer. Nor would it be surprising that the officer should express such views, given that they will have interviewed the claimant and should have relevant expertise to form a view; and given also that DGIE is an agency of the National Intelligence and Security Service which is one of the bodies which contributes a member to the RSDC. To the extent that any such analysis or recommendation is given, it is not satisfactory that the claimant is not shown it.

It will be appreciated that the asylum interview is of central importance because it may (subject to paras. 184-185 below) be the only opportunity which a claimant has to present their case orally. Paras. 4.3 and 4.4 of the APNV read:

“4.3

A decision whether a Relocated Individual is recognised as having a refugee protection need will only be taken following an appropriate examination that will, amongst other things, provide the Relocated Individual with the opportunity to:

4.3.1

make a written application and provide evidence in support;

4.3.2

attend an interview to explain their application and answer any questions the decision maker may have;

4.3.3

to further explain their claim, including any elements that may be missing from their written application or after the interviewer’s question.

4.4

The asylum interview will:

4.4.1

be transcribed or electronically recorded in full. If the interview is transcribed, the Relocated Individual will be given the opportunity to review and, if necessary, correct the transcript; and

4.4.2

be conducted under conditions which allow the Relocation Individual to present the grounds for their application in a comprehensive manner. In particular;

4.4.2.1 the person who conducts the interview will be competent to take account of the personal and general circumstances surrounding the application, including the applicant’s cultural origin, gender, sexual orientation, gender identity or vulnerability;

4.4.2.2 wherever possible, the interview with the Relocated Individual will be conducted by a person of the same sex if the Relocated Individual so requests, unless there is reason to believe that such a request is based on grounds which are not related to difficulties on the part of the Relocated Individual to present the grounds of his or her application in a comprehensive manner;

4.4.2.3 be in the presence of an interpreter who is able to ensure appropriate communication between the Relocated Individual and the person who conducts the interview. The communication shall take place in the language preferred by the applicant unless there is another language which he or she understands and in which he or she is able to communicate clearly.”

It is clear that some of those provisions were included in response to criticisms of the conduct of the asylum interview which had already emerged in the course of the two visits to Rwanda by Home Office officials and/or in their discussions with UNHCR.

I turn to consider UNHCR’s criticisms of the asylum interview. UNHCR is not generally permitted to attend asylum interviews but its staff in Kigali speak to asylum-seekers about their experience of them.

First, it is said at para. 41 (a) of LB 2 that the interview with DGIE is brief and perfunctory, lasting only about 20-30 minutes, and that it does not give the claimant a fair opportunity to explain the basis on which they are claiming asylum. That is particularly serious since DGIE is said (LB 2 para. 34) to encourage claimants not to exceed one or two pages in their initial applications or to submit lengthy documents, such as country information reports, in support so that there may be much that has to be amplified or explained at the interview. Mr Bottinick also says that claimants are not given an opportunity in the interview to address “adverse points” or to submit further information following it in order to address points which have arisen during it (LB 2, paras. 38 (c) and 41 (b)).

The GoR’s primary response to para. 41 (a) reads:

“This is not true. The interview takes as much time as necessary for the applicant [to] explain clearly his or her case. The interview guiding questions are set in way that provides the applicant the possibility to provide all information to support their application. Applicants can be invited for more interviews if necessary.”

No copy of “the interview guiding questions” is supplied, and the source of the information is not stated. That general and unsourced denial does not justify a wholesale rejection of UNHCR’s evidence.

The question then is whether the position will be different as a result of the MEDP. Para. 4.3.1 of the APNV expressly says that claimants will have the right to provide supporting evidence as part of the initial claim, and para. 4.3.3 allows them to further explain their claim following the interview. The APNV does not prescribe a minimum length for the interview, but the intention is evidently that it will be long enough to allow the claimant to explain their asylum claim and answer questions about it; and clearly in many cases thirty minutes would be inadequate for that purpose, particularly since usually both questions and answers will have to be translated.

I do not believe that it will be a straightforward matter for DGIE officials to change the way in which they have been accustomed to conduct interviews. No doubt it can be done, but it requires an appreciation that their previous approach was inadequate and effective training and monitoring. The response quoted above suggests that GoR does not accept that there is a problem which requires to be addressed. I return to the issue of training below.

Second, it is said that no transcript or other record of the interview is provided to the claimant: LB 2, para. 41 (e), LB 3 para. 29 (a). This is a point of obvious importance since the facts elicited at interview will form a crucial part of the file which goes to the RSDC.

It is not entirely clear whether the GoR accepts that this is not part of its current practice. The supplementary GoR e-mail says:

“Records of the DGIE interview: The DGIE conducts interviews with the asylum seeker in the initial stages of the asylum process with a view to submit the information to the RSDC and to grant the asylum seeker a temporary residence permit. The DGIE interviews consist of the asylum seeker describing their reasons for seeking asylum in Rwanda. The interview is recorded electronically and at the end of the interview, the asylum seeker is presented with a written record of the interview. The asylum seeker verifies the information and can confirm the record with a signature or can amend the record by correcting the information or providing more information. A copy of the DGIE record of interview as verified by the asylum seeker will be made available to the asylum seeker. The legal representative retained by the asylum seeker can assist with reviewing the records of the interviews.”

That passage starts by using the present tense but changes to the future tense.

I do not believe that that evidence is an adequate basis for rejecting UNHCR’s evidence as to past practice, and it is more likely that it reflects the intended post-MEDP system: as we have seen, provision of a transcript, with the opportunity to approve it, is specifically addressed by para. 4.4.1 of the APNV. As to that, I should add that the PDA comments on this obligation as follows:

“Ability for Relocated Individual to review transcription

RIs will have the ability to review the transcript and correct the record immediately after the interview process. The transcript can be printed there and then. After the interview is conducted and information recorded by DGIE on their digital system, it will be reviewed again to confirm the transcript is correct, including where relevant, that translations are correct as transcribed by the interpreter. Once all parties are content the document will be signed.”

It is clear from para. 34 of Mr Williams’ witness statement that this information derives from the DGIE Director.

The real question thus is whether the requirements of the APNV will be complied with. We are here concerned with a specific procedural requirement rather than one involving the exercise of judgment, such as the conduct of the interview as considered above. I see no reason to suppose that DGIE will deliberately disregard the explicit requirement of the APNV, confirmed and amplified in the statements of the DGIE Director, that it should supply claimants with a copy of the transcript of the asylum interview in their own language. However, the process described in the APNV is not straightforward, and training will be required to ensure that it can be accomplished in practice. Again, I return to this below.

Third, Mr Bottinick says (LB 2, para. 41 (c); LB 3 para. 28 (a)) that claimants are not permitted at the asylum interview to be accompanied by a lawyer. The evidence about this, or in any event about whether it will continue to be the case under the MEDP, is not clear. Because the same question arises in relation to stages (3) and (4) I address it separately below. As will appear, my conclusion is that legal representatives are certainly not permitted at present but it appears to be intended that henceforward they will be. (It is accordingly unnecessary to consider whether that is in all cases necessary in the interests of fairness: I will only say that it is very likely that in some circumstances it will be.) However, it is not clear what consideration has been given to the role that legal representatives can play at the interview.

Mr Bottinick addresses the role of the Eligibility Officer at paras. 42-47 of LB 2. He points out that it is potentially very important because the additional information which the Eligibility Officer may obtain, and the notes of any interview that they conduct with the claimant, will presumably go before the RSDC; but he complains that there is no transparency about this stage of the process. It is not clear on what basis claimants may be selected for interview nor is any record of the interview shared. Nor is it clear what the file compiled for the RSDC consists of, including what country information, UNHCR guidance, or summary of legal principles it may contain. He also says that there is at present only one Eligibility Officer, which is unsatisfactory given the nature and importance of the role (and also because the current incumbent does not speak English).

The GoR responses do not comprehensively address these criticisms. The GoR statement appears to say that the Eligibility Officer interviews every claimant (para. 4); but the statement is in very general terms. As for numbers, the answer to qu. 24 in the tabular response states that there are now two Eligibility Officers and that MINEMA is recruiting three more and plans to recruit others in future. The answer to qu. 36 is similar but not identical. It says:

“There is currently one Eligibility Officer which is adequate for the usual volume of claims received. (ToR of an Eligibility Officer are attached below). MINEMA is actively recruiting more Eligibility Officer to cater for the expected increase in asylum claims under this partnership.”

The “attached ToR” would seem to be the job description for the role as advertised (apparently now to be described as “Eligibility and Protection Specialist”) exhibited by Mr Armstrong to his witness statement dated 2 July 2022. This reads:

“JOB PURPOSE

The Eligibility and Protection Specialist will be in charge of checking, record keeping, filing, and advocating on behalf of migrant families. He/She will determine whether or not migrant families’ various needs are met and propose corrective measures.

DUTIES AND RESPONSIBILITIES

Under the direct supervision of the Program Manager, the Eligibility Specialist will perform the following duties:

· Assist in the monitoring and analysis of statistics related to migrant case processing in order to identify and respond to developments or issues affecting decision‐making quality, and to propose corrective measures.

· Ensure the reception of migrants and asylum seekers and assist them in providing feedback on individual cases.

· Conduct Migrants, asylum seekers’ interviews and draft their Assessments in accordance with set guidelines.

· Conduct research on country of origin information and legal issues concerning migrants and asylum seekers, and assist in the maintenance of a local database of relevant information.

· Maintain accurate and up‐to date records and data related to all work on individual cases.”

The reference in the third bullet to conducting interviews and drafting assessments cannot, I think, be to any involvement at the DGIE stage but to the Eligibility Officer’s own interview (though it is not clear whether this will occur in every case) and assessment carried out for the benefit of the RSDC.

Mr Armstrong also says, at para. 46, that in the course of the negotiation of the MEDP

“… Rwandan officials explained that individuals will be supported throughout the process by a caseworker from the Eligibility and Protection Office in Ministry of Emergency Management (MINEMA) who will collate information related to an individual’s case and submit it to the National Refugee Status Determination Committee for decision.”

I see no reason to doubt that the intention of the GoR is that there will be a sufficient number of Eligibility Officers to deal with the increase in asylum-seekers going through the RSD process as a result of the MEDP; or that they will have for the future the functions and responsibilities described in the job description, whether or not they precisely correspond to what their role has been in the past. I need not therefore reach a conclusion about UNHCR’s criticisms of the current position. However, it is clear that the role as envisaged is important and valuable: given the part-time and non-specialist nature of the RSDC, it is in truth essential that it be serviced by full-time officials who can provide it with necessary information, materials and technical advice. That being so, it is essential that the Eligibility Officers be properly trained: I return to this below.

Finally, I should note that it is not clear from the APNV, or any of the other evidence, whether any interview conducted by the Eligibility Officer will be subject to the requirements specified in para. 4.4 – i.e. that it will be conducted through a qualified interpreter and the claimant will have the opportunity to address the record.

As regards the substantive determination of an RI’s asylum claim – that is, the decision by the RSDC – the APNV provides as follows:

“4.5

For the purpose of taking decisions on asylum claims, decision makers will obtain up-to-date information as to the general situation prevailing in the countries of origin of the Relocated Individual. This information will be available to decision-makers, and they will have appropriate resources to further research and access expertise where needed.

4.6

A decision on a Relocated Individual’s asylum application will:

4.6.1

be taken on the merits of the individual application; and

4.6.2

will be objective and impartial.

4.7

Arrangements will be made to ensure that the decisions taken on individual claims are recorded.

4.8

Relocated Individuals will be notified in writing of the decision that has been taken on their asylum claim.

4.9

A decision will:

4.9.1

be in one of the official languages of Rwanda and, if needed for understanding, it will be translated in writing by an interpreter into a language that the Relocated Individual understands, free of charge;

4.9.2

include the reasons for the decision in both fact and law; and

4.9.3

a decision that is a refusal of the asylum claim will notify a Relocated Individual that they will be able to appeal the decision on their asylum claim and provide an explanation of how to do this.”

Paras. 48-65 of LB 2 advance a number of criticisms of the RSDC stage of the process. I consider them under the following heads.

Mr Bottinick’s evidence is that in the majority of cases the RSDC takes its decision at a meeting which the claimant does not attend (and which indeed they will not know has taken place until they receive the decision) and thus without receiving any submissions from him or her: see paras. 56 and 59 of LB 2. This means that in such cases the quality of the information before the Committee, at least as regards the claimant’s individual circumstances, is dependent on the quality of the asylum interview conducted by DGIE, and any further interview conducted by the Eligibility Officer, and the records of such interviews. The claimant has no access to those records and so is in no position to correct any errors or clarify any misunderstandings. More generally, as he says in para. 65 (d), they have

“no opportunity … to present their claim in full, or address provisional adverse findings (eg to address a credibility concern that has arisen in the view of the decision-maker(s), including through a lawyer).”

If that is a fair summary of the procedure followed it clearly has the potential for serious unfairness.

The GoR responses do not directly address this part of Mr Bottinick’s evidence (although para. 4.4.1 of the APNV provides for them to receive a copy of the transcript of the asylum interview). However, at para. 36 of his witness statement Mr Williams says that the DGIE director told him that

“… following their asylum interview and receipt of their translated interview transcript, relocated individuals will be given the opportunity to make oral and written representations directly to the RSDC and that the RSDC may request further information from relocated individuals if required for the purposes of reaching a decision on their claim”.

This is the only reference in the evidence to an RI being given the opportunity to make “oral and written representations” directly to the RSDC. Nothing is said about when any written representations would be expected to be lodged or considered. The reference to oral representations is presumably to representations made by the claimant at the meeting at which the Committee makes its decision: a separate “hearing” would be a radical departure from its practices which I would expect to have seen fully explained. In that case there might not be much difference between the right to make oral representations and the “interview” referred to by Mr Bottinick: the important point would be that there would be an interview in every case.

If that is indeed what the DGIE Director intended to convey, it appears inconsistent with the understanding of the Chief Technical Adviser to the Ministry of Justice, since the passage from the supplementary e-mail quoted at para. 230 below carefully refers to “any interview at the RSDC”. It is also inconsistent with what the Home Office officials understood from their meetings in January and March 2022: para. 4.7.3 of the Asylum System CPIN says that “RSDC can [my underlining] request to meet the applicant to verify information (in a 20-40 min interview)”. It is very unsatisfactory not to have a clear and consistent account of how the RSDC will proceed in determining the RIs’ cases.

Mr Bottinick complains, on the basis of accounts from individual asylum-seekers and others who have been present, that where the RSDC interviews the claimant the interview is short (cf the CPIN reference to 20-40 minutes) and often conducted unprofessionally (e.g. by preventing claimants giving proper answers or submitting further information or by hostile questioning) or in a way that appears to show a poor knowledge of the file and/or a failure to appreciate the requirements of refugee law (e.g. by a focus not on the claimed fear of persecution but on why the claimant has not sought asylum “closer to home”): see para. 60 of LB 2. In addition, he says that professional interpreters are rarely employed. These complaints are not addressed in the GoR’s response.

Mr Bottinick also complains that where the RSDC has conducted an interview no transcript is made available to the claimant. That criticism is not addressed directly by the GoR, but its supplementary e-mail says:

“The refugee status determination by the RSDC is an administrative process wherein minutes of the decision-making process are recorded. These minutes will be made available to the relocated individual attached to their notification of decision by the RSDC.”

Whether that is a complete answer depends on what procedure is followed in the case in question. If the RSDC proceeds to a decision immediately after the interview the only value to the claimant of having a transcript would be in case what they had said was relevant to an appeal; and in those circumstances the promised minute might be adequate, depending how full it was. But if the decision is deferred to a later occasion it might be of real importance for the claimant to have a record, both so that any errors could be corrected and as the basis for any submissions that they might wish to make (subject to the point I have to consider next). The APNV says nothing about the agreement of a transcript in this situation. This uncertainty is unsatisfactory.

Finally, and most importantly, Mr Bottinick complains that claimants are not entitled to make submissions to the RSDC through a lawyer. Para. 60 (j) of LB 2 reads:

“There is no opportunity for asylum seekers (whether or not interviewed) to make submissions (in person or through a lawyer) to the RSDC. Lawyers are not permitted at the RSDC stage. UNHCR and its legal aid partners have been told repeatedly that if a person was telling the truth, they had no need for a lawyer. Over the years, when legal aid partners have inquired about the possibility of legal representation, they have been told that as the relevant national refugee law does not specifically refer to provision of legal representation, it cannot be permitted.”

That evidence, so far as concerns representations by a lawyer, is not challenged by the GoR. As appears at para. 230 below, it does now say that a claimant may be accompanied by their lawyer if the RSDC conducts an interview, but it has not changed its position that it will not entertain submissions from them: Sir James Eadie confirmed this in the course of his oral submissions. In my view this is a serious defect in the process. Any representations will in most cases only be effective if made by a lawyer because 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. It is true that in many, perhaps most, asylum claims the determinative issue will be whether the claimant is telling the truth, and that what they say in their interview(s) will be of central importance. But even on that issue there may be points to be made beyond the bare narrative; and sometimes there will be important issues about matters outside the claimant’s knowledge, such as developments in their country of origin or issues of law. It is also important to bear in mind that many RIs are likely to be especially vulnerable as a result of their experiences, which may include a history of torture. The need to have a lawyer to present their claim will be particularly important in such cases.

Para. 4.5.1 of the APNV acknowledges the need for decision-makers – in practice, the RSDC – to have access to up-to-date country information. Mr Bottinick observes that because claimants do not see the contents of their files it is impossible to know what information is in fact provided to the Committee. In its tabular response (at qu. 10) the GoR says that the RSDC is provided with “country information reports from ministry of foreign affairs and other open sources country information reports”. That is not very specific, though I note that the duties of the Eligibility Officer will apparently include maintaining a relevant database (see para. 177 above). On any view it will be particularly important for RSDC members to have access to good objective information about the kinds of country from which most RIs are likely to come, of which they will have had little previous experience.

Para. 4.9.2 of the APNV requires that a written decision will “include the reasons for the decision in both fact and law”. That is particularly important where a claim is refused. The intention is plainly that decisions should not be “stereotyped” but should address the particular factual case advanced by the individual claimant and identify the legal basis for the decision.

Para. 61 of LB 2 deals with the reasons given by the RSDC for decisions refusing an asylum claim. Mr Bottinick says that UNHCR has seen the reasons given in 116 cases and that in none of them were the reasons in sufficient detail to allow the claimant to understand why their claim had been rejected. In 36 out of 50 such refusals in the previous year (of which he reproduces the relevant parts in a table) the reasons given amounted to no more than some such formula as “because you don’t meet the eligibility criteria and the reasons you provided during the interview are not pertinent”. On the occasions where something more is said it remains cursory in the extreme. UNHCR makes the further point that decisions in this form are not only inadequate in themselves but indicate a poor quality of decision-making.

In its primary response to para. 69 of LB 2 the GoR states that:

“The reasons are briefly provided in the notification and more detailed reasons are communicated to the applicant in person. Templates are being adjusted to provide detailed reasons on the notification.” (Footnote: ⁵)

That tacitly acknowledges that reasons in the form previously given are inadequate; and in any event Sir James accepted that before us. No examples have been produced of more detailed reasons given in recent cases before the RSDC.

I see no reason to reject the GoR’s statement that it intends that the RSDC will henceforth give more detailed reasons for its decisions. But the giving of proper written reasons, particularly where this has not been the practice to date, is not a straightforward matter, and it is not possible to be confident that it will occur unless those drafting them receive proper training.

Para. 63 of LB 2 contains a table headed “Overview of cases processed by RSDC as known by UNHCR for 2020 to 2022 (as of 21 June 2022)”. Mr Bottinick accepts that the table may be incomplete because the GoR does not share statistics with UNHCR, but he explains the sources from which it is derived. The table shows a total of 156 cases in the relevant period, with an overall rejection rate of 77%.

Mr Bottinick draws attention to the fact that the table shows three asylum claimants from Syria, three from Yemen and two from Afghanistan, and that all of their claims were rejected. He also refers to eighteen claims from Eritrea, of which ten were rejected. He says that the situation in the countries in question, and the profiles of the individual claimants, are such that all their claims are very likely to have been well-founded and that the numbers who were in fact rejected casts doubt on the quality of the RSDC’s decision-making. More particularly, as regards the former group he suggests at para. 114 of LB 2 that the reason for the rejection of their claims is likely to be a bias against asylum seekers from the Middle East and Afghanistan. As to that, he relies not only on the figures in the table but on the cases of airport refoulement discussed at para. 146 above. He also relies on statements which he says that UNHCR’s staff have heard senior government officials make to the effect that asylum-seekers from the Middle East and Afghanistan should claim asylum in their own region. The latter allegation was not made for the first time in his evidence: it has been made in UNHCR’s meetings with Home Office officials in both March and April 2022.

The GoR has provided a different table, covering substantially the same period. This shows a total of 108 claimants, including only one from each of Syria, Yemen and Afghanistan, all of whose claims were rejected, and twenty from Eritrea, fourteen of whose claims were rejected. It has also provided short notes on the reasons for the rejections in question, rebutting any suggestion of prejudice, but without more detail it is not possible objectively to assess their validity.

In LB 3 Mr Bottinick addresses the differences in the numbers as regards the claimants from Syria, Yemen and Afghanistan and confirms that he is confident in the accuracy of his figures. He suggests that the discrepancy may arise, at least in part, from the GoR treating linked cases as a single entry.

It is not possible definitively to resolve the difference between the two tables, but I believe we should proceed on the basis of UNHCR’s figures, both because they are carefully explained and defended in both LB 2 and LB 3 and because we should take a “substantive grounds/real risk” approach. In my view the surprisingly high rejection rate of claimants from known conflict zones, where UNHCR recommends against returns, does indeed suggest a poor quality of decision-making.

As regards the more particular allegation of a bias against claimants from the Middle East and Afghanistan, UNHCR’s figures are of course statistically frail. But I do not believe they can be disregarded, particularly when taken with its evidence about the views expressed by senior GoR officials that they should have sought asylum nearer to home. Such views are not uncommon generally, and they are consistent with questions reported to have been asked by RSDC members in cases where the claimants have been interviewed (see para. 186 above); they would also be a plausible explanation for the airport refoulement cases. They constitute evidence that RSDC members may hold such views and that they may influence its decision-making.

I accept that the changes associated with the introduction of the MEDP may improve the quality of decision-making and help to eliminate prejudices of the kind which UNHCR alleges are operative. But the extent to which it will do so will depend on the effectiveness of the training which DGIE and RSDC officials receive.

Para. 41 (h) of LB 2 suggests various grounds for believing that DGIE may seek information about asylum-seekers from the authorities in their countries of origin or their embassies in Rwanda. The GoR’s primary response is to the effect that it is standard practice for the Ministry of Foreign Affairs to “liaise with Embassies to gather background information on the applicant and/or to gather country information”. That is ambiguous, but in the GoR supplementary e-mail it is explained that the reference is to the Rwandan embassies (or High Commissions) in the country of origin, to which application may be made if information is required about a “specific event/situation”. The e-mail states in terms that “the DGIE and RSDC do not share the personal data of an asylum seeker with any third party during the processing of the asylum seeker’s application”. This is one of the two criticisms on which the Divisional Court made a finding: at para. 56 of its judgment it said that it was satisfied with the GoR’s explanation. I find it difficult to reach a concluded view, and for the reason given at para. 144 above I do not believe it is necessary to do so.

The Claimants allege that the drafting of article 7 (1) of Law no. 13, which governs eligibility for asylum in Rwanda, fails properly to implement the Refugee Convention because it does not cover persecution for imputed political opinion or by non-state actors. That criticism is adopted by Mr Bottinick in LB 2 (para. 82). As the Divisional Court points out at para. 55 of its judgment, it is impossible to form a view about whether that is in fact how article 7 would be interpreted by a Rwandan court without expert evidence of Rwandan law. The AAA Claimants’ skeleton argument said that the absence of such evidence was itself a breach of the Ilias duty, but Mr Husain did not develop the point in his oral submissions.

At paras. 83-88 of LB 2 Mr Bottinick expresses a number of concerns about the ability of inexperienced Rwandan decision-makers to understand some of the more subtle important concepts of substantive asylum law, such as the principle established in HJ (Iran) v Secretary of State for the Home Department [2010] UKSC 31, [2011] 1 AC 596. This, however, is more relevant to the issue of training, which I consider separately below.

Under head (9) of its Written Observations UNHCR says:

“UNHCR has observed a lack of training or sufficient knowledge at all stages of the Rwandan RSD system. The changing, part-time and non-specialist composition of what is in principle the main decision-making body on asylum claims, the RSDC, in UNHCR’s view compromises the quality and integrity of the Rwandan RSD procedure. The RSDC’s members are high level functionaries from an array of ministries, whose primary responsibilities lie elsewhere and many of whose portfolios do not otherwise include matters relevant to the asylum procedure. UNHCR’s repeated offers to provide training to the Rwandan RSD authorities have only been taken up on two occasions, with a gap of three years in between and those trainings were short and basic.”

The passage goes on to develop the criticism in relation to training.

I would not accept – if UNHCR intends to go this far – that a body with the composition of the RSDC is inherently incapable of making proper decisions on asylum claims, provided always that its members approach their task conscientiously, and that they are provided with full information, including representations from or on behalf of the claimant, and proper specialist support. Para. 4.5 of the APNV recognises this, and it seems that it is intended that the support and information there referred to will be provided by the Eligibility Officers. However, the main focus of UNHCR’s criticism appears to relate to the absence of training. I agree that this is fundamental, and I return to it below.

Mr Bottinick addresses the right of appeal to MINEMA at paras. 66-75 of LB 2. In short his criticisms are: that claimants whose claims are rejected by the RSDC are not notified in writing of their right of appeal and only sometimes notified orally; that the right of appeal is ineffective because in the absence of a reasoned decision from RSDC it is impossible to address the basis on which the claim was rejected; that it is unclear whether the appeal is (adopting English terminology) by way of review or re-hearing; that MINEMA is not independent, because its Permanent Secretary is the Secretary to RSDC; that free legal advice is not available; and that MINEMA does not give reasons for its decisions. He also observes that UNHCR is unaware of any cases where an appeal to MINEMA has succeeded.

Those criticisms are only partly addressed in the GoR’s primary response. It appears to say that claimants are informed by the Eligibility Officer of the reasons for MINEMA’s decision and of their right of appeal; but in fact I think the reference may be to the decision of the RSDC. It also says that two out of the five appeals to MINEMA in 2021 succeeded.

Paras. 5.1-5.2 of the APNV read:

“5.1

A Relocated Individual may appeal a refusal of their asylum application to the Minister responsible for considering such appeals.

5.2

A Relocated Individual who wishes to appeal to the Minister will have the opportunity to make oral and or written representations. Any legal representative engaged by the Relocated Individual will have the opportunity to make submissions when appropriate before the end of the process of appeal to the minister.”

That makes clear that, whatever may have been the position in the past, MINEMA will consider representations from the claimant’s lawyer.

It will be seen that the evidence about the right of appeal to MINEMA is very limited. That makes it difficult to assess its effectiveness as a safeguard against wrong decisions by the RSDC. I find it hard to believe that following the MEDP claimants will be left unaware of their right of appeal or that fuller reasons for a refusal by MINEMA will not be given (as is intended with the RSDC). But it is impossible to form a useful view about the quality or independence of its decisions. The issue may not be central since the ultimate safeguard should be the appeal to the High Court, to which I now turn.

The various problems identified above about the administrative stage of the RSD process make it particularly important that there be a right of appeal to an independent judicial tribunal which can examine the claim afresh. Para. 9.1.3 of the MoU provides:

“If a Relocated Individual’s claim for asylum is refused, that Relocated Individual will have access to independent and impartial due process of appeal in accordance with Rwandan laws.”

That commitment is amplified by paras. 5.3-5.5 of the APNV, which read:

“5.3

A Relocated Individual whose appeal has been refused by the Minister will be permitted to appeal that decision to the High Court of Rwanda.

5.4

The court will be able to conduct a full re-examination of the Relocated Individual’s claim in fact and law in accordance with Rwanda rules of court procedure.

5.5

A Relocated Individual and their representative will have the opportunity to make full representations as to fact and law at their appeal in accordance with Rwandan rules of court procedure.”

That does not give details of the “independent and impartial due process of appeal” afforded by Rwandan law, but article 47 of “Law 30/2018 of 02/06/2018 determining the jurisdiction of courts” provides that “the High Court also adjudicates cases relating to the applications for asylum”. It is common ground that that provision gives an asylum-seeker whose claim has been rejected by the RSDC and MINEMA what is in practice a right of appeal to the High Court. It is also common ground that no appeal under article 47 has been brought since it became available five years ago, and there is no evidence about how it would work in practice.

At para. 143 of LB 2 Mr Bottinick says:

“Even if the safeguards of representation and High Court appeal are now put in place for UK-Rwanda arrangements, judges and lawyers do not have relevant experience. This raises a serious question about the effectiveness of any appeal. In any event, UNHCR does not consider that the possibility of an appeal to the High Court provides a sufficient safeguard against a decision-making process which is flawed from the outset.”

He also repeats in this context his points that claimants are not reliably notified of the right of appeal, in this case to the High Court, and that an appeal will be difficult in the absence of proper reasons from the RSDC and/or MINEMA.

In the tabular response the GoR gives further information about the right of appeal to the High Court in answer to two questions.

First, qu. 11 asked some questions about the procedural aspects of the appeal. The response reads:

“Individuals are allowed to appeal to high court to request a judicial review of the decision given by the Minister.

- The High Court tries cases by a bench of one (1) or three (3) judges assisted by a Court Registrar. The President of the court determines the appropriate number of the sitting judges depending on the importance of the case.

- Evidence admissible under the rules of evidence can take the form of testimony, documents, photographs, videos, voice recordings, DNA testing, or other tangible objects.

- High Court judgments can be appealed to the Appeal Court.”

That information is useful as far as it goes, but it does not appear to be specific to asylum appeals. If the term “judicial review” is being used in its English sense it would appear inconsistent with the reference in para. 5.4 of the APNV to “a full re-examination of the Relocated Individual’s claim in fact and law”; but the GoR is not consistent in its usage (see below).

Second, qu. 31 asked whether there had been provision for appeals to the High Court prior to the 2018 law and whether it was the case that there had indeed been no appeals so far, concluding “Can you provide any information that will reassure our courts that anyone refused asylum will have access to this system?”. The response says that other routes of challenge been available under different legislation prior to 2014 and had been used on four recorded occasions. It says that there is no statutory impediment to asylum seekers appealing under the 2018 law and that “there are currently 44 asylum seekers who were refused refugee status after their appeal to the Minister who are within their right [to] get an appeal/judicial review of this decision at the High Court”. This is not, of course, a statement that any of the 44 have in fact done so.

The Claimants advanced a distinct challenge to the effectiveness of the right of appeal based on the lack of independence of the judiciary. A submission that the Rwandan judiciary were not independent of the government was first made at paras. 219-223 of the AAA Claimants’ skeleton argument before the Divisional Court. It was not specifically related to the effectiveness of the right of appeal against the refusal of an asylum claim but was part of a general submission about human rights in Rwanda. Perhaps for that reason, it was not addressed in the Divisional Court’s judgment. However, in the AAA Claimants’ skeleton argument before us the submission was deployed in this context (see paras. 9 and 18), and it was fully developed in Mr Husain’s oral submissions.

The Claimants relied primarily on the decision of the Divisional Court (Irwin LJ and Foskett J) in Government of Rwanda v Nteziryayo [2017] EWHC 1912 (Admin), in which the Court upheld the refusal of Senior District Judge Arbuthnot to order the extradition of five Rwandan nationals on charges relating to the genocide on the basis that there was a real risk they might suffer a flagrant breach of their rights to a fair trial if extradited. At para. 234 of its judgment the Court said:

“We reject firmly the submission of the GoR that the judge should unequivocally have stated the judiciary was independent and that there was no risk of bias or interference in these cases, because they are not ‘political’. The clearly authoritarian nature of the regime; the long and continuing history of the influence of political will over the justice system where the case is perceived to matter; the evidence of Gahima [a Rwandan lawyer and activist] and others; the evidence of threats arising from criticisms of the regime, not in a political context but in the context of the justice system; the ‘Osman’ warnings [warnings given to Rwandan exiles in London that they were at risk from the GoR]; the experiences of witnesses who gave evidence in genocide cases unfavourable to the prosecution, whether in Rwanda or abroad: all these point to a high level of risk of pressure within the system.”

At para. 373 it said:

“The evidence suggests that judges are not appointed unless they have party membership of the RPF [i.e. the governing party]. Their appointments have moved from being indefinite to appointment for definite terms. The Rwandan executive can achieve the dismissal of serving judges: it has done so in recent times in respect of around 40 individual judges. We are not in a position to say whether the suggested misconduct or corruption was established in these cases: it may be so. But the capacity of the executive to get rid of judges is established. In such circumstances, there can be little doubt that judges will feel exposed.”

The Claimants also place reliance on three other matters:

In his submissions before us Sir James Eadie did not dispute that the conclusions in Nteziryayo and the evidence relied on by the Claimants gave, at the least, real grounds to believe that the Rwandan judiciary was susceptible to political pressure. But he pointed out that the context in each of those instances was the trial of political opponents. The context in the case of asylum appeals would be completely different. The GoR would have no political interest in the outcome of particular appeals and would have no reason to seek to manipulate them. In so far as judges might nevertheless wish to determine appeals in accordance with what they understood to be government policy, the GoR’s declared policy was to ensure compliance with the MoU, which promised an independent and impartial appeal.

There is force in those points. However, I do not believe that they afford a complete answer. In the first place, on an appeal under article 47 of the 2018 Law the High Court is being asked to overturn the decision of a Government Minister and, indirectly, the decision of a Committee comprised of senior representatives of the principal government bodies and agencies, including the Prime Minister’s Office and the National Intelligence and Security Service. Given what the Divisional Court in Nteziryayo calls “the influence of political will over the justice system where the case is perceived to matter”, and the insecurity of the position of Rwandan judges, 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. Further, I do not think it can be assumed that the GoR will never have an interest in the outcome of particular asylum appeals: the willingness to deny access to the asylum process for nationals of “country X” (see para. 151 above) suggests that in particular circumstances it may well have an interest in denying asylum to RIs of particular nationalities.

These concerns could in principle be met if there were evidence of how the appeal system has worked in practice. But since not a single appeal has so far been brought there is no such evidence.

There are distinct issues about legal representation and training, but I address these separately below.

As we have seen, para. 9.1.2 of the MoU guarantees RIs “access to procedural or legal assistance, at every stage of their asylum claim”. What “legal assistance” amounts to is amplified in paras. 7 and 8 of the APNV as follows:

“7.

Procedural and Legal Assistance

7.1

A Relocated Individual will be provided with orientation that includes details of the asylum process and support that is available to them free of charge.

7.2

Each Transferee will be permitted to seek legal advice or other counsel from any non-governmental or multilateral organisation, at any stage of the asylum application process at their own expense including from an organisation providing that support free of charge.

7.3

The legal representative or other counsel engaged by a Relocated Individual in accordance with 7.2 above will be permitted to provide legal assistance at every stage of the claim, in accordance with Rwandan law.

8.

Legal assistance at appeal

8.1

Should a Relocated Individual wish to appeal their decision to the court of Rwanda they will be provided with legal assistance and representation from a legal professional qualified to advise and represent in matters of asylum, free of charge. This shall include, at least, the preparation of the required procedural documents and participation in the hearing before the appeal court on behalf of the applicant.

8.2

Rwanda shall provide the legal advisor access to the information provided by the applicant’s file upon the basis of which a decision is or will be made. Rwanda may make an exception where disclosure of information or sources would jeopardise national security, the security of the organisations or person(s) providing the information or the security of the person(s) to whom the information relates or where the investigative interests relating to the examination of applications for international protection by the competent authorities of Rwanda or the international relations of Rwanda would be compromised.”

Although the phrase “legal assistance” is used in the heading to both paragraphs, the text clearly distinguishes between legal assistance simpliciter (para. 7) and legal assistance and representation (see para. 8.1). At the administrative stages of the RSD process the asylum-seeker will be “permitted” to obtain legal assistance, at their own cost or from a pro bono organisation, but a right to representation (which will be free) is only assured for any High Court appeal.

The assurances in the APNV give rise to two questions:

Legal Assistance

There is no reason to doubt the assurance at para. 7.1 of the APNV that RIs will be given information about how to access legal assistance or to suppose that any obstacles will be put in the way of their doing so (subject to the issue about its availability). The issue here is what role the lawyers will be able to play over and above giving advice and assistance with documents.

The UNHCR evidence is that current practice is that lawyers are not permitted to accompany claimants to DGIE interviews or to make any submissions to the RSDC: see paras. 174 and 183 above. That evidence is consistent with the Asylum System CPIN, paras. 4.8.2-3 of which read (so far as material):

“4.8.2

During the meeting with the Rwandan Government on 18 January 2022, HO officials asked about the availability of legal advice and support for asylum seekers during the RSD process. The Director of Response and Recovery Unit at MINEMA explained:

‘Legal assistance is provided for the 2nd level claim [referral to minister for review]. Up to now there have been no cases of an asylum seeker having a lawyer before the RSDC decision because the initial decisions are based on analysis of facts and explanations provided by the asylum seeker …’

4.8.3

HO officials asked whether claimants were allowed to have a legal adviser for the first level claim if they wanted one and the Director explained: ‘No, only at the level where a case goes before the court. ...’”

(That passage appears to contain an internal contradiction about whether legal assistance is available for the appeal to MINEMA; I return to this below.)

The GoR’s response documents initially appeared to confirm that lawyers would not be entitled to accompany claimants to interviews or to make any submissions to the RSDC. Para. 22 of the GoR statement says that legal advice is available during “the administrative phase of the RSD process” but that legal representation is only available if an appeal is made to the High Court. Likewise the primary response says:

“… the Rwandan RSD process is an administrative process with the possibility of appeal at the High Court. During the administrative phase of the process lawyers’ role is limited: they can assist applicants in preparing their submissions to the RSDC but they cannot attend the RSDC sessions. At the High Court level, the RSDC lawyers are permitted to represent the asylum seekers in accordance with the law.”

However, a rather different account is given in the GoR supplementary e-mail, which says:

“In accordance with their constitutional right to due process, the relocated individual/asylum seeker has the right to retain the services of a lawyer at any stage of the asylum process. The legal representative of the asylum seeker is permitted to attend the interviews at DGIE level and any interview at the RSDC.”

Similarly, Mr Williams says at para. 34 of his witness statement that he was told by the DGIE Director that RIs would be permitted to be accompanied by a lawyer at their asylum interview. That evidence does not contradict the GoR’s previous position that a claimant’s lawyers may not make submissions to the RSDC; but it is clearly a departure from what it had previously said about attendance at interviews.

The evidence is clear that under the pre-MEDP practice claimants were not permitted to be accompanied by their lawyers at any interviews: although the supplementary e-mail uses the present tense, it is only reconcilable with the other evidence if it is taken as stating the GoR’s post-MEDP intentions. I am prepared to proceed on the basis that that is now what is intended, though it is very unsatisfactory that it is unacknowledged in the evidence that this is a major change in the process. But it is another matter how easy it will be to introduce a significant change of practice of this kind. There are bound to be questions both about its administrative implementation and about defining the role that the lawyer should be permitted to play in an interview at each level. On any view there will be a need for planning and training of a kind which it is very unlikely has occurred in view of the late emergence of this evidence and which is not likely to be straightforward.

There is some support for that view in an e-mail dated 31 March 2022 from Finnlo Crellin, part of the Home Office team based in Kigali, reporting a meeting with his counterparts in the Rwandan Ministry of Justice. This reads (so far as material):

“The main purpose was to talk them through, and get their initial views on, the latest draught of the Asylum Process NV ... A few key points:

…

•

As expected, 7.2 and 7.3 were the biggest sticking points. They felt that this brought us back to the same issues we discussed last week, particularly in terms of compatibility with, and potential impacts on, their immigration system – including the risk of creating 2 asylum systems. I emphasised this was a red line for us …. Ultimately they agreed to keep the wording as it is but, in Providence’s [evidently a member of the GoR team] words, “we’ll see how it works in practice” …”

It is not necessary to understand the nuances: I refer to this passage only to illustrate that the assurances about legal assistance (the subject of paras. 7.2 and 7.3 of the APNV) were regarded by the relevant GoR officials as likely to be difficult to implement.

The position as regards the appeals to MINEMA is rather different. We are not in this context concerned (at least generally) with the presence of a lawyer at an interview but with whether they are entitled to make submissions to the Minister. As to that, the evidence tends to suggest that that was not permitted pre-MEDP; but para. 5.2 of the APNV appears to confirm that it will be the case in future. I see no reason to doubt that it is intended that that assurance will be complied with, but again adjusting to a change of this kind will not be straightforward.

Accessibility of legal assistance/representation

It is unrealistic to suppose that RIs will be in a position to pay for legal assistance at the administrative stage of the RSD process. They will accordingly be dependent on pro bono assistance. The GoR says that such assistance will be available from one of two NGOs in Rwanda, the Prison Fellowship (“PFR”) and the Legal Aid Forum (“LAF”).

Mr Bottinick gives evidence about the assistance available from these two NGOs at para. 100 of LB 2. In summary, he says that PFR has only a single legal officer who regularly provides assistance in the RSD process, as part of a number of other duties: she is not a qualified lawyer, though she has a law degree. When she is not available, there is another lawyer at PFR who can act as backup, though this is not part of his regular work. (PFR also has some legal officers who are engaged in the etm project referred to above but these are not available to work on RSD cases.) As for LAF, this primarily operates in camps and urban centres outside Kigali. Mr Bottinick says that his colleagues in Kigali have spoken to LAF for the purpose of his statement and have been told that four of its lawyers have some previous experience in assisting in the RSD process but that they currently do so very rarely: it routinely refers its RSD cases to PFR. He says that it was unfortunate that in its assessment visits the Home Office team had met LAF but not PFR.

The GoR’s response appears in several places:

I should also refer to para. 8.1 of the PDA, which states:

“During the initial application process RIs can seek legal assistance at their own cost for private advice or through NGOs during the initial stage. Two NGOs, the Legal Aid Forum and Prison Fellowship, have confirmed to the GoR they will provide advice at no cost if RIs ask for it at the initial stage. The GoR does not have a formal agreement with them for this.”

The GoR’s statements on this aspect are unsatisfactory. It is in my view clear from UNHCR’s evidence that the NGOs with which the GoR says it has a “partnership” – i.e. PFR and LAF – 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. The statements quoted at (2) and (4) above give general (though in fact different) numbers for the lawyers/legal officers qualified to give advice in this field, but it does not follow that they would be available to advise RIs, and the effect of Mr Bottinick’s evidence is that they are in fact engaged on other work. It is true that the answer to qu. (6) also refers to a “standing agreement with Rwanda Bar Association”, but no details are given of the agreement and it is far from clear that the agreement is relevant to legal assistance of the kind with which we are concerned here. It is also true that in the statement quoted at (1) the GoR speaks of “building on” its existing partnerships and funding the organisations in question so as to increase their capacity. But that is extremely unspecific, and it appears from the PDA that there are in fact no formal agreements in place. I believe I should accept the UNHCR evidence as stating the current position.

I turn to the question of legal representation for RIs on any appeal to the High Court, as promised in para. 8.1 of the APNV. The only concern expressed by UNHCR about this commitment is whether there are a significant number of qualified lawyers in Rwanda with the expertise to conduct such appeals, given that none have so far been brought. Figures supplied by the GoR in the tabular response (see qu. 9) show that there are over a thousand “senior registered lawyers” in Rwanda and a further 300 “intern advocates”. Refugee law will have been part of the curriculum in their original training and the Ministry of Justice has agreed with the Institute of Legal Practice and Development for courses to be run on refugee law.

Mr Bottinick does not in LB 3 respond to that evidence. In my view the most that can be said is that lawyers conducting the first appeals in the High Court are likely to have a steep learning curve, but it does not follow that they will be unable to give proper representation.

The great majority of asylum-seekers relocated to Rwanda under the MEDP will not be fluent in any of its official languages (English, French and Kinyarwanda). They will therefore need the services of an interpreter (a) to make their initial claim; (b) in their interview with DGIE; (c) in any subsequent interview required by the Eligibility Officer or the RSDC; (d) for the purposes of any appeal to the Minister or to the High Court; and (e) in any dealings with lawyers or other advisers.

That need is recognised in the APNV. Para. 9 reads:

“Interpreter

9.1

If a Relocated Individual requires it, an interpreter will be provided, free of charge, whenever the Relocated Individual meets with a legal representative provided to them free of charge in accordance with 8.1 above or a representative or employee of the Government directly involved in the Relocated Individual’s asylum application.

9.2

All written correspondence and information that a Relocated Individual receives concerning their claim and the asylum process will be translated by an appropriate interpreter, free of charge, if they require it to understand.

9.3

A Relocated Individual who has the opportunity to consider a written transcript of their interview will have the assistance of an interpreter, free of charge, if needed for understanding.”

(I note that para. 9.1 does not provide for access to an interpreter in making the initial claim; but that is unlikely to matter in practice if the claimant has access to a lawyer for that purpose.)

It appears from para. 9 of the PDA that although the GoR has contracted for the provision for RIs of interpreters in a number of languages, including Arabic, it has not been able to do so for a number of other languages, including Kurdish, Vietnamese and Albanian (all languages spoken by the Claimants in this case) and also Pashto (which will be relevant to many Afghan nationals). The PDA says:

“To mitigate this, they have agreed to use the virtual interpretation facilities the Home Office have offered to them through the Home Office contract with The Big Word (TBW). They have opted for telephone interpreting, face to face interpreting by way of an online platform and translation services (for the SOPS, orientation pack and other written documents where necessary). They intend to use TBW only where necessary and when their current contractors cannot provide a service. This was set up ahead of the scheduled charter flight on 14/6 and has been tested to ensure the service is functioning. Longer term, GoR will look to negotiate their own separate contract with the TBW with assistance from the dedicated HO team.”

It appears from Mr Williams’ witness statement that this difficulty only emerged in a meeting with the DGIE Director on 26 May.

In para. 35 of LB 3 Mr Bottinick expresses concerns about this arrangement, on the basis that not all officials involved in the process speak fluent English. An interpreter based in Rwanda could assist such officials by using Kinyarwanda where necessary, and interpreting anything said in Kinyarwanda to the asylum-seeker, which a remote interpreter could not do. I understand the concern, and remote interpretation is no doubt sub-optimal; but there is no reason to suppose that a DGIE official who was not fluent in English would be asked to conduct an interview in which English was the medium of interpretation. This problem by itself may not undermine the fairness or effectiveness of the system, but it reinforces the need for those conducting an interview to have the skills and experience to cope professionally with difficulties of this kind.

It is obvious, and undisputed, that officials making asylum decisions, or otherwise contributing to the process, need to be properly trained. That is recognised in the APNV, para. 4.2 of which reads:

“Asylum decisions will be taken by decision-makers who are appropriately trained to take a decision on an asylum claim in accordance with the Refugee Convention and are able to seek advice from senior officials or external experts if necessary.”

However, there are two particular reasons why proper training is essential in the circumstances of this case.

First, as noted above, 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. On its own figures the total number of claims determined by the RSDC between 2019 and June 2022 is 152, of whom 115 were from DRC and Burundi, and another 20 from Eritrea. It has determined no claims requiring consideration of conditions in the countries from which the Claimants originate – that is, Syria, Iraq, Iran (most of these being ethnic Kurds), Vietnam, Sudan or Albania. These Claimants are likely to be broadly representative of the cohort of asylum-seekers liable to be relocated under the MEDP, except that they include no-one from Afghanistan (from which the RSDC has so far determined only one claim). A table exhibited to LB 3 gives figures, on the basis of the information available to the UNHCR (which it accepts may be incomplete), of all cases in the process over the relevant period, including those not yet determined: although that shows a handful of cases from Syria, Sudan and Afghanistan, the proportionate picture is the same. The RSDC, together with DGIE and the Eligibility Officers in so far as their work feeds into its decisions, will be effectively starting from scratch in acquiring an understanding of the situations in those countries.

Second, the system in place at the date of the conclusion of the MEDC had the serious deficiencies identified in the earlier discussion, which need to be addressed by effective training.

In those circumstances particularly thorough and effective training is required in order that the relevant institutions and individuals can deal properly with RIs. Para. 18 (i) of the UNHCR Note reads:

“There is a need for an objective assessment of the fairness and efficiency of the asylum procedures, followed by a range of capacity development interventions including, but not limited to, sustained capacity building and training for all actors working in the Rwandan national asylum system [my emphasis].”

I agree.

The question of the current level of training of those operating the system is addressed at paras. 89-98 of LB 2. Mr Bottinick summarises his evidence in para. 89 as follows:

“UNHCR has observed serious shortcomings in knowledge and training regarding RSD among relevant officials at all levels. UNHCR considers that this lack of training gives rise to a serious risk that refugees will be refused recognition by the Rwandan Government and refouled.”

He goes on to say that UNHCR provided a three-day training course to officials in June 2017 but that, although it had repeatedly offered further training, the offers had not been taken up until December 2021 when it was invited to co-facilitate, together with MINEMA, the Rwanda Law Reform Committee and the University of Rwanda, a workshop for DGIE and RSDC staff and officials. The workshop was intended to last four and a half days but in fact lasted just over three because of the unavailability of the intended attendees. Paras. 93-97 of LB 2 read as follows:

“93.

The training was attended by only 15 participants. Out of 11 RSDC members at the time, eight attended, but even some of those could only attend partially because of their conflicting professional schedules and ministerial commitments (and one attended for only two days). The RSDC chair (who was new to the process at the time and had not yet attended any RSD-related adjudication) and secretary missed at least the first day of the training which covered basic principles of refugee law.

94.

At the time of the training, most RSDC members were new, had no prior exposure to RSD and had not attended RSDC deliberations. One of the officials remarked that he did not understand why he was required to undertake RSD given that his departmental role was not connected to asylum.

95.

The training was targeted at an extremely basic level. It included, in the main, general principles of refugee law, in addition to brief and basic training on assessing individual claims and interviewing techniques. My colleagues felt that the basic knowledge of the attendees did not allow them to cover crucial areas such as how to deal with claims based on membership of a particular social group.

96.

The participants’ lack of relevant knowledge and skills was particularly apparent during a simulation of RSDC interviews and decision making. Observations from UNHCR’s trainers noted that the participants lacked interviews skills and had very limited or no understanding of how to assess refugee status. In a simulation involving a husband and wife, the ‘couple’ were interviewed together and the husband was allowed to answer for the wife. In addition, there was no opportunity for the ‘asylum seeker’ to express relevant gender-based violence related elements of her claim. It was also noted that elaborate leading questions were asked by participants and that the ‘asylum seeker’ was not given an opportunity to respond in full to questions, nor were they alerted to adverse credibility points. When making their assessments of the cases, participants were unable to demonstrate knowledge of how to assess credibility and COI; or of key concepts in refugee law. This is not surprising given that the participants are senior civil servants with no background in RSD.

97.

In UNHCR’s view, this short (and truncated) one-off workshop cannot be considered adequate training to ensure fair RSD decision making, especially for training participants with little or no prior knowledge and experience of refugee law. RSDC members still at the end of the training lacked by some distance the requisite knowledge and skills to make fair, reliable RSC decisions. RSDC members require significant further in-depth on-the-job training and shadowing of appropriate procedures. However, in UNHCR’s view, while that is necessary to rectify some of the problems in the RSDC process, it would be far from sufficient: the non-specialist composition of the RSDC is inimical to fair, reliable RSD decision-making. UNHCR was further concerned by attitudes expressed by Rwandan authorities during this training that DGIE are within their rights to deny access to its territory or to RSDC procedures if they consider the profile of an individual applicant unpalatable, including on unspecified grounds of national security. The Rwandan staff and officials present at UNHCR’s December 2021 training did not appear to consider such ‘screened out’ persons as asylum seekers or consider that their deportation would constitute refoulement.”

He says at para. 91 that he is unaware of any other outside body providing training for participants in the process.

The GoR’s response to that evidence appears in a number of places and cannot readily be summarised.

I start with the GoR statement. Para. 23 says:

“The Government of Rwanda has … taken measures to build the expertise of persons involved in processing the claims of asylum seekers. The Rwanda Institute of Legal Practice and Development (ILPD) will be providing bloc courses, periodic trainings and workshops on refugee law and other related laws to Eligibility Officers, RSDC Members, lawyers, and high court judges.”

The primary GoR response cross-refers on this aspect to the tabular response. This has several answers referring to training.

First, qu. 4 asked what training “interviewing officers” – i.e. the DGIE officers who conduct the asylum interview – receive. The answer is:

“The interviewing officers at DGIE have received different trainings on international protection of refugees, international law of refugees, rights-based approach to migration law, national laws relating to refugees and migrants, interview skills, etc. - These various trainings were provided through:

1.

Rwanda Institute of legal practice and development (ILPD).

2.

The International Institute of Humanitarian Law/San Remo, Italy.”

Second, qu. 7 asked various questions about the RSD Committee, including what training they have received. The answer is:

“They have received training on refugee status determination. … In addition to the periodic training/workshops on international refugee law and asylum process offered by UNHCR (the latest trainings by UNHCR were offered in 2018 and 2021). UNHCR also offered training to RSDC members at International Institute of Humanitarian Law/San Remo, Italy. The members also bring on board complementary expertise from their respective specialized institutions. The diverse expertise which is uniquely relevant to the work the committee ranges from human rights perspective, diplomacy & global trends, refugee management, security and migration matters, etc. So, RSDC is purposely comprised of members with varying knowledge and expertise that enables objective consideration of asylum claims.”

Third, qu. 26 asked if there was “any record of how many people attended the UNHCR training for the RSD”. The answer is:

“Two Eligibility Officers and one RSDC members were trained at San-Remo. 9 out of the 11 RSDC members participated in the training co-organized by MINEMA and UNHCR in December 2021 and 10 out of 11 in the one organized by MINEMA in December 2018. Two RSD members completed online training in eligibility and RSD process. These training normally serves to harmonize on principles that guide the decision making. In addition, each of the RSDC members has completed training in her/his area (human right, humanitarian protection, international justice, migration, socio economic inclusion, etc) that build analytical skills for the member to contribute efficiently during the committee sessions.”

The primary GoR response adds two further points. First, it disputes Mr Bottinick’s account that not everyone attended the whole of the December 2021 workshop, saying that “the figures provided [in the tabular response] are accurate to our knowledge”. Second, the response to para. 144 of Bottinick 2 says:

“A training for RSD members is also being organized and shall be facilitated by local learning institutions (University of Rwanda and Institute of Legal Practice and Development) but also by institutions concerned by the RSD process including MINEMA, MINIJUST, NCHR, DGIE, MINAFFET.”

Finally, Mr Williams says at para. 44 of his witness statement, amplifying what appears in the relevant part of the PDA:

“The DGIE Director informed me that all DGIE Immigration Officers are trained on asylum and international protection, including how to register asylum applications, conduct asylum interviews, and write reports for the Refugee Status Determination Committee (RSDC). The Director informed me on 17 June 2022 that DGIE Immigration Officers undergo a minimum of six months training at a dedicated training college followed by on-the-job training once they commence their duties. The Director also told me on 17 June that most of the training is in-house but DGIE Immigration Officers have received training from international organisations including the International Organization for Migration (IOM) and international partner countries.”

The Secretary of State’s evidence is responded to in para. 34 of LB 3. I need not reproduce it in full. In short:

In my judgment, Mr Bottinick’s evidence raises a clear case to answer that the level of training made available to the key players in the asylum process – the DGIE officials who conduct the interviews, the Eligibility Officers, the members of the RSDC, the MINEMA Minister or the officials who advise him or her on appeals – is not sufficient to equip them to perform their functions properly. I do not believe that the Secretary of State’s evidence, based on the information obtained from the GoR, provides a satisfactory answer. In particular:

It is not ideal that this important point should turn on the Court’s assessment of hearsay evidence from the GoR produced at short notice in response to the evidence of Mr Bottinick. But in truth the nature and extent of the training of officials involved in the asylum process should have been assessed in depth, with reference to documents and records so far as available, as part of the investigations carried out when the MEDP was still in gestation.

I start by acknowledging that there is nothing in the evidence that would justify the conclusion that the GoR has entered into the commitments in the MoU and the APNV in bad faith. There is no reason to suppose that it does not wish to ensure that RIs have their asylum claims determined fairly and effectively. But aspiration and reality do not necessarily coincide. As we have seen, the RSD process is a recent creation and it has so far had little experience of dealing with asylum-seekers with the characteristics of those liable to be relocated under the MEDP. The UNHCR evidence in my view clearly shows that there are important respects in which it has not so far reliably operated to international standards. It is its case, and the Claimants’, that even if there is a will to make the necessary changes they cannot be achieved in the short term and certainly have not been achieved yet. At paras. 143-144 of LB 2 Mr Bottinick says:

“143.

Rwanda’s serious capacity issues cannot be addressed within a short space of time. …

144.

Moreover, at the time of making this statement, UNHCR is unaware of any steps being initiated that might, after a sustained period of capacity building, eventually permit certain of the commitments in the Notes Verbales and MOU to be fulfilled. UNHCR is not, for example, aware of interpreters, lawyers or decision makers being hired or trained by the Rwandan Government at present.”

I should say that when Mr Bottinick refers to “capacity” it is clear from the context that he is not referring primarily to ability to cope with numbers but to skills and experience more generally. It is not, therefore, an answer to say that the GoR can decline to accept more RIs than the RSD process can cope with in the early stages of the MEDP.

The essential question is thus to my mind whether the changes necessary to ensure compliance with the GoR’s assurances had been, or in any event would be, implemented before relocations under the MEDP began to take place: that was of course initially intended to be on 14 June 2022 (only two months after the conclusion of the MEDP), although as I have said the position should now be judged as at the dates of the hearing in the Divisional Court.

I have not found it entirely straightforward to answer that question. The evidence is not as complete as could be wished; and, as already noted, we have not had the advantage of the kind of detailed examination, with the benefit of expert evidence and cross-examination, that would have been possible in the FTT if the Secretary of State had not certified the Claimants’ human rights claims. In the end, however, I have reached the conclusion that the Rwandan system for refugee status determination was not, as at the relevant date, reliably fair and effective.

In reaching that conclusion I have taken into account the totality of the defects identified in the discussion above. Those which have weighed with me particularly, not least because they are not clearly addressed in the APNV, are:

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). But 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: see in particular paras. 245-260.

Like the Master of the Rolls (see his Issues 8 and 9), I believe that that conclusion means that it is unnecessary to consider separately the issues relating to the adequacy of the inquiries conducted by the Secretary of State, whether by reference to Ilias or to Tameside. I would only make two observations.

On the one hand, I would accept 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. There were evidently dedicated civil servants genuinely trying to establish how the RSD process worked and to obtain assurances that addressed the perceived problems.

On the other hand, however, 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: the late emergence of the problem about interpreters is an illustration of this. We were referred by Mr Husain to a review of the Rwanda CPINs which was undertaken in July 2022 for the Independent Advisory Group on Country Information (“IAGCI”). IAGCI acts on the instructions of the Independent Chief Inspector of Borders and Immigration, to whom it provides advice to the Chief Inspector of the UK Border Agency to allow him to discharge his obligation under section 48 (2) (j) of the UK Borders Act 2007. The researcher responsible for the review criticised various aspects of the way in which the Asylum System CPIN was prepared, including “very limited critical information on the Rwandan asylum system” and “fundamental gaps of information and unanswered questions with regards to procedural practicalities and implications”, together with more specific methodological criticisms of the conduct of the interviews contained in the Annex A CPIN. As the Divisional Court pointed out at para. 59 of its judgment, and as Sir James emphasised in his oral submissions, the Chief Inspector has not himself made any recommendations, and it is not known whether or to what extent he will endorse those criticisms. But I note that they are consistent with my own conclusion.  I should also say in this connection that I believe that it is unfortunate that officials did not engage with UNHCR on their first visit to Rwanda in January 2022. Their initial intention was to do so; but it seems that, for reasons that are unclear, they did not receive the necessary clearance from the Secretary of State.

I have not so far addressed the reasoning of the Divisional Court. It says, at paras. 64-66:

“64.

In the present case we consider the Home Secretary is entitled to rely on the assurances contained in the MOU and Notes Verbales, for the following reasons. The United Kingdom and the Republic of Rwanda have a well-established relationship. This is explained in the witness statement of Simon Mustard, the Director, Africa (East and Central) at the Foreign, Commonwealth and Development Office. This has comprised a development partnership set out in various agreements (referred to as Development Partnership Agreements) since 1998.  The relationship is kept under review.  In 2012 it was suspended by the United Kingdom government in response to Rwanda’s involvement in the so-called ‘M23 Rebellion’ in the Democratic Republic of Congo, and in 2014 the relationship was further reviewed in response to the assassination in South Africa of a Rwandan dissident.  Since then, the United Kingdom has continued to provide Rwanda with financial aid, but this has been tied to specific activities.  Thus, while there is a significant history of the two governments working together, the Rwandan government has reason to know that the United Kingdom government places importance on Rwanda’s compliance in good faith with the terms on which the relationship is conducted. 

65.

The terms of the MOU and Notes Verbales are specific and detailed.  The obligations that Rwanda has undertaken are clear.  All, in one sense or another, concern Rwanda's compliance with obligations it already accepts as a signatory to the Refugee Convention.  The Claimants have placed particular emphasis on whether the Rwandan asylum system will have the capacity to handle asylum claims made by those who are transferred under the terms of MOU.  It is a fair point that, to date, the number of claims handled by the Rwandan asylum system has been small.  It is also fair to point out, as Mr Bottinick has, that it will take time and resources to develop the capacity of the Rwandan asylum system.  However, significant resources are to be provided under the MEDP, and by paragraph 3.3 of the MOU the number of persons that will be transferred will depend on the consent of the Rwandan government, taking account of its capacity to deal with persons in the way required under the MOU and the Notes Verbales.  The MOU also contains monitoring mechanisms in the form of the Joint Committee (paragraph 21 of the MOU) and the Monitoring Committee (paragraph 15 of the MOU).  For now, at least, there is no reason to believe that these bodies will not prove to be effective.  Lastly, the MOU makes provision for significant financial assistance to Rwanda. That is a clear and significant incentive towards compliance with the terms of the arrangement.

66.

Moreover, Mr Mustard explains that HM Government is satisfied that Rwanda will honour its obligations.  At paragraph 20 of his statement, he says this:

‘The British High Commission in Kigali led initial conversations with the [Government of Rwanda] regarding the [MEDP] and participated in negotiations in support of the Home Office. Since these negotiations began, there has been a renewed focus on our bilateral relationship with an increase in contact at an official and ministerial level.  Prior to signing the agreement, Home Office officials visited the Rwanda on many occasions, meeting government and non-governmental interlocutors, and carried out further discussions virtually.  The Rwandan Permanent Secretary to the Ministry of Foreign Affairs also led a delegation to London for further talks. These negotiations have been conducted transparently and in good faith throughout.  In light of the considerations described in this witness statement, and the manner in which the negotiations [with] our Rwandan counterparts were conducted, we are confident that Rwanda will honour its commitments under the MEDP.’

We consider that we could go behind this opinion only if there were compelling evidence to the contrary.  We do not consider such evidence exists.”

As discussed at paras. 128-129 above, it seems from that passage that the Divisional Court erred by approaching its task in this part of its judgment as one of review rather than seeking to reach its own conclusion. But even if I am wrong about that, I do not believe that its reasoning can be supported. It did not seek to engage with the details of UNHCR’s criticisms of the RSD process. As I read it, the principal reason why it thought that this was unnecessary was the weight that it attached to Mr Mustard’s assessment, representing the view of the UK Government, that the GoR would honour its commitments under the MEDP. As to that, I would accept that great weight should indeed be given to the Government’s assessment that the GoR negotiated the MEDP in good faith and with a genuine willingness to comply with its obligations under it, for the reasons which the Court gives at paras. 64-65 of its judgment. I agree that there is no reason to doubt the genuineness of the GoR’s intentions: it is for that reason that I have found that there is no risk of RIs being denied access to the RSD process (see para. 155 above). But the real issue here is not the good faith of the GoR at the political level but its ability to deliver on its assurances in the light of the present state of the Rwandan asylum system.

The Divisional Court does in fact in para. 65 acknowledge that “it will take time and resources to develop the capacity of the Rwandan asylum system”; but it believes that that concern is sufficiently answered by the facts that significant resources are to be provided to the GoR under the MEDP and that it has the right to control the numbers of RIs admitted so that they do not exceed the capacity of the RSD process at any given time. However, the provision of resources does not mean that the problems in the Rwandan system can be resolved in the immediate term; and even if the flow of RIs is restricted for the time being in order for improvements to take effect that does not justify the denial of a fair and effective asylum system to the earlier arrivals.

In short, the relocation of asylum-seekers to Rwanda under the MEDP would involve their claims being determined under a system which, on the evidence, has up to now had serious deficiencies, and at the date of the hearing in the Divisional Court those deficiencies had not been corrected and were not likely to be in the short term.

The result of my conclusion in the preceding section is that there are substantial grounds for believing that there is a real risk that the asylum claims of RIs may be wrongly refused. On the face of it, it would appear to follow that there was a real risk of them being refouled. Where an asylum-seeker’s claim is rejected the country in question will typically require them to leave the country (in the absence of any other basis on which they might claim residence), and since they will have been found to be at no risk in their country of origin, there is no reason why they should not be returned there; and even if they are in the first instance returned to some other country that does not exclude the possibility of indirect refoulement. The reason why the ECtHR in Ilias insisted on the need to establish that there was an adequate asylum system in the country of return is that the existence of such a system is regarded as an essential protection against the risk of refoulement.

It may be, however, that the Secretary of State wishes to submit that a finding that the Rwandan asylum system is inadequate would not in the circumstances of this case mean that there is a real risk of refoulement. At the start of the part of her skeleton argument dealing with the safety of Rwanda she advances five “overarching submissions”. The fifth, which appears at para. 8 of the skeleton, is that the Court should not seek to interpret Rwandan asylum law or predict how it might apply in particular cases because it was sufficient to rely on para. 9.1.1 of the MoU. The paragraph continues:

“Furthermore, even if some deficiency in Rwandan asylum law were identified ... it would not give rise to a risk of refoulement or Article 3 ill treatment unless there were to be evidence of intention to send the asylum seeker back to their country of origin”.

A footnote reads:

“Rwanda has no returns agreement with any of the countries in question ... The MOU provides, at para 10.3, for relocated individuals to apply for residence even if refused asylum.”

The statements made in the footnote are referenced to passages in the witness statements of Mr Williams and Mr Armstrong. It is sufficient to refer to the latter. Mr Armstrong says, at para. 85:

“The statement of Chris Williams sets out what he was told by DGIE officials about what would happen to relocated individuals who are refused asylum. Senior officials from DGIE in Rwanda confirmed that if an individual who was relocated to Rwanda under the MEDP had their asylum claim refused and all their appeal rights were exhausted, they would be eligible to be issued a Resident Card. They said that they envisaged that all relocated individuals would be permitted to remain in Rwanda and that no relocated individuals would be forcibly returned to their country of origin. They also explained that the Government of Rwanda did not have returns agreements or arrangements in place with any country with the exception of neighbouring countries.”

The statements there attributed to the DGIE officials are consistent with statements made in the GoR statement (see para. 5) and the tabular response (answers to qus. (12) and (22)).

Para. 10.3 of the MoU, also referenced in the footnote, reads:

“10.3

For those Relocated Individuals who are neither recognised as refugees nor to have protection need in accordance with paragraph 10.2, Rwanda will:

10.3.1

offer an opportunity for the Relocated Individual to apply for permission to remain in Rwanda on any other basis in accordance with its domestic immigration laws and ensure the Relocated Individual is provided with the relevant information needed to make such an application;

10.3.2

provide adequate support and accommodation for the Relocated Individual’s health and security until such a time as their status is regularised or they leave or are removed from Rwanda.”

Reference should also, I think, be made to para. 10.4, which reads:

“For those Relocated Individuals who are neither recognised as refugees nor to have a protection need or other basis upon which to remain in Rwanda, Rwanda will only remove such a person to a country in which they have a right to reside. If there is no prospect of such removal occurring for any reason Rwanda will regularise that person’s immigration status in Rwanda.”

Although the submission which I have quoted from the skeleton argument is ostensibly addressed only to a situation where there is “some deficiency in Rwandan asylum law”, the logic of the point might be thought to apply to a situation where the asylum system as a whole was inadequate and thus liable to produce wrong outcomes: that is, it might be said that its inadequacy did not matter “unless there were to be evidence of intention to send the asylum seeker back to their country of origin”.

The issue was not addressed by the Divisional Court, since in the light of its conclusion on the adequacy of the Rwandan asylum system it did not arise.

The submission in question was not developed in the hearing before us. In his opening submissions Mr Husain referred to it but only to contend that the point was not open to the Secretary of State because it had not been pleaded in a Respondent’s Notice. He said that it was in any event bad but that he would address it in his reply if necessary. In the event, Sir James Eadie made no oral submissions on this point and Mr Husain did not revert to it.

Mr Husain is right to say that this submission should have been raised in a Respondent’s Notice. With some hesitation, however, I think that I should consider it, although in the absence of oral argument I can and should only deal with it briefly.

I start by setting out what is said in the GoR statement and the tabular response as referred to above:

In my view none of those statements can be treated as a reliable assurance that an RI whose asylum claim is refused will be permitted to remain in Rwanda and enjoy basic rights equivalent to those granted by the Refugee Convention. I would make three points.

First, as to the possibility of the RI being granted legal residence on other grounds, it is – unsurprisingly – not said that this occurs in every case (the tabular response says “most”, but the GoR statement says “a significant portion”). It is in fact easy to see how the grant of “work/business permits” and “dependent/relatives permits” may be appropriate for migrants from neighbouring countries, who have historically constituted the great majority of asylum-seekers in Rwanda; but the circumstances of RIs will be wholly different. Whether or not some RIs may be granted legal residence on another basis, the question remains of what will happen to those who are not, who are likely to be the great majority.

Second, para. 10.3.2 refers in terms to the removal of RIs whose asylum claims fail and who do not qualify for a right to remain on any other basis. Para. 10.4 provides that they will only be removed to a country where they have the right to reside, but that would of course include their country of origin, in which the GoR would (ex hypothesi wrongly) have decided that they did not face a risk of persecution or other ill-treatment.

Third, while there is no reason to doubt the statement that Rwanda has at present no return agreements with the countries of which RIs are likely to be nationals, and no intention “at the moment” to conclude any such agreement, that falls well short of a guarantee that it will not do so in future. In any event direct return is not the only way that refoulement can occur, as the cases referred to at para. 146 illustrate.

In short, even if the Secretary of State is seeking to advance the (prima facie surprising) argument that it does not matter if Rwanda’s asylum system is inadequate because RIs whose claims are wrongly refused will in every case be allowed to stay I would not accept that argument.

As noted above, the Claimants also allege that the repressive nature of the Rwandan regime means that asylum-seekers and refugees will be at risk of inhuman and degrading treatment within the meaning of article 3 if they engage in protests against or other criticisms of the GoR. The Master of the Rolls sets out the nature of the case in rather more detail at para. 103 of his judgment.

This aspect of the case is addressed in the judgment of the Divisional Court at paras. 73-77 under the heading “Conditions in Rwanda generally”. It acknowledges that there is clear evidence that the GoR is intolerant of dissent; that there are restrictions on the right of peaceful assembly, freedom of the press and freedom of speech; and that political opponents have been detained in unofficial detention centres and have been subjected to torture and article 3 ill-treatment short of torture. However, it concludes that there is no real risk that persons returned under the MEDP would be ill-treated even if they expressed dissent or protest (whether about their own treatment or on other issues). Para. 77 concludes:

“… [T]he Claimants’ submission is speculative. It does not rest on any evidence of any presently-held opinion. There is no suggestion that any of the individual Claimants would be required to conceal presently-held political or other views. The Claimants’ submission also assumes that the response of the Rwandan authorities to any opinion that may in future be held by any transferred person would (or might) involve article 3 ill-treatment. Given that the person concerned would have been transferred under the terms of the MEDP that possibility is not a real risk. It is to be expected that the treatment to be afforded to those transferred will be kept under the review by the Monitoring Committee and the Joint Committee (each established under the MOU). Further, the advantages that accrue to the Rwandan authorities from the MEDP provide a real incentive against any mis-treatment (whether or not reaching the standard of article 3 ill-treatment) of any transferred person.”

In the light of the conclusion that I have reached on the refoulement issue I do not need to decide whether the Divisional Court’s conclusion was correct, and I prefer not to do so since we heard only very limited oral submissions about it. I confine myself to two observations.

First, I respectfully doubt whether it is relevant that the Claimants themselves have not been shown to hold any opinions of a kind which are likely to attract adverse attention from the GoR. (Footnote: ⁶) We are, as I have said, concerned with a generic challenge, and the question must be whether there is a real risk that RIs generally may suffer serious ill-treatment in Rwanda if they engage in any protest or express dissent.

Second, while I see the force of the point made by the Divisional Court that the GoR is likely to be very chary about any ill-treatment of RIs, it is right to note that documents produced by the Secretary of State show officials expressing concern about this very aspect: an FCDO official reviewing the draft CPIN noted that asylum-seekers “would need to be 100% compliant and subservient to very stringent top-down rules in Rwanda”.

I note the Master of the Rolls’ observation at para. 106 above that it would have been better if the Divisional Court had dealt with this aspect as an undifferentiated part of the Soering test. For myself, I see some advantages in analysing it separately because the nature of the article 3 risk is different. But I agree that they are not wholly distinct. The nature of the Rwandan government is a relevant background consideration in considering some aspects of the RSD process.

For the reasons given above, I believe that the evidence before the Divisional Court established that there were substantial grounds to believe that asylum-seekers relocated to Rwanda under the MEDP were at real risk of refoulement, and that accordingly such relocation would constitute a breach of article 3 of the ECHR and contravene section 6 of the 1998 Act.

I have thus reached the same conclusion as the Master of the Rolls about the overall safety of Rwanda issue. The Lord Chief Justice has reached the opposite conclusion. It will be sufficiently apparent from my reasoning above why I respectfully take a different view from him.

I have not found it necessary to address separately each of the eleven issues identified by the Master of the Rolls under this heading, though I have in substance covered most of them in my analysis. However, I should say something more about his Issues 10 and 11.

The effect of my conclusions thus far is that a decision to remove an asylum-seeker to Rwanda would be unlawful because it would involve a breach of their article 3 rights. It follows that a published policy which positively authorised or approved such removals would also be unlawful: the relevant principles derive from the decision of the House of Lords in Gillick v West Norfolk and Wisbech Area Health Authority [1986] AC 112 and have recently been restated by the Supreme Court in R (A) v Secretary of State for the Home Department [2021] UKSC 37, [2021] 1 WLR 3931.

The question then is whether the Secretary of State has promulgated such a policy. On 9 May 2022 she published version 6 of Guidance to case-workers headed “Inadmissibility: safe third country cases” (“the Inadmissibility Guidance”). (Footnote: ⁷) As regards relocation to Rwanda, the Guidance says:

“If a case assessed as suitable for inadmissibility action appears to stand a greater chance of being promptly removed if referred to Rwanda (a country with which the UK has a Migration and Economic Development Partnership (MEDP), rather than to the country to which they have a connection, TCU should consider referring the case to Rwanda. An asylum claimant may be eligible for removal to Rwanda if their claim is inadmissible under this policy and (a) that claimant’s journey to the UK can be described as having been dangerous and (b) was made on or after 1 January 2022. A dangerous journey is one able or likely to cause harm or injury.

...

Those progressed for consideration for relocation to Rwanda under the MEDP will be taken from the detained and non-detained cohort and be identified in line with processing capacity.  Priority will be given to those who arrived in the UK after 9 May 2022.

...

Decision makers must take into account country information of the potential country/countries to where removal may occur in deciding whether referral into a particular route is appropriate in the particular circumstances of that claimant.”

Country information relating to Rwanda was published on the same day in the form of the CPINs identified at para. 139 above. Consistently with the guidance that decisions must be taken by reference to the particular circumstances of each claimant, these do not purport to reach a definitive conclusion on whether Rwanda is a safe third country. However, they state the views of the CPIT on a number of issues relevant to that question, the most important of which are summarised as “key judgments” in the Assessment CPIN. These include, under the heading “refoulement”, the judgment that:

“There are not substantial grounds for believing that a person, if relocated, would face a real risk of being subjected to treatment that is likely to be contrary to Article 3 ECHR by virtue of being refouled or returned to a place where they have a well-founded fear of persecution.”

The Master of the Rolls considers that in view of his conclusion on the substantive issue of the safety of Rwanda it is unnecessary to decide whether the Inadmissibility Guidance is unlawful: see para. 114 of his judgment. However, I think I should say that I believe that the Guidance and the Assessment CPIN, taken together, do constitute policy guidance which, in the light of the conclusions reached above, is indeed unlawful.

The Divisional Court held that the Inadmissibility Guidance was lawful: see para. 72 of its judgment. But that simply reflects its substantive decision on the issue of the safety of Rwanda.

Ground 6 of the consolidated grounds of appeal, advanced before us by Ms Sonali Naik KC, reads:

“The Court was wrong to conclude that R’s inadmissibility policy on removals to Rwanda was not unlawful either under the conventional Gillick test or under a Gillick test necessarily modified in cases involving a real risk of Article 3 ECHR breach.”

Since for the reasons which I have given I would hold that the policy was indeed unlawful on what I believe to be a conventional basis, I see no advantage in considering the alternative submission that “the Gillick test” requires modification in cases based on article 3, though I feel bound to say that I had some difficulty in understanding it.

I have already made the point that we are not in this appeal concerned with the challenge to the Secretary of State’s decision to certify the human rights claims made by (most of) the Claimants under paragraph 19 of Schedule 3 to the 2004 Act: see para. 130 above. However, I agree with the Master of the Rolls that it follows that the Secretary of State’s decisions certifying the individual human rights claims could not be sustained on this ground, as well as on the grounds on which they were in fact quashed by the Divisional Court: see para. 116. As to whether it would also have followed, if the Court had reached the same conclusion, that the appeal to the FTT would have proceeded in priority to the generic challenge (see his para. 117), I refer to what I say at para. 131 above.

THE REMAINING ISSUES

Although our conclusion on the safety of Rwanda issue means that the Rwanda policy must be declared unlawful, it is necessary to consider the remaining issues raised by the Claimants’ grounds of appeal, not only in case of an appeal to the Supreme Court but also because they will remain relevant if the defects in the Rwandan asylum system are in due course resolved. I gratefully adopt the Master of the Rolls’ categorisation of the remaining issues, and I take them in turn. Responsibility for arguing these issues on behalf of the Claimants was divided between different counsel; they were argued on behalf of the Secretary of State by Lord Pannick KC.

Section 2 of the Asylum and Immigration Appeals Act 1993 is headed “Primacy of the Convention” and reads:

“Nothing in the Immigration Rules (within the meaning of the 1971 Act) shall lay down any practice which would be contrary to the Convention”.

(“The Convention” is defined in section 1 as the Refugee Convention.)

In R (European Roma Rights Centre) v Immigration Officer at Prague Airport [2004] UKHL 55, [2005] 2 AC 1, Lord Steyn said, at para. 41 of his opinion:

“It is necessarily implicit in section 2 that no administrative practice or procedure may be adopted which would be contrary to the Convention. After all, it would be bizarre to provide that formal immigration rules must be consistent with the Convention but that informally adopted practices need not be consistent with the Convention. The reach of section 2 of the 1993 Act is therefore comprehensive.”

Although in the Secretary of State’s skeleton argument that statement is described as obiter it is not suggested, nor did Lord Pannick submit, that we should not follow it. It seems that the Divisional Court may have had some doubts about its correctness because at para. 122 of its judgment it observed that “section 2 of the 1993 Act is directed only to ensuring consistency between the Immigration Rules and the Refugee Convention” and questioned whether the “practice” challenged in these cases fell within the terms of the section because paragraphs 345A-345D said nothing about removal of asylum-seekers to Rwanda; but it went on to say that it would not dismiss this part of the case on that basis. Lord Steyn’s statement has been quoted with approval in at least two subsequent decisions of this Court – EN (Serbia) v Secretary of State for the Home Department [2009] EWCA Civ 630, [2010] QB 633, (per Stanley Burnton LJ at para. 58); and Secretary of State for the Home Department v RH [2020] EWCA Civ 1001 (per Baker LJ at para. 12). I proceed on the basis that it is correct.

It is the Claimants’ case that the removal of asylum-seekers to Rwanda on the basis of paragraphs 345A-345D of the Immigration Rules and the Inadmissibility Guidance constitutes a practice which is contrary to the Convention and accordingly unlawful under section 2. That case is formulated in four distinct grounds of appeal, namely grounds 9-12 in the consolidated grounds. These read:

“9.

The Court erred in failing to address the question of whether asylum-seekers removed to Rwanda would be accorded their rights under the Refugee Convention as a matter of vires as opposed to rationality.

10.

The Court erred in concluding at [126] that the MEDP Scheme as set out in paragraphs 345A-D of the Immigration Rules was consistent with the Refugee Convention and therefore not ultra vires s2 of the 1993 Act.

11.

The Court erred in finding that inadmissibility and/or removal to Rwanda did not constitute a penalty for the purposes of Article 31 of the Refugee Convention.

12.

The Court was wrong to find, for the purposes of Article 31 of the Refugee Convention, that the removal of RM, before his asylum claim would have been considered, to a third country with which he has no prior connection, with the avowed aim of deterring them or others from seeking asylum in the UK after arriving by unlawful means, did not constitute a penalty and therefore was consistent with s.2 of the Asylum and Immigration Appeals Act 1993.”

Grounds 9 and 11 were pleaded in AAA and were advanced by Mr Husain; grounds 10 and 12 were pleaded in ASM and RM respectively and were advanced by Mr Richard Drabble KC. As will appear, there is some overlap between the grounds, but I will take them as my structure in addressing this issue. I will consider ground 10 first, then grounds 11 and 12 together, and finally ground 9.

For the purpose of grounds 10-12 the relevant provisions of the Convention are articles 31 and 33, and it will be convenient to set them out now. Article 31 reads:

“Refugees unlawfully in the country of refuge

1.

The Contracting States shall not impose penalties, on account of their illegal entry or presence, on refugees who, coming directly from a territory where their life or freedom was threatened in the sense of article 1, enter or are present in their territory without authorization, provided they present themselves without delay to the authorities and show good cause for their illegal entry or presence.

2.

The Contracting States shall not apply to the movements of such refugees restrictions other than those which are necessary and such restrictions shall only be applied until their status in the country is regularized or they obtain admission into another country. The Contracting States shall allow such refugees a reasonable period and all the necessary facilities to obtain admission into another country.”

Article 33 reads:

“Prohibition of expulsion or return (‘refoulement’)

1.

No Contracting State shall expel or return (‘refouler’) a refugee in any manner whatsoever to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion.

2.

The benefit of the present provision may not, however, be claimed by a refugee whom there are reasonable grounds for regarding as a danger to the security of the country in which he is, or who, having been convicted by a final judgment of a particularly serious crime, constitutes a danger to the community of that country.”

This ground is pleaded in very general terms. However, Mr Drabble’s primary submission, as formulated at para. 58 of ASM’s skeleton argument, is that “there is an implied obligation on a receiving state, inherent in the basic structure of the Convention, to process a claim for asylum made by a refugee physically present in its territory”. It would follow that it was a breach of the Convention for a state to remove a person who has made an asylum claim to another country, however safe, without determining their claim and according them the rights of a refugee under the Convention if the claim is established.

The factors relied on in support of that primary submission are identified in the same paragraph of the skeleton argument as follows:

“That implied obligation arises as a combination of the declaratory nature of refugee status (which requires investigation of the individual’s circumstances and the nature of his claim) and Convention provisions, including the prohibition on refoulement (Article 33), the prohibition on penalties for illegal entry or presence (Article 31), the duty to afford refugees the same treatment as ‘aliens’ (Article 7), the prohibition on discrimination on grounds of country of origin and the right of access to courts (Article 16). Even Article 9, which permits provisional measures against an individual where ‘essential to national security’, does so only ‘pending a determination by the Contracting State that that person is in fact a refugee’.”

(It is unnecessary to set out the full terms of articles 7, 9 and 16 as there referred to: their effect is adequately summarised.)

In support of the submission that the obligation in question could be implied from the terms of the Convention the skeleton argument refers to the following provisions of article 31.1 of the Vienna Convention on the Law of Treaties 1969 (“General rule of interpretation”), which reads:

“1.

A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.”

2.

…

3.

There shall be taken into account, together with the context:

(a)

…;

(b)

any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation;

(c)

… .”

That submission is contrary to the practice which has prevailed in the European Union since the coming into force of the “Dublin system” in 1997 (initially in the form of the Dublin Convention, but since succeeded by the Dublin II and III Regulations), which permits member states to decline to entertain asylum applications from claimants who had previously applied to another member state. It is also contrary to articles 25-27 of EU Council Directive 2005/85/EC “on minimum standards on procedures in Member States for granting and withdrawing refugee status” (“the Procedures Directive”), which explicitly recognise the legitimacy of treating an asylum application as inadmissible, and refusing to consider it, if a country other than a member state “is considered as a safe third country for the applicant” (article 25.2 (c)). The “safe third country concept” is defined in article 27, which reads:

“1.

Member States may apply the safe third country concept only where the competent authorities are satisfied that a person seeking asylum will be treated in accordance with the following principles in the third country concerned:

(a)

life and liberty are not threatened on account of race, religion, nationality, membership of a particular social group or political opinion;

(b)

the principle of non-refoulement in accordance with the Geneva Convention is respected;

(c)

the prohibition of removal, in violation of the right to freedom from torture and cruel, inhuman or degrading treatment as laid down in international law, is respected; and

(d)

the possibility exists to request refugee status and, if found to be a refugee, to receive protection in accordance with the Geneva Convention.

2.

The application of the safe third country concept shall be subject to rules laid down in national legislation, including:

(a)

rules requiring a connection between the person seeking asylum and the third country concerned on the basis of which it would be reasonable for that person to go to that country;

(b)

rules on the methodology by which the competent authorities satisfy themselves that the safe third country concept may be applied to a particular country or to a particular applicant. Such methodology shall include case-by-case consideration of the safety of the country for a particular applicant and/or national designation of countries considered to be generally safe;

(c)

rules in accordance with international law, allowing an individual examination of whether the third country concerned is safe for a particular applicant which, as a minimum, shall permit the applicant to challenge the application of the safe third country concept on the grounds that he/she would be subjected to torture, cruel, inhuman or degrading treatment or punishment.

3.

When implementing a decision solely based on this Article, Member States shall:

(a)

inform the applicant accordingly; and

(b)

provide him/her with a document informing the authorities of the third country, in the language of that country, that the application has not been examined in substance.

4.

Where the third country does not permit the applicant for asylum to enter its territory, Member States shall ensure that access to a procedure is given in accordance with the basic principles and guarantees described in Chapter II.

5.

Member States shall inform the Commission periodically of the countries to which this concept is applied in accordance with the provisions of this Article.”

Both the Dublin system and the Procedures Directive are parts of the “Common European Asylum System”, and the Dublin system proceeds on the basis that member states are safe third countries.

In tacit recognition of that difficulty, the skeleton argument advances what appears to be an alternative to the primary submission. Para. 63 reads:

“The operation of the Dublin III scheme indicates that a constrained application of the ‘safe third country’ concept, with clear procedures and a robust protective legal framework, may be consistent with the Convention.”

As I understand it, that alternative proceeds on the basis that, although the Convention itself says nothing about contracting states being entitled to decline to entertain an asylum application where the applicant could go to a safe third country, such a right, if sufficiently “constrained”, might be capable of being implied in the light of its object and purpose and by reference to subsequent practice. It refers only to the Dublin system, and it does not explicitly extend to the safe third country concept as defined in the Procedures Directive, which goes beyond the Dublin regime in as much as it applies to non-EU states; but the reasoning behind the alternative would seem to apply in the case of removal to any safe third country.

Mr Drabble contended that that alternative reading of the Convention does not assist the Secretary of State because the MEDP is not sufficiently “constrained”. Para. 73 of ASM’s skeleton argument reads:

“The MEDP scheme represents a significant departure from the Dublin III framework and an extension of the ‘safe third country’ concept beyond that permitted by the Convention. Rwanda is not a country included in Parts 2-4 of Schedule 3 [to the Asylum and Immigration (Treatment of Claimants, etc.) Act 2004], and is therefore not subject to the statutory presumption of safety; there has been no Parliamentary scrutiny of the Secretary of State’s decision to treat Rwanda as a safe country. Rwanda is, obviously, not bound by EU standards nor the Common European Asylum System …; it is not a signatory of the [European Convention on Human Rights]; it is not subject to the jurisdiction of the [Court of Justice of the European Union] or the European Court of Human Rights. The MEDP scheme is underpinned not by statutory provisions, but by the Rwandan MOU and accompanying Notes Verbales …, paragraphs 345A-C of the Immigration Rules and relevant policy guidance.”

Mr Drabble advanced three further arguments under this ground:

The starting-point in considering those submissions is that it is in my view settled law that the Refugee Convention does not prohibit a receiving state from declining to entertain an asylum claim where it can and will remove the claimant to another non-persecutory state. That was clearly stated by Lord Bridge in R v Secretary of State for the Home Department, ex p Bugdaycay [1987] AC 514. That decision in fact covers four distinct cases. The relevant case for our purposes is Musisi, where the applicant was a Ugandan national who had arrived in the UK from Kenya and claimed that he was at risk of persecution in Uganda. The Home Secretary proposed to return him to Kenya. That decision was quashed in the High Court, whose decision was upheld by the Court of Appeal, on the basis that the Home Secretary had not considered his claim that he would be refouled from Kenya to Uganda. Although the House of Lords upheld the Court of Appeal’s decision, Lord Bridge began his analysis by saying (at p. 532 B-C):

“… I can well see that if a person arrives in the United Kingdom from country A claiming to be a refugee from country B, where country A is itself a party to the Convention, there can in the ordinary case be no obligation on the immigration authorities here to investigate the matter. If the person is refused leave to enter the United Kingdom, he will be returned to country A, whose responsibility it will be to investigate his claim to refugee status and, if it is established, to respect it. This is, I take it, in accordance with the ‘international practice’ [referred to in the evidence]. The practice must rest upon the assumption that all countries which adhere to the Convention may be trusted to respect their obligations under it. Upon that hypothesis, it is an obviously sensible practice and nothing I say is intended to question it.”

He goes on to say that where that assumption was shown to be unsafe return to country A would be contrary to the prohibition on refoulement (in that case indirect refoulement) in article 33; but what matters for our purposes is his statement of the position where there is no such risk.

Lord Bridge’s statement is consistent with the academic commentary. The discussion of article 31 in Professor James Hathaway’s The Rights of Refugees under International Law (2^nd ed, 2021) is particularly useful because it considers the travaux préparatoires. He says, at p. 519, that “Art. 31 in no way constrains a state’s prerogative to expel an unauthorized refugee from its territory”. He observes that it might seem ironic that an asylum country which is generally prohibited from imposing penalties on refugees may nonetheless expel them; but he goes on at pp. 519-520 to demonstrate, by reference to the statements made on behalf of the original signatories in the course of the Conference which led to the Convention, that that was indeed the intention. The position was most clearly stated by the argument of the Canadian representative that no modification of the text as regards this issue was required, since “the consensus of opinion was that the right [to expel refugees who illegally enter a state’s territory] would not be prejudiced by adoption of Article [31]”. He goes on to point out that what he calls the “potentially devastating impact” of that decision is mitigated by the duty of non-refoulement in article 33: “any expulsion of a refugee must therefore not expose the refugee, directly or indirectly, to a risk of being persecuted”.

To the same effect, in his analysis and commentary on the Travaux Préparatoires to the Refugee Convention Dr Paul Weis says at pp. 302-303:

“Article 31 refers to ‘penalties’. It is clear from the travaux préparatoires that this refers to administrative or judicial convictions on account of illegal entry or presence, not to expulsion ...

Paragraph 1 does not impose an obligation to regularise the situation of the refugee nor does it prevent the Contracting States from imposing an expulsion order on him. However, a refugee may not be expelled if no other country is willing to admit him …”

See also Goodwin-Gill and McAdam, The Refugee in International Law (4^th ed, 2021), at p. 278.

Against that background, I see no room for the kinds of implied obligation contended for by Mr Drabble. Specifically, there is no warrant for implying a prohibition on removal except where the third country satisfies the particular requirements either of the Dublin system or of article 27 of the Procedures Directive (including the requirement that the application have a connection with that country). The straightforward question, so far as the Convention is concerned, is whether the third country is safe for the applicant in the sense that there is no real risk of their being refouled (directly or indirectly). Nor can any limitations be implied on the state’s right to take into account, when deciding whether to remove an asylum-seeker to a safe third country, the fact that they arrived in the UK irregularly, even in circumstances where there was no regular means to do so, or that their removal may have a deterrent effect. The state’s motivation is irrelevant to the object and purpose of the Convention: 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. (This conclusion is subject to the ostensibly distinct “penalty” issue discussed under grounds 11 and 12.)

My conclusion is reinforced by the observations of Lord Bingham at para. 18 of his opinion in the Roma Rights case. The case involved, in part, the interpretation of the Refugee Convention. Lord Bingham accepted counsel’s submission that the Convention “should be given a generous and purposive interpretation, bearing in mind its humanitarian objects and purpose”, but he continued:

“But I would make an important caveat. However generous and purposive its approach to interpretation, the court’s task remains one of interpreting the written document to which the contracting states have committed themselves. It must interpret what they have agreed. It has no warrant to give effect to what they might, or in an ideal world would, have agreed. This would violate the rule, … expressed in article 31(1) of the Vienna Convention, that a treaty should be interpreted in accordance with the ordinary meaning to be given to the terms of the treaty in their context. It is also noteworthy that article 31(4) of the Vienna Convention requires a special meaning to be given to a term if it is established that the parties so intended. … It is in principle possible for a court to imply terms even into an international convention. But this calls for great circumspection since, as was said in Brown v Stott [2003] 1 AC 681, 703,

‘it is generally to be assumed that the parties have included the terms which they wished to include and on which they were able to agree, omitting other terms which they did not wish to include or on which they were not able to agree,’

and caution is needed

‘if the risk is to be averted that the contracting parties may, by judicial interpretation, become bound by obligations which they did not expressly accept and might not have been willing to accept’.”

The pleaded challenge is to para. 126 of the Divisional Court’s judgment. However, that merely expresses its overall conclusions on the issue of compatibility with the Refugee Convention. Its essential reasoning as regards this aspect appears in para. 121 of its judgment, where it said:

“Mr Drabble KC submitted that the Refugee Convention imposes an obligation on contracting states to determine all asylum claims made, on their merits. We disagree. There is no such obligation on the face of the Convention. The obligation that is imposed is the one at article 33, not to expel or return a refugee to a place where his life or freedom would be threatened by reason of any of the characteristics that the convention protects. Mr Drabble’s submission was that an obligation to determine asylum claims would be consistent with the spirit and purpose of the Convention and could therefore reasonably be assumed. Again, we disagree. Obligations in international treaties are formulated with considerable care. They reflect balances struck following detailed negotiations between states parties. An obligation to determine every asylum claim on its merits would be a significant addition to the Refugee Convention. There is no reason to infer the existence of an obligation of that order; to do so would go well beyond the limits of any notion of judicial construction of an international agreement; and the protection that is necessary if the purpose of the Convention is to be met, is provided by article 33.”

I agree.

In both these grounds the contention is that the removal of asylum-seekers to Rwanda under the MEDP constitutes a “penalty” within the meaning of article 31 (1). They differ from ground 10 because they depend on the meaning of the word “penalty” and thus do not depend on the implication of any term. However the underlying issue is in fact in my view the same, and I can take them fairly shortly.

It follows from my reasoning in relation to ground 10 that the removal of an asylum-seeker to a safe third country without their claim being determined is not in itself a penalty: indeed the passages from the academic commentators which I have quoted are all in the context of article 31 (1). However, Mr Husain and Mr Drabble submitted that expulsion may become a penalty within the meaning of the article depending on the facts of a particular case. I summarise their submissions in turn.

Mr Husain contended (see para. 64 (ii) of the AAA Claimants’ skeleton argument) that “the reasons why, and the conditions to which, a person is being removed are highly relevant and may convert a removal which is lawful per se into an impermissible penalty”. As for the “reasons” element in that formulation, he relied on the fact that the asylum-seekers who were liable to relocation to Rwanda under the MEDP were being removed “for the purposes of imposing a detriment on them and deterring others from arriving in the same way”. He relies on an observation in a commentary on the term “criminal offence” in the International Covenant on Civil and Political Rights, adopted by Professor Guy Goodwin-Gill in a paper on article 31 commissioned by UNHCR, to the effect that “every sanction that has not only a preventative but also a retributive and/or deterrent character is ... to be termed a penalty”. As for the “conditions to which [the Claimants would be] removed”, he relied on the fact that they would be being removed to “significantly inferior processes and human rights protection”.

Mr Drabble submitted that the term “penalty” required a broad and purposive construction in line with the humanitarian purpose of the Convention: as we have seen, that was accepted by Lord Bingham in the Roma Rights case. Adopting that approach, he submitted that expulsion could constitute a penalty if it was detrimental to the applicant in some specific way such as separation from family members or a supportive community. He also relied on the deterrent purpose of the MEDP.

Both counsel relied on the decision of the Supreme Court of Canada in B010 v Canada [2015] 3 SCR 704. The case concerned a Canadian statutory provision – section 37 (1) of the Immigration and Refugee Protection Act 2001 – which rendered a person “inadmissible”, which effectively denied them access to refugee determination procedures, if they had engaged in “in the context of transnational crime, activities such as people smuggling”. The issue was whether on its true construction that provision applied to illegal migrants who had assisted other illegal migrants but had not done so in return for any financial or other benefit. The Court held that it did not. As part of her reasoning in support of that construction McLachlin CJ, delivering the judgment of the Court, held that “omitting a financial or other benefit limitation” would be inconsistent with article 31 (1) of the Convention: see para. 62 of her judgment. At para. 57 she sets out article 31 (1) and adopts a statement in a textbook that:

“an individual cannot be denied refugee status – or, most important, the opportunity to make a claim for such status through fair assessment procedures – solely because of the way in which that person sought or secured entry into the country of destination”

and that

“[o]bstructed or delayed access to the refugee process is a ‘penalty’ within the meaning of art. 31(1) …”.

At para. 63 she says:

“The respondents contend that art. 31(1) of the Refugee Convention refers only to criminal penalties. This interpretation runs counter to the purpose of art. 31(1) and the weight of academic commentary: J. C. Hathaway, The Rights of Refugees Under International Law (2005), at pp. 409-12; Gallagher and David, at pp. 164-68; G. S. Goodwin-Gill and J. McAdam, The Refugee in International Law (3rd ed. 2007), at p. 266. The generally accepted view is that denying a person access to the refugee claim process on account of his illegal entry, or for aiding others to enter illegally in their collective flight to safety, is a ‘penalty’ within the meaning of art. 31(1). The law recognizes the reality that refugees often flee in groups and work together to enter a country illegally. Article 31(1) thus does not permit a state to deny refugee protection (or refugee determination procedures) to refugees solely because they have aided others to enter illegally in an unremunerated, collective flight to safety. Rather, it targets those who assist in obtaining illegal entry for financial or other material benefit.”

Mr Husain relied on the words that I have italicised and submitted that they were directly applicable to the present case. He acknowledged that the Court in B10 was not concerned with a situation where the migrant would be being removed to a safe third country, but he submitted that that was immaterial. I do not agree. In my view it is a crucial distinction. The views endorsed by McLachlin CJ in paras. 57 and 63 of her judgment are concerned with denial of, or obstructions or delay to, access to the refugee determination process for a migrant who is in the country. They are not concerned with expulsion, as to which, as I have sought to show in connection with ground 10, the Convention imposes no restrictions save for the duty of non-refoulement imposed by article 33.

The same point applies to Professor Goodwin-Gill’s wide definition of “penalty” (see para. 324 above). I have no difficulty with the proposition that the term is not confined to sanctions of a criminal character; but the issue here is whether it extends to expulsion.

In short, it is in my view inconsistent with the well-recognised scheme of the Convention that the expulsion of a migrant to a safe third country should be treated as a penalty within the meaning of article 31 (1), whatever the reasons for taking that course may be and however unwelcome it may be to the migrant in question.

The Divisional Court addressed the effect of article 31 (1) at paras. 123-125 of its judgment. At para. 125 it says:

“There is, therefore, a clear consensus. Article 31 does not prevent a state expelling a refugee.  States must not act in breach of article 33; removal that is not contrary to article 33 is not a penalty for the purposes of article 31.  On this basis, neither decisions on inadmissibility under paragraph 345A of the Immigration Rules, nor decisions under paragraph 345C on removal to Rwanda are contrary to the Refugee Convention. The latter because one premise of a paragraph 345C decision is that the country concerned is a safe third country, as defined at paragraph 345B of the Immigration Rules. The deterrent purpose that the Home Secretary pursues in relation to removals to Rwanda does not, of itself, render removal to Rwanda contrary to article 31, let alone article 33 of the Refugee Convention. Further, the simple fact of removal to Rwanda is not sufficient to make good the Claimants’ submission that removal is a penalty contrary to article 31. That submission would succeed only when removal amounts to a breach of article 33. Looked at on this basis, the Claimants’ article 31 submission merges with their submission on whether Rwanda is a safe third country.  If it is a safe third country, decisions taken in exercise of the powers in paragraphs 345A - 345D of the Immigration Rules are not in breach of article 31; if, however, Rwanda is not a safe third country, removal would be both contrary to paragraph 345C of the Immigration Rules and to both article 31 and article 33 of the Refugee Convention.”

Again, I agree.

As pleaded this ground is decidedly opaque. However, as developed at paras. 53-58 of the AAA Claimants’ skeleton argument it appears to comprise four points.

First, it is contended that the Divisional Court’s error in approaching the question of whether Rwanda was a safe third country on a review basis rather than reaching its own conclusion infected its reasoning also on the Refugee Convention issues. I do not agree. Its dispositive reasoning, which I have set out at paras. 321 and 330 above, does not depend on its assessment of the safety of Rwanda. Even if it did the point would go nowhere since I believe that its conclusion on the Refugee Convention issues which it considered was correct in any event.

The second and third points appear to go together. The Claimants had argued in the Divisional Court that Rwandan law was incompatible with two articles of the Refugee Convention, as follows:

The Claimants complain that the Divisional Court failed to address those points. I accept that it did not expressly consider them, though the omission is venial in view of the plethora of arguments with which it was faced. But I do not believe that the points are good in any event. The material relied on falls well short of establishing that Rwandan law fails to give effect to either article 1C or article 1F. As we have seen, the MoU contains an express provision that RIs will be treated in accordance with the Refugee Convention (see para. 9.1.1). We do not have the text of either of the domestic instruments relied on, still less any expert evidence. Even if they are accurately reproduced, we do not know whether under Rwandan law the requirements of the Convention would trump any contrary provision of the domestic legislation (or in any event any ministerial instruction). But even if that is not the case, it would be surprising if a Court seeking to construe domestic law in accordance with Rwanda’s obligations under the Convention were unable to do so. It is for the Claimants to establish that the Rwanda policy contravenes section 2, and I do not believe that (in this respect) they have done so.

The fourth point is based on article 15 of the Convention (“Right of Association”), which obliges contracting states, as regards “non-political and non-profit-making associations and trade unions”, to “accord to refugees lawfully staying in their territory the most favourable treatment accorded to nationals of a foreign country, in the same circumstances”. (Article 7 (1) contains a more general non-discrimination provision, but that adds nothing to the argument.) At paras. 386-389 of their skeleton argument in the Divisional Court the AAA Claimants had relied on various materials in support of a submission that article 15 would not be complied with as regards RIs. Some of the materials simply showed that there were serious restrictions on freedom of speech and freedom of assembly in Rwanda, and at para. 77 of its judgment the Divisional Court held that that did not demonstrate any discrimination against refugees as opposed to other foreign nationals, which is the subject of article 15. However, the Claimants complain that that rebuttal fails to address one of the materials relied on, which was para. 3.8.3 of the Human Rights CPIN, which reads:

“In 2016, the Rwandan Government’s Ministry in Charge of Emergency Management (MINEMA), published Ministerial instructions determining the management of refugees and refugee camps. Article 2 refer to ‘Prohibited acts and behaviors for refugees’ and states that ‘Political activities’ and ‘Gatherings based on ethnicity, nationality, or any other sectarian ground’ and participating in, or inciting others into unlawful riots are prohibited.”

(The Ministerial Instruction seems to be the same as that referred to at para. 333 (1) above.) That Instruction does appear to impose specific restrictions on refugees. The Claimants complain that the Divisional Court failed to deal with it and submit that it shows that the article 15 rights of RIs would not be respected. In her submissions resisting the grant of permission to appeal the Secretary of State drew attention to article 12 of the Ministerial Instruction, which provides for refugees to enjoy (in effect) the full rights accorded by the Convention, including “membership to association of forums with non-political orientation”; but the Claimants contend that that is in very general terms, and article 2 is plainly a derogation from it.

I would reject the Claimants’ submission. I do not believe that the terms of article 2 of the Ministerial Instruction establish that the article 15 rights of RIs would not be recognised. Despite the distinction between “political activities” and “gatherings based on ethnicity, nationality, or any other sectarian ground”, I am not satisfied that the latter class of activities would qualify as “non-political” under article 15: on a purposive construction it seems very unlikely that a purely social, or other non-political, meeting of, say, Kurdish or Albanian refugees would be held to be caught by article 2 because it was a “gathering based on ethnicity [or] nationality”. It is also far from clear that the same restrictions would not apply to other non-nationals. In the absence of any authoritative expert evidence on these points I do not consider that the Claimants’ submission is made out.

I should add that other arguments were relied on by the Secretary of State, and the Divisional Court, in response to the Claimants’ points on Rwandan law; but the foregoing is sufficient for me to reject them.

For the reasons given, I would reject grounds of appeal 9-12.

It may in fact be arguable that the effect of my earlier conclusion that RIs would be at risk of refoulement from Rwanda because of the inadequacy of its asylum system is that relocation there would constitute a practice which contravened article 33 and would thus be unlawful under section 2 on a different basis; but the case was not put that way and I need not consider it here.

This issue concerns whether paragraph 345C of the Immigration Rules is compatible with retained EU law. Paragraphs 345A-345D of the Rules were introduced by Statement of Changes HC 1043, dated 10 December 2020, to take effect at 23.00 on 31 December 2020, i.e. at the moment of “IP Completion” as defined in the European Union (Withdrawal Agreement) Act 2020. For convenience I will set out paragraph 345C again here:

“When an application is treated as inadmissible, the Secretary of State will attempt to remove the applicant to the safe third country in which they were previously present or to which they have a connection, or to any other safe third country which may agree to their entry [my italics].”

The point to note for our purposes is that the italicised words explicitly contemplate the removal of an asylum-seeker to a safe third country with which they have no connection. (Footnote: ⁸)

The retained EU law which paragraph 345C is said to contravene is the Procedures Directive, to which I have already referred in connection with Issue 12. The Directive came into force in the UK, by virtue of section 2 of the European Communities Act 1972, on 2 January 2006. (I should also mention, because it is part of the wider picture, that a separate Directive relating to aspects of substantive asylum law – the “Qualification Directive” (2004/83/EU) – came into force shortly afterwards.)

The Procedures Directive imposes a number of requirements on member states as regards the procedure for determining applications for asylum, including (by article 6.2) a requirement that all adults with legal capacity should have the right to make an application for asylum and (by article 8) that any such application should be appropriately examined. So far as relevant for our purposes, article 25 allows member states to treat an asylum application as “inadmissible”, and accordingly not to examine it, in a number of particular situations identified in paragraph 2. These include, at (c), where “a country which is not a Member State is considered as a safe third country for the applicant, pursuant to Article 27”. I have already set out article 27 in full at para. 312 above. For present purposes I need only note that paragraph 2 (a) requires member states to implement “rules requiring a connection between the person seeking asylum and the third country concerned on the basis of which it would be reasonable for that person to go to that country”: I will refer to that as “the connection requirement”.

It is the Claimants’ case that the Procedures Directive remains part of UK law by reason of section 4 of the European Union (Withdrawal) Act 2018, which preserves the force, as “retained EU law”, of EU legislation given effect by the 1972 Act. On that basis, they submit, paragraph 345C of the Immigration Rules is unlawful, because, contrary to article 27.2 (a), it permits removal to a country with which the applicant has no connection.

The Secretary of State does not challenge the proposition that paragraph 345C is inconsistent with the requirements of article 27.2 (a) of the Directive. However it is her case that those requirements no longer form part of UK law. She relies on the effect of the Immigration and Social Security Co-ordination (EU Withdrawal) Act 2020 (“ISSCA”), which received the Royal Assent on 11 November 2020. I need to set out the structure of the Act and the relevant provisions in some detail.

The long title of ISSCA is as follows:

“An Act to make provision to end rights to free movement of persons under retained EU law and to repeal other retained EU law relating to immigration; to confer power to modify retained direct EU legislation relating to social security co-ordination; and for connected purposes.”

As foreshadowed by the long title, the Act has two substantive Parts. Part 1 is headed “Measures relating to ending free movement”, and Part 2 “Social security co-ordination”. We are concerned with Part 1, which came into force on 31 December 2020, i.e. coinciding with IP Completion Day.

Part 1 of ISSCA comprises sections 1-5. The principal substantive section is section 1, which gives effect to Schedule 1 to the Act. It reads:

“Repeal of the main retained EU law relating to free movement etc.

Schedule 1 makes provision to—

(a)

end rights to free movement of persons under retained EU law, including by repealing the main provisions of retained EU law relating to free movement, and

(b)

end other EU-derived rights, and repeal other retained EU law, relating to immigration.”

The other substantive sections are sections 2-3. Both address particular issues related to the ending of free movement. In summary:

Schedule 1, which as we have seen is given effect by section 1, is headed “Repeal of the main retained EU law relating to free movement etc.”. It comprises three Parts. Part 1 is headed “EU-derived domestic legislation” and revokes certain specified provisions of primary legislation and statutory instruments. Part 2 is headed “Retained Direct EU Legislation” and repeals the Workers Regulation (Regulation (EU) No 492/2011), which is the EU provision conferring the right of freedom of movement for workers within the EU.

We are concerned with Part 3 of the Schedule, which is headed “EU-derived rights etc”: it is thus concerned with rights other than those conferred directly by domestic or EU legislation, which are the subjects of Parts 1 and 2. It comprises paragraphs 5 and 6. Paragraph 5 is concerned with rights derived from a free movement agreement made between the European Union and the Swiss Confederation. Paragraph 6 reads (so far as relevant):

“(1)

Any other EU-derived rights, powers, liabilities, obligations, restrictions, remedies and procedures cease to be recognised and available in domestic law so far as—

(a)

they are inconsistent with, or are otherwise capable of affecting the interpretation, application or operation of, any provision made by or under the Immigration Acts (including, and as amended by, this Act), or

(b)

they are otherwise capable of affecting the exercise of functions in connection with immigration.

(2)

The reference in sub-paragraph (1) to any other EU-derived rights, powers, liabilities, obligations, restrictions, remedies and procedures is a reference to any rights, powers, liabilities, obligations, restrictions, remedies and procedures which—

(a)

continue to be recognised and available in domestic law by virtue of section 4 of the European Union (Withdrawal) Act 2018 (including as they are modified by domestic law from time to time), and

(b)

are not those described in paragraph 5 of this Schedule.

(3)

...”

(“EU-derived” is not a defined term, but it is obvious, and not disputed, that as a matter of language it would apply to the obligations imposed on the UK Government by the Procedures Directive.)

The effect of paragraph 6 (1) of Schedule 1 to ISSCA (to which I will refer simply as “paragraph 6 (1)”) is thus to disapply EU-derived rights to the extent that they are inconsistent with any provision made by or under “the Immigration Acts”. That term is defined by section 61 (2) of the UK Borders Act 2007, which provides that “[a] reference (in any enactment, including one passed or made before this Act) to ‘the Immigration Acts’” is to a specified list of statutes. For our purposes what matters is that those statutes include the Immigration Act 1971, but for reasons which will appear I should give the rest of the list:

“(a)

the Immigration Act 1971,

(b)

the Immigration Act 1988,

(c)

the Asylum and Immigration Appeals Act 1993,

(d)

the Asylum and Immigration Act 1996,

(e)

the Immigration and Asylum Act 1999,

(f)

the Nationality, Immigration and Asylum Act 2002,

(g)

the Asylum and Immigration (Treatment of Claimants, etc.) Act 2004,

(h)

the Immigration, Asylum and Nationality Act 2006,

(i)

this Act,

(j)

the Immigration Act 2014,

(k)

the Immigration Act 2016,

(l)

Part 1 of the Immigration and Social Security Co-ordination (EU Withdrawal) Act 2020 (and Part 3 so far as relating to that Part), and

(m)

the Nationality and Borders Act 2022.”

(Item (l) was added by section 4 (1) of ISSCA.)

Section 3 (2) of the Immigration Act 1971 provides, so far as relevant, that:

“The Secretary of State shall from time to time … lay before Parliament statements of the rules, or of any changes in the rules, laid down by him as to the practice to be followed in the administration of this Act for regulating the entry into and stay in the United Kingdom of persons required by this Act to have leave to enter.”

The Immigration Rules are made pursuant to that provision, and although their precise legal status is notoriously anomalous, it is not in issue before us that for the purposes of paragraph 6 (1) they are to be regarded as made “under” the 1971 Act. Part 11 of the Rules regulates applications for asylum: paragraphs 345A-345D fall under that Part.

It is the Secretary of State’s case that paragraph 27.2 (a) of the Directive is “inconsistent with, or … otherwise capable of affecting the … operation of” paragraphs 345A-345D of the Immigration Rules because it imposes a connection requirement which they do not; and that accordingly, by virtue of paragraph 6 (1), it ceases to that extent to be recognised and available in domestic law”. It follows that those paragraphs cannot be impugned on the basis that they fail to give effect to the connection requirement.

The Claimants’ answer to the Secretary of State’s case was advanced by Mr Drabble. He submitted that if paragraph 6 (1) is read in the context of the Act as a whole it is clear that the reference to provisions of “the Immigration Acts” is intended to be only to those provisions which are concerned with immigration as opposed to asylum, which I will call “immigration in the narrower sense”: immigration and asylum are distinct legal subject-matters, with very different origins and characters. He pointed out that the primary purpose of the Act as defined in the long title (ignoring its social security co-ordination aspect) is to end free movement, and that that is all that the heading to Part 1 refers to: free movement is wholly concerned with immigration and has nothing to do with asylum. He accepted that the long title and section 1 refer not only to ending rights to free movement but also to repealing “other retained EU law relating to immigration”; and that that additional element is reflected in the phrase “free movement etc” in the headings to section 1 and Schedule 1. But he submitted that the phrase “relating to immigration” takes its colour from the primary purpose and can only be read as referring to immigration in the narrower sense; and that the same goes for the “etc”. That being so, the reference to provisions of “the Immigration Acts” must be similarly limited so as to apply only to provisions relating to immigration as opposed to asylum.

Mr Drabble sought to reinforce that submission by referring to the Immigration, Nationality and Asylum (EU Exit) Regulations 2019, which were made under powers conferred by section 8 of the 2018 Act to correct “deficiencies in retained EU law” and which revoked a number of community instruments. He relied in particular on the fact that Schedule 1, which contains the relevant revocations, is divided into two Parts – Part 1 being headed “Revocations related to immigration and nationality” and Part 2 “Revocations related to asylum” (the latter including the Dublin Convention and the Dublin III Regulation). He submitted that that dichotomy illustrated how “immigration” and “asylum” were treated as distinct subject areas in the context of Brexit-related legislation, itself reflecting the fact that they are regarded as distinct concepts in EU law which are subject to entirely separate legal frameworks.

To the same effect, but more generally, Mr Drabble referred to the list of statutes in section 62 (1) of the 2007 Act and pointed out that the long titles, of which he handed in a schedule, are drafted on the basis that “immigration” and “asylum” are distinct: that is, if the statute contains provisions relating to asylum as well as immigration the long title always refers to both. For example, the long title of the Immigration Act 2016 begins “An Act to make provision about the law on immigration and asylum”.

Mr Drabble also referred to the Explanatory Notes to ISSCA. It is well established that Explanatory Notes are admissible aids to construction insofar as they “cast light on the objective setting or contextual scene of the statute, and the mischief at which it is aimed”: see Westminster City Council v National Asylum Support Service [2002] UKHL 38, [2002] 1 WLR 2956, per Lord Steyn at para. 5. Para. 1 of the Notes give a very full “Overview of the Act”. It focuses wholly (leaving aside the social security co-ordination aspect) on the need to end free movement between the EU and the UK and makes no mention of asylum. Paras. 70-71 of the Commentary on the particular provisions of the Act read as follows:

“70.

Paragraph 6 ensures any directly effective rights that will have been saved by the EUWA 2018 and would, in the absence of this paragraph, be retained, cease to apply insofar as they are inconsistent with, or are otherwise capable of affecting the interpretation, application or operation of, specified domestic immigration legislation or functions. For example, the residence rights that are derived from Articles 20 and 21 of the TFEU (rights of citizenship and free movement) will be retained EU law and, unless they are disapplied would provide a right to reside in the UK for certain groups, for example ‘Chen’ carers who are primary carers of an EU citizen child who is in the UK and is self-sufficient. However, the rights derived from Articles 20 and 21 would continue to apply in non-immigration contexts unless disapplied.

71.

The following is a non-exhaustive list of the directly effective rights relevant to this Paragraph. …”.

The list is in the form of a table which identifies seven treaties (broadly, the various EU and EEA/Switzerland treaties and the Association Agreements with Turkey) and some thirty particular provisions of those treaties. The subject area of each provision is identified. Several of the subject areas are identified in terms as “free movement”, either of workers or of services, but the remainder are also concerned with aspects of the free movement rights (such as discrimination on grounds of nationality or rights of family members of those exercising such rights). The Procedures Directive does not appear on the list and none of the subject matters has anything to do with asylum.

Mr Drabble submitted that the passages from the Explanatory Notes quoted above confirm, what was in any event clear from the language of the Act itself, that Part 1 of ISSCA is concerned only with “immigration” in the narrower sense and was not intended to have any effect on asylum law. He made what was substantially the same point by reference to various other Parliamentary materials relating to what became ISSCA, none of which refer to any impact on asylum law. I need not refer to them in detail, but they are the report of the House of Lords Delegated Powers and Regulatory Reform Committee (HL Paper 118, 25.8.20); the Delegated Powers Memorandum dated 24 July 2020, supplied to the Committee by the Home Office; the Government response to the Committee’s report (HL Paper 141, 14.10.20 – see Annex 1); and the report of the House of Lords Select Committee on the Constitution (HL Paper 120, 2.9.20).

Finally, Mr Drabble referred us to two recent decisions of the Supreme Court, Robinson v Secretary of State for the Home Department [2020] UKSC 53, [2022] AC 659, and G v G [2021] UKSC 9, [2022] AC 544. In fact it is the latter which is the more directly relevant for our purposes, and I take it first. A father resident in South Africa had brought proceedings under the Hague Convention for the return of his child who had been brought by the mother to England. The mother had claimed asylum for herself and her child. The High Court had imposed a stay on the Convention proceedings pending the determination of the asylum claim. The issue in the Supreme Court was whether it had been right to do so. The Home Secretary was an intervener in the appeal. The Court handed down its decision on 19 March 2021. The only judgment was given by Lord Stephens. For the purpose of his review of “the legal landscape governing asylum applications” he said, at paras. 83-84:

“83.

In so far as applicable to the United Kingdom the principal EU measures are (i) the Qualification Directive and (ii) the Procedures Directive (together, ‘the Directives’).

84.

The Secretary of State accepts, for the purposes of this appeal, and I agree, that the relevant provisions of the Directives are directly effective and remain extant in domestic law as ‘retained EU law’ after the United Kingdom’s withdrawal from the EU.”

He went on to summarise a number of the provisions of both Directives. Mr Drabble did not contend that Lord Stephens’ statement that the Directives remained law following the UK’s withdrawal from the EU was binding on us: it was debatable whether it formed part of the ratio but even if it did the point had not been the subject of any argument. He submitted that it was nevertheless weighty support for his proposition that not only the Court but the Secretary of State understood the position in the same way. He pointed out that although the Court was not addressed about the effect of ISSCA Lord Stephens was certainly aware of it, since in Robinson, which concerned Zambrano rights and in which he had delivered a judgment in December 2020, he had referred to the possibility that the law as declared in that case might change following IP Completion Day as a result of ISSCA, which he described as providing for “repeal of the main retained EU law relating to free movement” (see paras. 29-30).

The starting-point must be that on a straightforward reading of the statutory language paragraphs 345A-345D of the Immigration Rules are “provisions made … under the Immigration Acts”, with the consequence that paragraph 6 (1) has effect to disapply any inconsistent provision of retained EU law.

I accept, however, that it is necessary to construe the statute purposively, which means having regard not simply to the literal meaning of the words in question but to the legislative purpose, to be gleaned from the entirety of the provisions of the Act (in practice Part 1) and any admissible contextual materials. Both Mr Drabble and Lord Pannick referred us to paras. 29-31 of the judgment of Lord Hodge in R (O) v Secretary of State for the Home Department [2022] UKSC 3, [2023] AC 255 (the so-called “BritCits” case). I need not set the passage out in extenso, but the essential point is that the primary source for ascertaining the intention of Parliament must be the words used read in the context of the statute as a whole (para. 29). External aids to interpretation, such as Explanatory Notes, can only play a secondary role (para. 30).

Accordingly I start with what can be learnt about the purpose of Part 1 from the statute itself. It is clear from the long title, as well the language and structure of its provisions, that the main purpose is to end the right to free movement and to repeal legislation related to it; and I accept that that is a purpose that has nothing to do with asylum. However that is not the only purpose. Section 1 (1) (b) says in terms that Schedule 1 makes provision to “end other EU-derived rights, and repeal other retained EU law, relating to immigration”; “other” must mean other than related to free movement.

The question then is whether the phrase “relating to immigration” defines that secondary purpose in a way which excludes EU-derived rights relating to asylum. I do not believe that this is clear. Mr Drabble is right that in some legislative contexts “immigration” and “asylum” are treated as distinct subject-matters. But I am not persuaded that there is any settled practice. There are instances in the legislative context of the term “immigration” being used to cover both immigration, in the narrower sense, and asylum. One example is the Immigration Act 2016: although Mr Drabble relies on the fact that the long title refers to both asylum and immigration, it is also significant that the short title does not. Equally, the “Immigration Rules” include the entirety of the UK’s rules governing the determination of asylum claims (see Part 11) and are of course made under the Immigration Act 1971. (It may also be pertinent to note, though strictly it is outside the legislative context, that the leading textbook, Macdonald’s Immigration Law and Practice, treats asylum law as an integral part of its subject.) That broader use is not loose or illogical: the effect of a grant of asylum is to confer on the beneficiary the right to remain in the UK notwithstanding that they do not have British nationality, or other right of abode, and the relevant law can perfectly naturally be regarded as an aspect of immigration law.

If the use of the term “immigration” itself does not clearly connote an “asylum-exclusive purpose”, I see nothing else in the statute that does so. The fact that the primary purpose of Part 1 is to end free movement is in my view neutral: it does not follow that the “other” EU-derived rights relating to immigration law need be in some sense akin to freedom of movement rights or otherwise relate to immigration in the narrower sense. The only overall purpose of Part 1 that can be discerned from the Act itself is to remove or qualify EU-derived rights which are inconsistent with domestic immigration legislation, including the Immigration Rules: there is no a priori reason why the rights in question should not relate to asylum.

The upshot of that discussion is that there is nothing in the other provisions of the Act to suggest a legislative purpose that would justify departing from the literal language of paragraph 6 (1).

I turn to the extraneous materials relied on by Mr Drabble. I accept that the Parliamentary materials referred to strongly suggest that the Government, as the promoter of what became ISSCA, did not at the time that the bill was going through Parliament have any specific intention that the EU-derived rights affected by paragraph 6 (1) would include rights of asylum-seekers: if it did, it is hard to think that the responsible Home Office officials would not have referred to it in the Explanatory Notes and in the Government’s response to the relevant committee reports in the House of Lords. That is consistent with the Secretary of State’s stance in G v G (though it should be noted that her concession referred only to “the relevant provisions” of the Directives).

That has given me some pause. But it does not follow from the fact that the promoters of a statute have not foreseen all the particular consequences of its provisions that such consequences must be treated as falling outside its purpose so as to justify departing from their otherwise clear meaning. What matters in this case is not whether the Government specifically intended that “asylum rights” should fall within the scope of paragraph 6 (1) but whether it is clear that it specifically intended that they should fall outside its scope.

I do not believe that that is established. That is not simply because of the secondary role that extraneous materials of this kind play in the construction exercise. It is also because it is not clear that there was any reason at the time that the bill was going through Parliament for the Government to have considered the question of its impact on asylum rights one way or the other. When the Qualification and Procedure Directives first came into force Part 11 of the Immigration Rules was re-drafted so as to give effect to their provisions. There was accordingly no general reason to suppose that there was any inconsistency on which paragraph 6 (1) would operate: the Directives would simply remain in effect as retained EU law following IP Completion Day, as the Secretary of State accepted in G v G. It is true that in the particular case of paragraph 345C such an inconsistency has emerged, but paragraphs 345A-345D were only added to the Rules on 10 December 2020, after ISSCA had received Royal Assent. We do not know for how long it had been intended to introduce a rule which allowed removal to a safe third country with which the applicant had no connection; or whether it had been appreciated that such a rule would be inconsistent with article 27.2 (a); or, if so, whether it was intended that paragraph 6 (1) would or might provide the answer. But what matters for present purposes is that it is not clear that at any relevant time the Government anticipated an inconsistency between the Immigration Rules and retained EU law. That being so, it is not possible to draw any inference about the intended scope of paragraph 6 (1) from what was or was not said in the Explanatory Notes or other Parliamentary materials. It is necessary simply to apply the statutory language, which, as I have said, on its natural meaning applies to any provision made under the Immigration Rules.

I would accordingly uphold the Divisional Court’s conclusion on this issue. I have not found it necessary to refer to its reasoning, which is at paras. 106-118 of its judgment; but I do not detect any substantial difference between it and my own.

Section 33 (1) of the Asylum and Immigration (Treatment of Claimants etc) Act 2004 reads:

“Schedule 3 (which concerns the removal of persons claiming asylum to countries known to protect refugees and to respect human rights) shall have effect.”

Schedule 3, which is headed “Removal of Asylum Seeker to Safe Country”, is in five Parts, but Part 1 is introductory and we are concerned with the relationship between Parts 2-4 on the one hand and Part 5 on the other.

Parts 2-4 provide for, respectively, three “Lists of Safe Countries”. Part 2 sets out a list of countries which are regarded as safe for the purpose of both the Refugee Convention and the ECHR: they are all EU or EEA states. Parts 3 and 4 make provision for the Secretary of State to list, by order in the form of a statutory instrument, other states which are to be regarded as safe for the purposes of, in the case of Part 3, both the Refugee Convention and the ECHR and, in the case of Part 4, the Refugee Convention only. The statutory instrument requires the approval of both Houses of Parliament. The consequences of such listing vary between the different Parts. In short:

Part 5 is headed “Countries Certified as Safe for Individuals”. Paragraph 17 provides:

“This Part applies to a person who has made an asylum claim if the Secretary of State certifies that—

(a)

it is proposed to remove the person to a specified State,

(b)

in the Secretary of State’s opinion the person is not a national or citizen of the specified State, and

(c)

in the Secretary of State’s opinion the specified State is a place—

(i)

where the person’s life and liberty will not be threatened by reason of his race, religion, nationality, membership of a particular social group or political opinion, and

(ii)

from which the person will not be sent to another State otherwise than in accordance with the Refugee Convention.”

Paragraph 19 provides:

“Where this Part applies to a person—

(a)

…

(b)

he may not bring an immigration appeal in reliance on an asylum claim which asserts that to remove the person to the State specified under paragraph 17 would breach the United Kingdom’s obligations under the Refugee Convention,

(c)

he may not bring an immigration appeal in reliance on a human rights claim if the Secretary of State certifies that the claim is clearly unfounded, and

(d)

…”

There is accordingly no presumption of the kinds provided for in Parts 2-4. Part 5 is concerned only with a bar on the bringing of immigration appeals. The relevant certifications – under paragraphs 17 (c) and 19 (c) – necessarily require an assessment of the risk to the particular individual.

Rwanda is not listed in Part 2 and has not been specified in an order made under Part 3 or Part 4. The Secretary of State has, however, in the case of each of the Claimants certified under paragraph 17 (c) that Rwanda is a place where their life and liberty will not be threatened on any of the specified grounds and from which they will not be refouled – in short, that it is safe. (She has also, as we have seen, certified their human rights claims as clearly unfounded under paragraph 19 (c); but that is not the material certification for the purpose of this issue.) The decision letters in each case, which are in this respect in standard form, say that the decision is based in part on the judgment in the Asylum System CPIN.

The Claimants’ case as regards this issue can be summarised as follows:

I do not believe that that case is well-founded. Specifically, I do not accept that the assessments contained in the CPINs constitute presumptions of the same kind as those enacted in parts 2-4 of Schedule 3: they are of an essentially different character. The Preface to the Assessment CPIN begins, in what is evidently standard language for CPINs, by describing their purpose and nature generally. The fourth paragraph reads:

“In addition to background information obtained from a range of sources, [CPINs] also include relevant caselaw and our (CPIT’s) general assessment of the key aspects of the refugee status determination process (that is risk, availability of protection, possibility of internal relocation, and whether the claim is likely to be certified as ‘clearly unfounded’).”

It continues:

“This note provides an assessment of Rwanda’s asylum system, support provisions, integration opportunities as well as some of the general, related human rights issues for use by Home Office decision makers handling particular types of protection and human rights claims.

It is not intended to be an exhaustive survey of a particular subject or theme.

It analyses the evidence relevant to this note – that is: information in the contained in the separate country information reports (see below); refugee/human rights laws and policies, in particular paragraph 345B of the immigration rules which sets out when a country is a ‘safe third country of asylum’; and applicable caselaw – describes this and its inter-relationships, and provides an assessment of whether, in general, there are substantial grounds for believing that a person, if relocated to Rwanda, would face a real risk of being subjected to treatment contrary to Article 3 of the European Convention on Human Rights (ECHR).”

After referring to the other CPINs which I have identified at para. 138 above, it concludes:

“Decision makers must, however, still consider all claims on an individual basis, taking into account each case’s specific facts [emphasis in original].”

That approach recognises that the exercise for which CPINs are required is one of taking decisions on a case-by-case basis: that is stated in terms in the final sentence but is in fact clear from the entirety of the passage. The assessments contained in the CPIN will of course be general in character, but that is not inconsistent with individualised decision-making. As Lord Kerr observed at para. 70 of his judgment in R (EM (Eritrea)) v Secretary of State for the Home Department [2014] UKSC 12, [2014] AC 1321, “[t]he court must examine the foreseeable consequences of sending a claimant to the receiving country bearing in mind both the general situation there and the claimant’s personal circumstances [my emphasis]”. As regards the first element, it is not only legitimate but inevitable that caseworkers will have regard to general assessments of the kind contained in CPINs. As the Secretary of State observes at para. 91 of her skeleton argument,

“The notion of a completely ‘ad hoc’ or ‘one-off’ assessment is unreal: it is inevitable that recurring issues will arise in relation to particular countries which it is desirable to resolve on a consistent basis.”

I accept that some of the assessments contained in the CPINs are likely in practice to be treated as definitive of the particular question which they address – see, for example, the assessment about the risk of refoulement quoted at para. 297 above. But that does not mean that the nature of the exercise under Part 5, which requires individualised decision-making, can be equated with the application of a presumption under Parts 2-4.

That being so, the Secretary of State cannot in my view be regarded as circumventing Parliament’s intention by choosing in these cases to make use of her individual certification powers under Part 5 rather than seeking to have Rwanda listed as a generally safe country under Part 3 or 4. The provisions of the various Parts offer different possible mechanisms for limiting claimants’ rights to challenge safe third country decisions. Each mechanism will to some extent involve the Secretary of State making a general assessment of matters relevant to the safety of the country in question, and to that extent there is a degree of overlap between them. Nevertheless, each has its different characteristics and its different advantages and disadvantages. Parts 3 and 4 give the Secretary of State the advantage of being able to apply a general presumption of safety, without having to make an individualised assessment, but they involve the no doubt cumbersome exercise of enacting secondary legislation and obtaining Parliamentary approval. I can see no basis for implying into the statute an obligation to use the Parts 2-4 mechanism in every case where the Secretary of State makes a “general” assessment of some kind. Indeed, given the fact that, as I have said, there will be an element of such assessment in every individual decision, it is hard to see how such an implication would be workable.

I would accordingly uphold the Divisional Court’s conclusion on this issue also. Again, I have not found it necessary to set out its reasoning, which is at paras. 78-84 of its judgment; but it is to the same effect as mine.

The eighth issue considered by the Divisional Court was defined by it as follows:

“Have there been breaches of the Data Protection Act 1998 and/or the UK General Data Protection Regulation in the implementation of the Rwanda policy? Do such breaches invalidate decisions taking under either paragraph 345A or 345C of the Immigration Rules?”

That issue arose from grounds of challenge pleaded by SAA and argued before the Divisional Court, as also before us, by Mr Manjit Gill KC. I will refer to them as “the data protection grounds”.

The Divisional Court dismissed the data protection grounds and refused permission to appeal. In my judgment dated 14 March 2023 considering the outstanding applications for permission to appeal I directed that the application be adjourned to the hearing of the appeal on the other issues, on a “rolled-up” basis. In the event we heard oral submissions from Mr Gill directed simply to the issue of permission.

For the reasons I give below I would refuse permission to appeal. That being so, it is unnecessary that I set out the factual and legal background, which are fully set out in the judgment of the Divisional Court.

The Court addressed the data protection grounds at paras. 127-149 of its judgment. It explains at paras. 127-130 that, pursuant to the terms of the MoU, once SAA had been served with a Notice of Intent to relocate him to Rwanda the Home Office provided the Rwandan authorities with details of SAA’s name, date of birth, sex, nationality, and the date he made his asylum claim in the United Kingdom, together with a photograph. At para. 131 it summarised SAA’s case as follows:

“First, that transfer of personal data to Rwanda on the terms set out in the MOU is contrary to the requirements in Chapter V of Retained European Parliament and Council Regulation (2016/679/EU), better known as the United Kingdom General Data Protection Regulation (‘the UK GDPR’). Chapter V makes provision regulating the transfer of personal data to third countries. Second, that the Home Secretary has failed to comply with article 13 of the UK GDPR, which requires a data controller when obtaining personal data from a data subject to provide information, for example on the purposes for which the data obtained will be processed. Third, that the data protection impact assessment prepared by the Home Secretary in respect of the MEDP, to meet the requirements of article 35 UK GDPR, is defective.” 

The Divisional Court did not, however, decide any of those issues. Instead, at para. 132, it pointed out that there was what it called a “logically prior issue”, namely:

“Even assuming that SAA is correct on any or all of his submission on compliance with the UK GDPR, does that affect the legality of any decision the Home Secretary has taken under paragraph 345A or 345C of the Immigration Rules such that it would be appropriate to quash that decision for that reason?”

As to that, it recorded Mr Gill’s case as follows (para. 133):

“The submission for SAA is to the effect that the power to make decisions under the Immigration Rules (i.e., decisions under paragraph 345A and 345C) depended on compliance with whatever requirements might arise either under the UK GDPR or under its counterpart, the Data Protection Act 2018 …  In consequence, failure to comply with data protection law would require the conclusion that the immigration decisions were unlawful and should be quashed.”

The Divisional Court did not accept that submission. Between paras. 135 and 142 it considered each of the three complaints identified in para. 131 and concluded that even if each was made out it would not have the consequence that the decision to relocate SAA should be quashed. In short:

Having performed that exercise, it concluded, at para. 143:

“All this being so, the data protection law submissions in this case are not capable of producing the conclusion that the Home Secretary’s decisions under Immigration Rules are unlawful.” 

It went on at paras. 144-148 to express some views on the substance of SAA’s complaints but on an expressly obiter basis.

SAA now advances a single ground of appeal, ground 23 in the consolidated grounds, which reads:

“The court erred in holding that the Appellant’s data protection arguments were not relevant to any public law decision, and it erred in not addressing the Appellant’s data protection arguments lawfully, or adequately, or at all.”

That challenges the basis on which the Divisional Court decided the eighth issue, but in order to identify the reasoning behind the challenge it is necessary to look at the skeleton argument. This makes thirteen points, at paras. 32-47. I will consider those points in turn. I should say that Mr Gill did not seek to develop them in his oral submissions, which were addressed to broader questions which I have to say did not appear to me to bear on the Divisional Court’s dispositive reasoning.

The first point is that the Divisional Court’s “logically prior issue” had not been raised by the Secretary of State but was instead raised by the court itself in the course of oral submissions. Even if that is correct, it goes nowhere if the point is in fact good.

The second point is that the Divisional Court “did not address the relevant public law decisions ... or consider how the data arguments related to them”. I do not accept that. On the contrary, it is precisely the exercise which the Court carried out between paras. 135 and 142 of its judgment.

The third point arises from para. 134 of the Divisional Court’s judgment, where it said:

“As a matter of principle, it cannot be that any breach of any rule on the part of a public authority or for which that authority is responsible, occurring in the context of either making or executing a public law decision will necessarily affect the validity of that public law decision. To take an obvious example, if a person being removed from the United Kingdom was assaulted by a Home Office official on his way to the airport, that assault would be unlawful but would not in itself compromise the legality of the immigration decision that was the reason for removal.  On its facts, this example is some way distant from the cases now before us.  However, on the facts that are before us, the same conclusion should be reached.”

It is said that the assault example is not analogous with the circumstances of the present case. But the Court expressly acknowledged that. The example was simply offered as a vivid example of the general principle. The relationship between the alleged breaches of data protection law and the public law decisions taken in SAA’s case was carefully considered in the paragraphs to which I have referred.

The fourth point is that “compliance with the SSHD’s obligations under DPA 2018 and UK GDPR is a necessary prerequisite or a sine qua non in practice of a removal decision under para. 345C”. That may be correct, but it does not follow that non-compliance invalidates the decision.

The fifth point is that the decision whether Rwanda was a safe third country within the meaning of para. 345B required a consideration of whether its data protection regime was adequate. Mr Gill contended that there was a wealth of evidence that that was not the case and that the Secretary of State had failed to assess the relevant risks by carrying out a properly comprehensive DPIA. But a case of that kind does not fall within the eighth issue as defined by the Divisional Court, which is the subject of the ground of appeal. (It does not in fact appear that the Court believed that such a case was before it at all, and I have not been able to identify it in SAA’s original grounds of judicial review.)

The sixth point is that a DPIA had to be carried out before a decision was reached to transfer the personal data in question or under paragraphs 345B and 345C. That may be so, but it does not address the Divisional Court’s point that the carrying out of the DPIA was not integral to the removal decisions taken.

The seventh and eight points are, respectively, that “the burden was on SSHD to carry out such an inquiry and to undertake such a DPIA” and that a DPIA was necessary not only by reference to article 35 itself but also because it was required by article 71 of “the Withdrawal Agreement 2020”. But again, neither point addresses the question whether its carrying out was integral to the removal decisions.

The ninth point is that the scope of the DPIA should conform to the guidance given in the case-law of the Court of Justice of the European Communities in such cases as Schrems v Data Protection Commissioner C-362/14, [2016] QB 527, and Data Protection Commissioner v Facebook Ireland Ltd C-311/18, [2021] 1 WLR 751. The Divisional Court had in the obiter section of its judgment said that those cases did not provide the appropriate yardstick, and that point is challenged in the skeleton argument. But none of this goes anywhere if the Divisional Court’s logically prior point is good.

The tenth point develops the argument that the Secretary of State failed to comply with the requirements of article 13 of the UK GDPR, noting that the Divisional Court had in the obiter part of its judgment appeared to accept that that was so. But that only serves to emphasise that the issue here is not whether there was a breach but whether it impacted on “the validity of the public law decision” (see para. 383 (2) above). That question is not addressed.

The eleventh point is that “all such enquiries” – which I take to be a reference to the DPIA but perhaps also to the requirements of Chapter V of the UK GDPR – had to be made prior to the transfer of the data, which was itself required by the GoR prior to any relocation decision. But the answer is the same as to the fourth point: it does not follow that non-compliance invalidates the decision.

The twelfth point is that

“… the SSHD would have to investigate and inquire if there are any other risks in Rwanda to the essence of the asylum seekers correlated human rights under the ECHR, including to his rights to data protection under Art 8 ECHR, such as to make a removal decision under para 345C inappropriate”.

As with the fifth point, a case that paragraph 345C had not been complied with because of a risk of data protection breaches in Rwanda that would impact on SAA’s privacy is not within the scope of the eighth issue (nor have I been able to identify it in the original grounds of judicial review).

The thirteenth point appears to be that the Secretary of State is relying on the MEDP, which is a treaty taking effect at the level of international law, in order to justify a breach of domestic data protection law. That bears no relation to the basis on which the Divisional Court decided the eighth issue – or, so far as I can see, to any argument advanced by the Secretary of State.

Although I have thought it right to go through each of the points in the skeleton argument, the simple fact is that neither there nor in Mr Gill’s oral submissions has SAA advanced any response to the Divisional Court’s dispositive reasoning that has any real prospect of success.

I would accordingly refuse permission to appeal on ground 7.

In addition to the seven claims brought by individual Claimants, on 9 June 2022 the charity Asylum Aid brought judicial review proceedings seeking to challenge

“… [t]he procedure and arrangements (including the Inadmissibility Guidance …) adopted by the SSHD in respect of forced removals to Rwanda made pursuant to the Migration and Economic Development Partnership announced on 14 April 2022”.

As developed, its challenge was specifically to the fairness of the procedure leading up to the decision in the case of any individual that they should be relocated to Rwanda – or, more accurately, the several decisions which lead to that outcome. The challenge is to the overall fairness of the system and is generic in character: it is not concerned with the fairness of the decisions taken in the cases of any individual Claimants. That challenge is considered at paras. 380-429 of the Divisional Court’s judgment, together with some related challenges to the fairness of the procedure advanced by individual Claimants. It was dismissed.

Permission to appeal was given by the Divisional Court on six grounds, to which I will return below: they were originally numbered 1-6 but they have been re-numbered 15-20. Those grounds were advanced before us by Ms Charlotte Kilroy KC. The AAA Claimants and RM also have permission to advance grounds on procedural fairness (grounds 21 and 22 respectively). The AAA Claimants do not raise any point not advanced by Asylum Aid; I address RM’s ground at para. # below. The charity Freedom from Torture and the United Nations Special Rapporteur on Trafficking in Persons were given permission to intervene by written submissions. I have read their submissions and bear them in mind in considering the issues raised by Asylum Aid’s grounds.

For ease of reference, I repeat here the terms of paragraphs 345A-345D of the Immigration Rules:

“Inadmissibility of non-EU applications for asylum

345A. An asylum application may be treated as inadmissible and not substantively considered if the Secretary of State determines that:

(i)

the applicant has been recognised as a refugee in a safe third country and they can still avail themselves of that protection; or

(ii)

the applicant otherwise enjoys sufficient protection in a safe third country, including benefiting from the principle of non-refoulement; or

(iii)

the applicant could enjoy sufficient protection in a safe third country, including benefiting from the principle of non-refoulement because:

(a)

they have already made an application for protection to that country; or

(b)

they could have made an application for protection to that country but did not do so and there were no exceptional circumstances preventing such an application being made, or

(c)

they have a connection to that country, such that it would be reasonable for them to go there to obtain protection.

Safe Third Country of Asylum

345B. A country is a safe third country for a particular applicant, if:

(i)

the applicant’s life and liberty will not be threatened on account of race, religion, nationality, membership of a particular social group or political opinion in that country;

(ii)

the principle of non-refoulement will be respected in that country in accordance with the Refugee Convention;

(iii)

the prohibition of removal, in violation of the right to freedom from torture and cruel, inhuman, or degrading treatment as laid down in international law, is respected in that country; and

(iv)

the possibility exists to request refugee status and, if found to be a refugee, to receive protection in accordance with the Refugee Convention in that country.

345C.    When an application is treated as inadmissible, the Secretary of State will attempt to remove the applicant to the safe third country in which they were previously present or to which they have a connection, or to any other safe third country which may agree to their entry.

Exceptions for admission of inadmissible claims to UK asylum process

345D.    When an application has been treated as inadmissible and either

(i)

removal to a safe third country within a reasonable period of time is unlikely; or

(ii)

upon consideration of a claimant’s particular circumstances the Secretary of State determines that removal to a safe third country is inappropriate

the Secretary of State will admit the applicant for consideration of the claim in the UK.”

The decision-making process followed in the cases of the individual Claimants is fully set out at paras. 29-35 of the Divisional Court’s judgment. It was common ground that the same process was intended to be followed in all cases involving removal to Rwanda under the MEDP. I will gratefully reproduce the Court’s account, though in slightly edited and abbreviated form.

Persons considered as potentially appropriate for relocation to Rwanda would have been detained shortly after arrival in the United Kingdom, and would be subject to the usual steps applied to all newly-detained persons. Among other matters, each person detained was to be:

If they made an asylum claim, as all the Claimants did, shortly after the claim was made (usually within a day or so) they would attend an asylum screening interview.

Each claim was then considered by the Home Office National Asylum Allocations Unit (“the NAAU”). The Inadmissibility Guidance provided that if the NAAU suspects that the claimant “... may [in the course of travelling to the United Kingdom] have spent time in or have a connection to a safe third country ...” the case must be referred to the Third Country Unit (“the TCU”) for consideration of whether an inadmissibility decision should be taken. The TCU then reviewed claims referred to it to determine whether they “... [appear] to satisfy paragraphs 345A and 345B of the Immigration Rules”. If a case fell into this category, the TCU would issue the claimant with a standard-form Notice of Intent.

I need not set out the full terms of the Notice of Intent, but it begins as follows:

“We have evidence that before you claimed asylum in the United Kingdom, you were present in or had a connection to [name the safe country or countries]. This may have consequences for whether your claim is admitted to the UK asylum system.

We will review your particular circumstances and the evidence in your case, and consider whether it is reasonable to have expected you to have claimed protection in [country or countries] (or to have remained there if you had already claimed or been granted protection), and whether we should consider removing you there or elsewhere.”

It goes on to inform the recipient that if inadmissibility action appears appropriate a safe third country, including Rwanda, may be asked if it will admit him or her. The Notice includes the following statement:

“If you wish to submit reasons not already notified to the Home Office why your protection claim should not be treated as inadmissible, or why you should not be required to leave the UK and be removed to the country or countries we may ask to admit you (as mentioned above), you should provide those reasons in writing within 7 calendar days [for detained cases] or 14 calendar days [for non-detained cases] of the date of this letter. After this period ends, we may make an inadmissibility decision on your case, based on the evidence available to us at that time.”

Since in cases of the kind with which we are concerned the claimant will be in detention, the relevant period for the provision of such reasons would be seven days. It was the Secretary of State’s evidence, although this is not stated in the Notice of Intent, that that period can, as a matter of discretion, be extended on request: see para. 434 below.

Following the expiry of the seven-day period the Secretary of State proceeds to take the relevant decisions, being:

Although that fourfold break-down of the relevant decisions appears to reflect the structure of the decision letters, Ms Kilroy submitted that in substance the Secretary of State was obliged to make at least six decisions, identified at paras. 43-50 of Asylum Aid’s skeleton argument. It is unnecessary to go through her analysis because the decisions in question can be broken down in various ways: what matters is her overall point, which I accept, that the decision-making process may involve numerous questions and that that will be reflected in the matters on which claimants may wish to make representations. I note, because our focus so far has been on issues relating to the safety of Rwanda, that one issue that may arise under paragraph 345A is whether there were exceptional circumstances preventing a claimant from making an application for asylum in any safe third country through which they may have passed: see head (iii) (b).

Removal directions can be issued as soon as those decisions are communicated. In the case of the Claimants, removal directions were issued only a few days after the initial decisions in late May/early June 2022. In accordance with normal practice governing removal directions, there would be an interval of at least five days between the date of the direction and the proposed date of removal.

I should mention one other point about the structure of paragraphs 345A-345D. A claimant may make representations which do not go either to the issue of inadmissibility under paragraph 345A or to whether Rwanda is a safe third country under paragraph 345C (which brings in the criteria in paragraph 345B). Specifically:

Asylum Aid’s primary case, as Ms Kilroy confirmed, is that the seven-day period allowed for the making of representations following the service of a Notice of Intent will in virtually every case be too short to allow for the making of any effective representations. Its ground 17 challenges the Divisional Court’s rejection of that case.

The length of time reasonably required will of course depend on a number of factors, including (a) the scope of the matters on which representations may be made, (b) whether the claimants will need legal advice and (c) whether they need further information from the Secretary of State in order to be able to make effective representations. The Divisional Court made findings on those questions, and they are the subject of the challenges in grounds 15-16 and 18. Grounds 19 and 20 are of a rather different character. Although it might be more logical to re-order the grounds, I have found it simpler to take them in the order that they are advanced.

In the light of our decision on the issue of the safety of Rwanda no removals under the MEDP will be proceeding for some time. To that extent, the generic challenge to the system for taking decisions for such removals has become academic. That does not mean that we should not determine Asylum Aid’s appeal, because removals may resume in due course, whether as a result of a successful appeal or because the defects in the Rwandan asylum system are effectively addressed. However, we should confine ourselves to deciding the specific issues raised by the grounds, and should eschew offering any wider guidance about what procedural fairness might entail in future cases of this kind, where the circumstances may be different. In any event, the hearing before us did not afford an opportunity for a general review of the procedure adopted.

There was no dispute before us as to the fundamental requirements of common law fairness. They are classically stated by the House of Lords in R v Secretary of State for the Home Department, ex p Doody [1994] 1 AC 531, but we were referred also to the judgment of this Court in Balajigari v Secretary of State for the Home Department [2019] EWCA Civ 673, [2019] 1 WLR 4647, at paras. 45-61. It is well-recognised that procedural fairness is a matter of peculiar importance in asylum cases: see, e.g., the observations of Bingham LJ in Secretary of State for the Home Department v Thirukumar [1989] Imm AR 402, at p. 414. The application of those principles in a given case depends on the particular circumstances of that case, and I do not propose to burden this judgment with further citation of authority.

Ground 15 begins with a general averment that the Divisional Court was wrong to reject Asylum Aid’s case that the Rwanda policy was systemically unfair. However, that general pleading is followed by the specific criticism – introduced by the word “including” – of the Court’s finding that “procedural fairness did not require the provision of information relating to, or the opportunity to make representations on, the matters set out in §345B (ii)-(iv) …”; and the skeleton argument is directed only to that criticism. The issue raised by this ground is whether fairness required that asylum claimants served with a Notice of Intent should be able to make effective representations about the safety of Rwanda generally (i.e., broadly, the matters covered by heads (ii)-(iv) under paragraph 345B of the Immigration Rules) – “the general safety question” – or only about why it was unsafe for the particular claimant (broadly, head (i) under paragraph 345B) – “the specific safety question”. Ms Kilroy submitted that that dichotomy is artificial and breaks down when it has to be applied in practice. I have sympathy with that submission: see Lord Kerr’s observations in EM (Eritrea) referred to at para. 375 above. But the distinction plays an important part in the reasoning of the Divisional Court, and I will accept it for present purposes.

In the Divisional Court the Secretary of State’s initial position was that fairness required that claimants be entitled to make effective representations on the question whether “Rwanda is a safe country” simpliciter. But in the course of the October hearing, she changed her position as a result of questions from the Court. Her revised position was that fairness only required that claimants should be entitled to make representations about “whether there is any reason specific to the Claimant why Rwanda would not be a safe third country in the individual circumstances of the Claimant”. The Court permitted her to advance that revised case, but it was not possible to hear submissions about it at the hearing, and the issue was resolved on the basis of written representations lodged subsequently: more details can be found at paras. 386-387 of the judgment.

The Divisional Court accepted the Secretary of State’s revised submission. Its reasons appear at paras. 388-395. The relevant passage for our purposes starts with para. 390. The drafting is complicated by the fact that the Court was considering both this issue and the related issue of whether the Secretary of State should be expected to provide further information; but the gist of its conclusion was that fairness required that they be entitled to make representations on the specific but not the general safety question. At para. 391 the Court summarises Asylum Aid’s submissions to the contrary, which included the submission that it was impossible in practice to distinguish between the two questions. At para. 392 it responds to those submissions as follows:

“A distinction does exist between the criteria at paragraph 345B of the Immigration Rules. Criterion (i) is formulated by reference to the asylum applicant’s own circumstances and characteristics, criteria (ii) - (iv) are framed by reference to the general position in the country in question. The real issue is whether that distinction is material for the purposes of setting what is required by law for fair exercise of the paragraph 345C power to remove to a safe third country. Our conclusion is that the distinction between what an asylum claimant may be able to say about his own circumstances and how those might be relevant to whether he is removed to a particular country, and whether that country, generally, complies with its obligations under the Refugee Convention does determine the extent of the legal requirement of procedural fairness in this context. Procedural fairness requires that an asylum claimant should have the opportunity to make representations on matters within the criterion at paragraphs 345B(i) of the Immigration Rules. Those are matters relevant to any decision to remove (self-evidently) and matters the asylum claimant is uniquely placed to consider and explain. Matters known to the asylum claimant may be a relevant consideration; the Home Secretary must take it into account; and the duty to act fairly must apply to require the claimant to have an opportunity to make representations. The same applies to the criteria at paragraph 17(c) of Schedule 3 to the 2004 Act which are also directed to the specific position of the asylum claimant. Criteria (ii) - (iv) within paragraph 345B of the Immigration Rules are different, and require evaluation of whether, generally, the relevant country complies with its obligations under the Refugee Convention. Those matters will go well beyond the circumstances of any one asylum claimant; they are also criteria which the Home Secretary, given the resources available to her, is well-placed to assess. We do not consider that the duty that the Home Secretary act fairly in exercise of the power at paragraph 345C of the Immigration Rules requires an asylum claimant to have the opportunity to make representations on these matters. It is not enough to say that criteria (ii) - (iv) are relevant to the decision to remove and since the asylum claimant is the subject of that decision he must have a legal right to comment on those matters before the decision is made. That is a non-sequitur. The scope of the obligation to act fairly is measured in specifics. This is not to say that any individual faced with the possibility of removal to a third safe country could not seek to persuade the Home Secretary that one or other of criteria (ii) - (iv) was not met, and that if such representations were made, the Home Secretary should have regard to them. But such representations would not be made in exercise of any legal right arising out of an obligation to ensure procedural fairness. Further, to the extent that an asylum claimant may wish to make such representations he has sufficient information about what such representations must be directed to, by reason of paragraph 345B itself. That explains the matters the Home Secretary must consider. The legal duty to act fairly does not require the Home Secretary provide him with all the material available to her; the legal duty to act fairly does not in the present context require that an asylum claimant be put in the position to second-guess the Home Secretary's evaluation on criteria (ii) - (iv).”

I have italicised the parts of the passage which decide as a matter of principle the scope of the matters on which fairness requires that claimants should be entitled to make representations. It is true that the Court goes on to acknowledge that the Secretary of State might have to consider representations made otherwise than by way of legal right, but I need not for present purposes try to unpack that proposition, and I will focus on the italicised words. The last three sentences are concerned with the separate issue of whether the Secretary of State was obliged to provide asylum-seekers whom she proposed to relocate to Rwanda with “all the material she has relied on to decide that Rwanda is a safe third country (including the material she relied on to reach the conclusion that Rwanda would abide by its obligations under the MOU and the notes verbales)” (see para. 385 (2) of the judgment); but the issues are clearly related, and the Court’s reasoning clearly reflects its reasoning on the issue of the scope of the representations which they were legally entitled to make.

In the circumstances of the present case the Divisional Court’s conclusion in para. 390 of its judgment, amplified in para. 392, was in my respectful view wrong. As Ms Kilroy submitted, as a matter of principle a person should be permitted to make representations about any matter relevant to an adverse decision about their case, and it is no answer to say that they might not be as “well-placed” as the decision-maker to make the decision itself. I accept that there may in some cases be matters contributing to the decision about which representations would for some objective reason be pointless. But that is not the case here. An individual asylum-seeker may not personally know anything about Rwanda, but they are entitled to rely on the research or expertise of people or institutions representing their interests. The position might be different if there had already been a decision of the Court (or perhaps some other authoritative independent body). In that case fairness might indeed not require that claimants have the opportunity to seek to persuade the Secretary of State to go behind that decision (at least unless they could provide compelling evidence of a relevant change of circumstances (Footnote: ⁹)). The Divisional Court made this very point at para. 395 of its judgment, where it said:

“… [F]or the future … the generic issues raised by the Claimants as to why relocation to Rwanda would be unlawful have now been determined by this court (subject to any appeal) and subject to any relevant new information emerging.”

But those were not the circumstances here. There had been at the date of the removal decisions no determination by the Court; and fairness required that the Claimants have the opportunity to try to persuade the Secretary of State that her conclusion on the general safety issue was wrong.

I should say that the Secretary of State did not make any substantive submissions in support of the impugned statement. Rather, Lord Pannick’s submission to us was that the issue was academic because the Secretary of State had in fact undertaken to consider all submissions, whether on the general or on the specific safety question. Whether or not that is so, the point is important in principle.

I should record that the Secretary of State filed a Respondent’s Notice in relation to this ground, contending that if there was in fact “a common law obligation to invite representations on the general safety of Rwanda, it was sufficient that the Secretary of State gave notice that she considered Rwanda to be a safe country and invited representations as to whether there was any reason the person should not be removed from the UK to Rwanda”. I am inclined to think that that is correct, but even if it is it does not affect the substance of Asylum Aid’s challenge.

In summary, I believe the point raised by ground 15 is good, but it does not necessarily impugn the Divisional Court’s overall conclusion.

At para. 403 of its judgment, in the section addressing procedural failures alleged by individual Claimants, the Divisional Court said:

“There have been criticisms of the lack of access to legal advice. Given the scope of the right to make representations in this context, we do not consider that procedural fairness requires that a person who is at risk of action under the Inadmissibility Guidance be provided with legal representation for the right to make representations to be an effective right [my italics]. It is essentially a matter of fact as to why he did not claim asylum in a third country on route to the United Kingdom. It is essentially a matter of fact for him to give his reasons why he should not be removed to Rwanda.”

Notwithstanding that general conclusion, it went on briefly to consider what access to legal advice the individual Claimants had in fact had and it found that the process was in that regard fair.

Asylum Aid’s challenge is only to the statement which I have italicised. Even if the Claimants in these cases did in fact all have access to legal advice, if that statement is wrong it is capable of leading to injustice if relied on in other cases, and I believe we should address it.

The impugned statement proceeds on the basis that fairness only requires that claimants should be entitled to make representations on circumstances arising from their specific factual histories, so it would in any event have to be reconsidered in the light of my conclusion on ground 15. But Ms Kilroy submitted that, even on the basis on which the Court proceeded, the statement cannot be defended. She submitted that the right to make representations must be effective, and that that went beyond a right simply to supply relevant facts to the decision-maker. It was unrealistic to believe that asylum-seekers who had just arrived in the UK and (usually) spoke no English could make effective representations, even on the specific safety question, without legal assistance, let alone understand the relevant law – and all the more so in the light of the scope and complexity of the decision-making scheme governing removals to Rwanda under the MEDP. She also pointed out that the Secretary of State’s decisions in the present cases show that she insists on the provision of documentary materials, such as medical reports, by way of supporting evidence, and that asylum-seekers in the Claimants’ situation would plainly be unable to comply with those requirements without professional assistance.

As with ground 15, the Secretary of State did not, either in her skeleton argument or in Lord Pannick’s submissions before us, advance any substantial submissions in support of the impugned statement, pointing out that in the Divisional Court she had sought to resist this part of Asylum Aid’s challenge on the factual basis that claimants were provided with access to state-funded legal advice.

In my view it is impossible, for essentially the reasons given by Ms Kilroy, to support a general proposition (if this is really what the Court intended) that procedural fairness will never require that an asylum-seeker who is at risk of removal to Rwanda under the MEDP be provided with legal assistance to make representations before the removal decision is taken, even if they are not addressing the general safety question. There may be cases where a decision is fair even where there has been no access to legal assistance, but they are likely to be exceptional. As we have seen, the Secretary of State does not contend otherwise and it is her policy to ensure that legal assistance is indeed available.

I therefore believe the point raised by ground 16 is good, but, again, it does not necessarily impugn the Divisional Court’s overall conclusion.

It is, as I have said, Asylum Aid’s case that seven days was simply too short a period for claimants to prepare effective representations in response to a Notice of Intent to remove them to Rwanda. The Divisional Court rejected that submission on the basis that its finding that fairness did not require that claimants have an opportunity to make representations about the general safety question meant that seven days were adequate: see para. 421 of the judgment. That reasoning is undermined by my conclusion on ground 15. In any event, Ms Kilroy did not accept that even on that basis seven days would be adequate. In those circumstances, I will focus directly on the parties’ submissions before us rather than considering the issue through the lens of the judgment.

I start with the applicable legal principles. At para. 27 of his judgment in Lord Chancellor v Detention Action [2015] EWCA Civ 840, [2015] 1 WLR 5341, another case involving the adequacy of procedural timetables, albeit in a different context, Sir John Dyson MR reviewed the relevant authorities and accepted counsel’s summary of their effect, as follows:

“… (i) [I]n considering whether a system is fair, one must look at the full run of cases that go through the system; (ii) a successful challenge to a system on grounds of unfairness must show more than the possibility of aberrant decisions and unfairness in individual cases; (iii) a system will only be unlawful on grounds of unfairness if the unfairness is inherent in the system itself; (iv) the threshold of showing unfairness is a high one; (v) the core question is whether the system has the capacity to react appropriately to ensure fairness (in particular where the challenge is directed to the tightness of time limits, whether there is sufficient flexibility in the system to avoid unfairness); and (vi) whether the irreducible minimum of fairness is respected by the system and therefore lawful is ultimately a matter for the courts. I would enter a note of caution in relation to (iv). I accept that in most contexts the threshold of showing inherent unfairness is a high one. But this should not be taken to dilute the importance of the principle that only the highest standards of fairness will suffice in the context of asylum appeals.”

That summary was approved by the Supreme Court in A: see para. 68 of the judgment of Lord Sales and Lord Burnett. Point (v) is of particular relevance for our purposes. I return to the actual decision in Detention Action at para. 439 below.

Ms Kilroy referred us to a wealth of evidence identifying the difficulties faced by claimants, even if they have good legal assistance. It is simplest if I reproduce verbatim, with some minor editing, the summary of that evidence at paras. 23-25 of her skeleton argument.

“23.

First, there are a wide range of relevant issues that may be specific to an individual, such as mental or physical health, family ties in UK, or the need for family reunion once recognised as a refugee, which require instructions and evidence. Particular difficulties meeting the tight time limits may arise where there is a history of trauma, including torture, trafficking, and/or sexual or gender-based violence, which makes it difficult to take full instructions at speed. In addition to inadmissibility and safe third country issues, investigations and representations must be conducted on more standard issues, such as lawfulness of detention, applications for bail, trafficking indicators, and NRM referrals, and any age dispute issues.

24.

Second, Asylum Aid’s evidence [from Toufique Hussain, a solicitor at Duncan Lewis with extensive experience in this field] was that it was generally not possible to prepare witness statements addressing even individualised matters relevant to inadmissibility and removal to Rwanda within the seven-day period. As shown in [tables produced to us showing the timetables for the representations made in these cases], in only one case was a witness statement served in advance of the initial decisions, and that had to be supplemented later with further instructions. In all other cases, witness evidence providing relevant individualised information was produced after removal directions had been set.

25.

The Court has recognised the relevance of medical evidence to the decisions in individual cases. Again, Asylum Aid’s evidence showed that it was not possible to obtain medical evidence within a seven-day timescale. In all the Claimants’ cases where medical evidence was produced by the individual (and all but one when it was produced through a rule 35 report) it was only obtained after removal directions were set.”

Lord Pannick did not seek to address the details of that evidence. As he put it in his submissions, “seven days may or may not be adequate: it depends on the case”. His essential point was that claimants were entitled to ask for an extension if they needed more time, and that it was the Secretary of State’s policy to grant an extension if reasonable grounds were shown. He pointed out that such extensions had in fact been granted in several of the Claimants’ cases and that if one was unreasonably refused they were entitled to seek judicial review. He acknowledged that the Notice of Intent makes no reference to the right to seek an extension, nor does it appear in any guidance to caseworkers. He referred us, however, to para. 23 of the witness statement of Mr Ruaridh McAskill, the Acting Head of the TCU in Glasgow, which reads:

“Where individuals request an extension of time to respond to an NOI these requests are considered on a case-by-case basis taking into account the reasons requested for the extension, how long they have had to respond, their access to legal representation and whether it would be reasonable to have expected reasons to have been submitted prior to the extension request. There are no fixed criteria for an extension to be granted or refused. Factors weighing in favour of an extension of time include if someone has not had access to adequate legal advice, if they provide good reasons why they have not been able to make representations (such as illness, trauma or electronic communication failures); factors weighing against include if they had the opportunity to instruct legal representatives and chose not to do so. Discretion has been used in most cases to extend the time period to respond to the NOI. If an extension is refused further representations are generally considered.”

Lord Pannick drew attention to the final sentence. The seven-day period did not represent an absolute cut-off, even where an extension was not sought.

Lord Pannick submitted that the existence of that policy meant that the process was sufficiently flexible to avoid systemic unfairness and accordingly could not itself be said to be unfair or unlawful. He submitted that there were good reasons why “the standard period should be as short as seven days”. It was in the public interest that removal decisions should be made without delay: indeed that was in the claimant’s interest also, particularly if they were in detention.

In support of his submissions Lord Pannick relied on the decision of this Court in R (Refugee Legal Centre) v Secretary of State for the Home Department [2004] EWCA Civ 1481, [2005] 1 WLR 2219. That case, which is one of the decisions on which Sir John Dyson drew in his summary of the relevant principles in Detention Action, involved a challenge to the fast-track system for adjudication of asylum claims which the Secretary of State believed to be clearly unfounded. Under that system the entire process from initial interview to final decision was compressed into a period of three days, with no opportunity for the claimant to put in further representations following the interview. The basis of the challenge was that that timetable was quite inadequate for the preparation of the claimant’s case, including in particular any medical or other evidence that might need to be obtained. A central part of the Secretary of State’s answer, as here, was that the system was operated flexibly, so that if it became clear that a case could not be dealt with fairly within that timetable caseworkers would remove it from the fast track. This Court believed that the system was defective in the absence of what it described (see para. 18 of the judgment of the Court given by Sedley LJ) as “a clearly stated procedure – in public law, a policy – which recognises that it will be unfair not to enlarge the standard timetable in a variety of instances”. It continued, at para. 19:

“To assert, without any real evidence to support it, that a general principle of flexibility is ‘deeply ingrained’ is not good enough. Putting the relevant issues in writing - and we assume without question that what is put in writing will be made public - is not simply a bureaucratic reflex. It will concentrate official minds on the proper ingredients of fair procedure; it will enable applicants and their legal representatives to know what these ingredients are taken to be; and if anything is included in or omitted from them which renders the process legally unfair, the courts will be in a position to say so.”

However, it declined to hold that that defect meant that the policy itself should be declared to be unlawful. As it put it at para. 23:

“… [P]rovided that it is operated in a way that recognises the variety of circumstances in which fairness will require an enlargement of the standard timetable - that is to say lawfully operated - the … system itself is not inherently unfair. A written flexibility policy to which officials and representatives alike can work will afford a necessary assurance that the three-day timetable is in truth a guide and not a straitjacket.”

The significance of the Refugee Legal Centre decision does not depend on comparing the lengths of the periods in that case and this, or other details of the two systems. But Lord Pannick relied on it as demonstrating that there was nothing unlawful in the Secretary of State adopting a standard procedural timetable which would in some cases not give asylum claimants a fair opportunity to make representations provided that it was also her policy to allow for departures from the timetable where that was necessary in the interests of fairness. He also noted that the Court did not regard a failure to promulgate that policy in writing as rendering the system inherently unfair and thus unlawful.

It was put to Lord Pannick in the course of oral submissions that, notwithstanding the Court’s ultimate conclusion in the Refugee Legal Centre, it evidently expected that the Secretary of State would develop and publish its flexibility policy: it said at para. 25 that it had “indicated what in our view needs to be done to obviate [the risk of injustice]”. He was asked whether he accepted that the Secretary of State ought now to publish as a policy the approach which was said by Mr McAskill to operate in practice, and to refer to it in the Notice of Intent. He was not prepared to accept that proposition in those terms, arguing that published policies specifying detailed criteria were sometimes positively disadvantageous as encouraging a tick-box approach; the most that he would say was that when there had been some more experience of the system the Secretary of State might think it desirable to publish some guidance about the basis on which extensions should be granted.

Ms Kilroy in her reply submitted that the practice of granting extensions could not justify a standard timetable which on the evidence was too short in every case. She countered Lord Pannick’s reliance on the Refugee Legal Centre case by referring us to the decision in Detention Action. That case raised a challenge to the fairness of the Fast Track Rules introduced in the Immigration and Asylum Chamber of the First-tier Tribunal, which required asylum-seekers to prepare and present their appeals within seven days of the refusal of their claim. The Court found that that timetable was so tight that it was inevitable that a significant number of appellants would be denied a fair opportunity to present their cases. The response of the Lord Chancellor, as the rule-maker, was that rule 14 of the Fast-Track Rules provided that if the Tribunal was satisfied that the case could not justly be decided within those timescales it must disapply those Rules. The Court did not accept that that was an adequate safeguard, essentially because the procedural structure created constraints which would make it very difficult in practice for appellants to make an application under rule 14 or for the Tribunal to accede to it: see paras. 42-44 of the judgment of Sir John Dyson.

I have not found this issue entirely easy, but in the end I have concluded that the seven-day period specified in the Notice of Intent does not render the decision-making process “structurally unfair and unjust”, to adopt the language of Sir John Dyson in Detention Action; and I would accordingly dismiss ground 17. My reasons are as follows.

The evidence clearly establishes, and it is in any event obvious as a matter of common sense and experience, that in many cases it will indeed be impossible for claimants to submit effective representations within seven days of receipt of a Notice of Intent, even if they have ready access to legal assistance and only wish to make representations on matters specific to their particular circumstances. But I do not believe that it establishes that it will be impossible in every case. Not every case, for example, will require the submission of medical evidence; nor in every case will there be a factual basis for a claim of exceptional circumstances under paragraph 345A (b) (ii). It cannot be assumed that the Claimants’ cases are representative of the range of cases in which Notices of Intent in relation to relocation to Rwanda might be served: quite apart from the possibility of selection bias in those who brought proceedings and whose cases were heard by the Divisional Court, the process of identifying issues and gathering evidence is likely to be more uncertain and require more consideration when dealing with a newly-applied policy. Even if – as I accept may well be the case – the limit is too short in the majority of cases, it is impossible to assess the relative proportions. In short, I agree with Lord Pannick’s laconic summary that seven days may or may not be adequate.

That being so, I see nothing wrong in principle in the Secretary of State imposing a “base-line” timetable which is realistic at least in the most straightforward cases and allows those cases to be decided promptly, provided that she is ready and willing to grant extensions in those cases where more time is reasonably required. I do not believe that it is inherently unfair to employ a model where there is a minimum period available to all claimants to make representations, together with consideration of what longer period may be required in particular cases. There is, as Lord Pannick submitted, an important public interest in decisions on removals under the MEDP being made – one way or the other – as expeditiously as is consistent with fairness. Seven days is no doubt the shortest realistic period, but a deadline of, say, fourteen or twenty-one days would very likely still require extensions in some cases and would mean that decisions were delayed in others for a longer period than was necessary or desirable.

That puts the focus on the Secretary of State’s policy of flexibility. For the reasons given by Sedley LJ in the Refugee Legal Centre case, I do not think it is good enough that that policy is not formally published in the form of guidance to caseworkers nor referred to in the Notice of Intent. Claimants and their advisers need to know that the seven-day timetable can be extended where that is shown to be necessary in the interests of fairness. I do not accept Lord Pannick’s submission that publication of formal guidance may do more harm than good: there should be no risk of a tick-box approach if the guidance (which need not be elaborate) is expressed in appropriately flexible terms. It is also in my opinion important that the guidance makes it clear that the seven-day period should not be treated as a norm and that the grant of extensions is not necessarily exceptional. It may be that if experience shows that extensions are required in a very large number of cases, the Secretary of State might wish to consider providing for a rather longer base-line period; but that must be a matter for her.

I should make it clear that I should not be taken to be endorsing the Secretary of State’s current policy in the precise terms given by Mr McAskill, which were not the subject of submissions before us, beyond recording an observation by Ms Kilroy that they contained no specific reference to the overriding requirement of fairness. It must ultimately be a matter for the Secretary of State how she chooses to formulate any guidance which she may give.

Although for that reason I believe that this aspect of the decision-making process requires improvement, it follows from the approach of this Court in the Refugee Legal Centre case that that does not justify a declaration that the process is unlawful. I do not believe that that conclusion is inconsistent with Detention Action. The crucial feature in that case was that the power ostensibly available to the Tribunal under rule 14 was in reality highly constrained.

At para. 389 of its judgment the Divisional Court said:

“Contrary to the submission made by some of the Claimants, fairness did not require that the Claimants have the opportunity to make representations in response to some form of provisional view that such circumstances existed. What fairness requires in the context of a decision under paragraph 345A(iii)(b) of the Immigration Rules is an opportunity for the Claimant to provide any explanation he has for not making an asylum claim before reaching the United Kingdom. Fairness did not require the opportunity to make representations in response to the Home Secretary’s evaluation (or provisional evaluation) of that explanation.”

Ground 18 reads:

“The court is wrong to conclude that the common law does not require individuals to have access to the SSHD’s provisional conclusions against them.”

In its skeleton argument Asylum Aid makes clear that the challenge in ground 18 is to para. 389 of the Divisional Court’s judgment, as quoted above. It refers to various authorities which establish, as Lord Mustill put it in Doody (p. 563 F-H), that “the right to make representations is of little value unless the maker has knowledge in advance of the considerations which, unless effectively challenged, will or may lead to an adverse decision”.

In her skeleton argument in response the Secretary of State pointed out that the Notice of Intent informed claimants that she was considering whether it was reasonable to expect them to have claimed protection in the specified countries through which they had passed and that put them sufficiently on notice of the need, if they could, to advance reasons why that was not the case.

In her oral submissions Ms Kilroy made it clear that because of the pressure of time she largely relied on the contents of Asylum Aid’s skeleton argument; such short points as she did make did not substantially develop what appears there. Lord Pannick did not make any oral submissions. That being so, I need only say that I accept the Secretary of State’s submission and would dismiss this ground.

I should clarify one related point. I have concluded that ground 15 is well-founded. We did not, however, hear any submissions on the specific question of what, if any, information over and above what the Secretary of State published in the MoU and the CPINs she was obliged to provide in order to give the Claimants a fair opportunity to make representations on the safety of Rwanda; and I accordingly say nothing about that question. I would, however, observe that the materials which were in fact disclosed in these proceedings appear sufficient to satisfy any such obligation as there may have been.

It was part of Asylum Aid’s case that the inadequacy of the time given for representations had a knock-on effect on the adequacy of the standard five-day notice given in removal directions: see para. 385 (5) of the Divisional Court’s judgment. The Court addressed that submission at para. 420 of its judgment, where it said:

“One point to note in the present case is that the access to court submission is parasitic on the unfair system submission. Ms Kilroy accepted that if the period permitted for representations before the decisions was lawful, then removal directions within the standard form would be lawful.” 

Ms Kilroy told us that that did not accurately reflect what she had said to the Court; but, irrespective of whether in fact she made the concession attributed to her it seems to us to be plainly right. At paras. 29-32 of its skeleton argument (which Ms Kilroy did not, again for reasons of time, significantly amplify in her oral submissions) Asylum Aid argues that the dismissal of its case on the other grounds makes ground 19 unanswerable because if there has been inadequate opportunity to make representations prior to the removal decision claimants will need longer than five days to issue Court proceedings. But that depends on the basis on which the other grounds failed. If, for example, I had accepted the Divisional Court’s statement that it was unnecessary for claimants to have access to legal advice prior to a decision on removal under the MEDP, there would be force in the point that five days was an inadequate time to find and instruct a lawyer from scratch and to bring legal proceedings. But my conclusion on ground 17 means that the system has sufficient flexibility to ensure that claimants will in fact have adequate time to make effective representations (ignoring aberrant decisions, which can only be addressed on a case-by-case basis). If that is the case, the claim based on a systemic denial of access to justice does indeed fall away.

I would accordingly dismiss ground 19.

Ground 20 reads:

“The Court’s analysis of the Immigration Rules and para. 17 of Schedule 3 to the 2004 Act is flawed.”

Asylum Aid’s essential contention is that the distinction drawn by the Divisional Court, in connection with the “Scope of Representations” issue, between head (i) and heads (ii)-(iv) under paragraph 345B does not properly reflect the requirements of the Refugee Convention and is inconsistent with the approach required of the Secretary of State in making a certificate under paragraph 17 of Schedule 3. It is thus not a distinct ground but a further argument in support of ground 15. Since I have accepted Ms Kilroy’s point on ground 15 for other reasons I need say no more about it.

I do not accept all aspects of the reasoning of the Divisional Court on the issue of whether the decision-making process was inherently unfair, as is reflected in my conclusions on grounds 15 and 16. But, as I have said, the determinative ground from the point of view of the fairness of the process is ground 17. Since I would dismiss that ground I believe that the Court was right to reject the claim of inherent unfairness, and I would dismiss Asylum Aid’s appeal.

Ground 22, which is pleaded in some detail, raises both a general challenge to the fairness of the Secretary of State’s decision-making process as regards relocation to Rwanda and a specific challenge to its conclusion that in his particular case (unlike those of most of the other Appellants) the inadmissibility decision did not fall to be quashed. The pleaded criticisms were developed at paras. 11-36 of RM’s skeleton argument, but in his oral submissions Mr Drabble (who was also under some pressure of time) did no more than refer us to the fact that the skeleton argument contained a procedural chronology of his particular case which he invited us to read as illustrating with particularity why seven days would almost never give enough time for proper submissions to be made.

In the version of this judgment handed down (subject to editorial correction) on 29 June, I did not refer specifically to ground 22, and in post-judgment submissions counsel have asked for clarification as to whether the intention was that it should be dismissed. I should confirm that that is indeed the Court’s intention. RM’s general challenge raises no points that I have not considered in relation to Asylum Aid’s appeal. As regards the decision in his particular case, the Divisional Court gave reasons at paras. 357-358 of its judgment why the inadmissibility decision of 5 July 2022 was not unfair. That specific reasoning is not addressed either in ground 22 itself or in RM’s skeleton argument, and I can identify no error of law in it. I should note that RM did in fact make further submissions on 9 July, addressed to the question why he had not claimed asylum in France: those will no doubt be carefully considered by the Secretary of State if the issue of relocation to Rwanda becomes live in his case.

These appeals concern the lawfulness of the Home Secretary’s decisions to remove to Rwanda a cohort of people who arrived irregularly in the United Kingdom by small boats across the English Channel and then claimed asylum. There are ten individual appellants, all single men, whose countries of origin are Syria, Iran, Iraq, Vietnam, Sudan and Albania. To qualify for removal, it must have been reasonable for them to have claimed asylum in a safe country on the way. All arriving by small boats across the Channel have necessarily been in a safe third country but it may not be reasonable for all to claim asylum there. The claimants would be free to claim asylum in Rwanda. We are not concerned with the political merits of the underlying policy.

The appeals relate to generic claims which contend, for a variety of reasons, that it would be unlawful for the Home Secretary to remove anyone from the United Kingdom to Rwanda. The individual circumstances of the claimants play no part in their arguments. Before the Divisional Court of the High Court the individual appellants (and others) also raised challenges based upon the circumstances in which their individual cases were considered by the Home Office. Many were successful in those challenges with the consequence that new decisions will need to be taken on the individual cases whatever the outcome of these appeals on generic issues. The proceedings were issued between 8 and 14 June 2022. The evidence upon which the Divisional Court decided the issues was filed over the course of the summer of 2022. We have considered the appeal on the same evidence which, in consequence, is now out of date.

Asylum Aid was also a claimant for judicial review before the Divisional Court. Its core submission is that the procedures surrounding the decision making and proposed removals are unfair and thus unlawful. It appeals the Divisional Court’s dismissal of its claim.

The judgment of the Divisional Court was handed down on 19 December 2022 following hearings in September and early October: Lewis LJ and Swift J [2022] EWHC 3230 (Admin). By its orders the court dismissed the generic claims for judicial review. There has been no appeal by the Home Secretary in respect of the individual decisions.

At the heart of the claims for judicial review is the proposition that it would be unlawful to remove anyone to Rwanda because there are “substantial grounds for believing that they would be at real risk” of treatment in Rwanda contrary to article 3 of the European Convention on Human Rights (“ECHR”). That is the well-known test first articulated by the Strasbourg Court in Soering v. United Kingdom (1989) 11 EHRR 439 at [88] et seq in the context of an extradition case. It was soon applied to removal cases generally. The Soering argument in these appeals has two components which make up the “safe country” issue. First, the appellants submit that the conditions that they would face in Rwanda would give rise to the relevant risk. Secondly, they submit that defects in Rwanda’s consideration of asylum claims would give rise to a risk of good claims being refused and the further risk that Rwanda would return individuals (refoule in the language of the 1951 Refugee Convention) to the countries from which they claimed protection. 

The principal decision of the Strasbourg Court dealing with that test in cases of removal of asylum seekers by an ECHR state to another state it considers safe is Ilias v. Hungary (2020) 71 EHRR 6.

It is the second aspect of the article 3 issue that was the focus of argument before us. On that second aspect the task of the Divisional Court was to determine whether deficiencies in the Rwandan system for dealing with asylum applications are such that there are substantial grounds for believing that they would face a real risk of being returned to their countries of origin despite having a valid claim for asylum.

The claimants, supported by the United Nations High Commissioner for Refugees (“UNHCR”), submit that such grounds remain despite the bespoke agreement reached between the Government of the United Kingdom and the Government of Rwanda, known as the Migration and Economic Development Partnership (“the agreement”). The agreement is contained in a Memorandum of Understanding (“MoU”) dated 14 April 2022 and diplomatic Notes Verbales which provide guarantees by Rwanda regarding “the asylum process of transferred individuals” and “the reception and accommodation of transferred individuals” respectively. The position of the claimants and the UNHCR is that the system for dealing with asylum claims in Rwanda is such that the aspirations set out in these documents are undeliverable for any person seeking asylum there, including the 11 claimants involved in this appeal. It also entails the proposition that the proposed monitoring arrangements which are designed to ensure that Rwanda deals with asylum applicants appropriately with no relevant risk of refoulement for those with valid claims are likely to be ineffective.

It is the opinion of His Majesty’s Government that Rwanda will comply with the terms of the agreements and abide by the assurances it has given. The intense scrutiny upon those who would be sent to Rwanda coupled with the monitoring mechanisms contained in the agreement and those of the British High Commission in Kigali, with embedded Home Office officials, provide additional and necessary safeguards. The UNHCR disagrees, believing that the asylum system in Rwanda does not have the capacity to deliver consistently accurate and fair asylum decisions and that there is a concomitant risk that some refugees will not be recognised as such and will be subject to refoulement. Capacity in this sense is not concerned only with the number of asylum seekers who arrive in Rwanda but the ability of the various parts of the system there to deliver reliable decisions on their applications.

The evidence in support of the claims for judicial review was largely provided by the UNHCR through Lawrence Bottinick, Senior Legal Officer and Acting Representative for the UNHCR in London, expressing an institutional view. The evolution of the evidence was unusual in that it joined issue with the British Government point by point, with exchanges of evidence. The UNHCR, an interested party in these proceedings, assumed the mantle of claimant. That is not to criticise it but to draw attention to the unusual nature of its engagement in these claims. Moreover, the UNHCR made clear that it is opposed as an institution to any attempt “to offshore” asylum claims in the manner contemplated by the agreement in question. It has an institutional interest in the outcome of these claims for judicial review. It nonetheless has unrivalled practical experience of the working of the asylum system in Rwanda through long years of engagement.

Rwanda is a country which has emerged from one of the most shocking and destructive periods of violence in recent history. Ethnic rivalry led to genocide in 1994 when over 500,000 Tutsi were murdered by Hutu. Many Hutu were also murdered and estimates of the total death toll are much greater. There was protracted violence thereafter before stability was restored. The ethnic violence in Rwanda had a long history and erupted repeatedly in the second half of the twentieth century. This is not the place for even a short description. Following the 1994 genocide the International Criminal Tribunal for Rwanda was established. It gave a relatively short account of the genocide in its first judgment delivered on 2 September 1998 (Akeyescu ICTR 96-04) to which reference might be made. The government which emerged after the genocide has been the subject of much criticism for its human rights record. The region has been unstable for decades with large population movements. Rwanda has received hundreds of thousands of refugees from the Democratic Republic of Congo and Burundi (according them refugee status prima facie, rather than considering individual circumstances) and has worked with the UNHCR to house and support those refugees. It has also worked with the UNHCR to provide sanctuary for over 500 refugees from Libya while the UNHCR decides their asylum claims and seeks to resettle them. For all this the UNHCR gives the Government of Rwanda credit. But it is critical of the way in which individual asylum claims have been dealt with hitherto and considers that the aspirations in the agreement between the two governments are unachievable at present. They will require, in particular, changes in attitude and training which are not yet in place.

The UNHCR does not question the good faith of the Government of Rwanda. In my view, there is nothing in the materials before the Divisional Court which credibly questions that good faith or the genuine determination of Rwanda to deliver its side of the agreement reached with the British Government. Similarly, there is no suggestion that the British Government will do other than seek to ensure that Rwanda will deliver on the agreement.

At the heart of this issue, therefore, is whether despite the efforts of both Governments and the protections built into the agreement, the relevant risk of bad decision making and subsequent refoulement remains. This calls for an evaluative exercise where past relevant events inform the evaluation and must be coupled with judgements about good faith, intentions to deliver, the capacity of the system to deliver and the effect of monitoring. Good faith and intentions to deliver fall squarely within an assessment of the value of these diplomatic assurances. Capacity issues are perhaps hybrid, falling within diplomatic assessment and also more straightforward questions of fact. There are questions about whether the will of the central Government of Rwanda will be effective in dictating the approach and conduct of officials down the chain. There must also be consideration of the practicalities of removing someone from Rwanda to their home country (assuming a “wrong” asylum decision). None of the appellants has a passport and Rwanda has no arrangements in place for returns to the countries from which any of these appellants hails. If sent to Rwanda each will travel on a British document for that purpose alone.

The Divisional Court has been criticised for saying that it would need “compelling” evidence to differ from the evaluation of the British Government. I think that description is apt to describe the approach of a court to the evaluation of a diplomatic assurance. The worth of a diplomatic assurance calls for an exercise of judgement in which the government has expertise and access to advice which a court does not have. A court is not institutionally well-equipped to make such a judgement. The situation is analogous to assessments of risks to national security where a court is slow to differ from the assessment of the government: Secretary of State for the Home Department v Rehman [2003] 1 AC 153, [50]; R (on the application of Begum) v Special Immigration Appeals Commission [2021] AC 765, [70] to [71]. But where the assessment of future conduct engages practical considerations which arise from past conduct the position is not as stark. In this case there is very detailed evidence of the way in which the Rwandan asylum system has operated when considering individual claims before the summer of 2022. There were undoubted deficiencies. Whether they are capable of being made good is not an issue on which the government has special institutional expertise.

This evaluative exercise of assessing future risk does not require the resolution of all conflicts of evidence between the Government of Rwanda and the UNHCR in the sense of deciding on the balance of probability whether something did or did not happen. That is in any event impractical in the context of this litigation for at least two reasons. There is material before the court from Rwanda provided to the British Government in answer to questions designed to counter concerns raised by the UNHCR. The Government of Rwanda is not a party to these proceedings nor, diplomatically, could it be expected to engage as if it were a litigant. Moreover, there is no practical way to test the evidence in these proceedings, still less to explore ambiguities in language and the like which were drawn to our attention. But it would not be the correct approach when evaluating a future risk of this nature.

The approach to evaluating the ultimate relevant future risk, which is of refoulement, is analogous, but not identical, to the evaluation of the risk under consideration in Rehman. That case concerned the evaluation of whether a person’s presence in the United Kingdom constituted a risk to national security. Such an evaluation did not depend upon a point-by-point consideration of past events by reference to a standard of proof: Lord Hoffmann at [55].

Before turning to the issues which arise in this appeal it is, in my view, necessary to consider what the Divisional Court (and in turn this court) was being asked to determine on the “safe country” issue. Mr Husain KC, who appeared as the principal advocate for the appellants on this issue, recognised that the proposition being advanced on their behalf was that it was not safe for anybody to be sent to Rwanda. All would face the relevant risks irrespective of the number removed or their personal makeup. But the arguments both before us and the Divisional Court at times became confused by the introduction of issues which were hypothetical and moved far from the consideration of a single generic case or the concrete cases represented by these appellants. For example, in much of the political hyperbole which surrounded the announcement of the Rwanda policy there was talk of Rwanda, within a few years, being a destination for thousands of asylum seekers who arrived irregularly in the United Kingdom. The UNHCR evidence questions whether Rwanda can cope with the volumes apparently contemplated. Yet the evidence before the Divisional Court was that the physical capacity for housing asylum seekers in Rwanda was limited to 100; that of the 47 originally identified for removal the Home Office expected in fact to remove about 10; and that the starting point for any removal under the agreement was for the two governments to agree who would be sent to Rwanda. That would be determined by the capacity of the Government of Rwanda to receive and process the individuals concerned. It also gave the Government of Rwanda complete control so that they might reject any proposed name.

The Divisional Court was not considering whether it is “safe” for Rwanda immediately to receive substantial numbers. Similarly, the voluminous papers in this case identify hypothetical special problems it is said that some groups of people would face. But we are not considering whether it would be “safe” for every conceivable type of person to be sent to Rwanda. For example, the UNHCR have provided evidence which suggests that were nationals of an unnamed country with which Rwanda has close relations to seek asylum it is unlikely they would ever receive it. It is not difficult to deduce the identity of that country. Were the British Government unwise enough to seek to remove any such nationals to Rwanda, and were Rwanda improbably to agree to accept them, they would have strong legal grounds to resist. The UNHCR also raised concerns about gay and lesbian asylum seekers in Rwanda. We are not concerned, and nor was the Divisional Court, to determine whether, hypothetically, there may be individuals bearing particular characteristics who would face the relevant risks in Rwanda.

The argument of the appellants before the Divisional Court was, that in respect of each of them, there were substantial grounds for believing that there was a real risk that they would be returned to their home countries after a wrong refusal of asylum in Rwanda. They also suggest a relevant risk of article 3 ill-treatment in Rwanda itself. They submit that the same risks would attach to anyone sent to Rwanda irrespective of any personal characteristics. That encapsulates the core safety issue in these proceedings.

There is a multitude of grounds of appeal on which permission to appeal has been granted and one (concerning data protection) where the Divisional Court’s refusal to grant permission to apply for judicial review is the subject of an application for permission to appeal. They break down into the following categories:

I have had the advantage of reading in draft the judgment of Underhill LJ and agree, for the reasons he gives, in respect of issues (vi) and (viii) to (xii) as I have identified them above. I shall not burden this judgment with any further discussion of those issues. On several issues, however, it is my misfortune to differ in my conclusion from both the Master of the Rolls and Underhill LJ.

First, on whether the Divisional Court applied the wrong test when considering the safety of Rwanda on the refoulement issue. Secondly, on whether there are substantial grounds for believing that an asylum seeker sent to Rwanda would face a real risk of refoulement following a flawed decision. Thirdly, with the Master of the Rolls, on whether there are substantial grounds for believing that a removed asylum seeker would be at real risk of article 3 ill-treatment in Rwanda. In my view, the Divisional Court did not err in the way suggested and the relevant risks are not established on the evidence. It follows that I do not consider that the underlying policy is unlawful in a Gillick sense. Moreover, I agree with the Divisional Court that the posited removals, and the underlying policy, are not unlawful for want of investigation either in accordance with Ilias or Tameside.

On issue (vii), the Master of the Rolls and Underhill LJ have concluded that the Secretary of State was wrong to certify the individual claims on the basis that Rwanda is a safe third county. Their conclusion followed inevitably from their ruling that Rwanda is not a safe third country. Despite having reached the contrary conclusion on the central issue of safety I nonetheless agree that these claims should not have been certified. The whole question of safety, as our judgments demonstrate, is contestable. For the reasons given by Underhill LJ at [130] I agree that this conclusion does not affect the outcome of these appeals.

The first issue arises from grounds of appeal that suggest that the Divisional Court made a fundamental error in its approach to the article 3 ECHR question. It failed to appreciate that it was for the court to make the judgement about whether there are substantial grounds for believing that the claimants would be at real risk of treatment contrary to article 3 through being returned to their countries of origin having been wrongly refused asylum. Instead, it is argued that the court applied the conventional domestic public law test by asking whether the Home Secretary was entitled to come to the conclusion that Rwanda was safe for these purposes. As Lewis LJ observed when the point was raised at the hearing which dealt with permission to appeal:

“…paragraph 45 of the judgment said that the issue was whether the [Home Secretary] could lawfully reach the conclusion that the arrangements governing relation to Rwanda would not give rise to a risk of refoulement. The [Home Secretary] could only do that if there was no risk. That is the issue the court then considered from paragraphs 46 to 71.”

In that passage Lewis LJ was using “risk of refoulement” as shorthand for “substantial grounds for believing there is a real risk”.

It would indeed be remarkable if the Divisional Court failed to appreciate that its function, when considering article 3 risks (both arising from refoulement and conditions in Rwanda itself), was to make an assessment for itself. It could properly be described as the most basic of errors in an ECHR based claim. It would be all the more remarkable given the composition of the court. In my view, a reading of the judgment dealing with these issues as a whole demonstrates that no such error was made.

In the introductory section of its judgment at [4] the court referred to section 6 of the Human Rights Act 1998. That requires public authorities to act compatibly with the ECHR. It is axiomatic that where section 6(1) of the Human Rights Act applies, it is unlawful for the Home Secretary to act in a way which is incompatible with an ECHR right. In judicial review proceedings challenging a removal decision on the basis that it is contrary to section 6 because the claimant will be subjected to a violation of his article 3 ECHR rights, it is for the court to determine whether or not the decision would result in such a violation. The court will have due regard to the evaluation of the decision maker and be sensitive to matters such as institutional competence in making evaluative judgements about the future where that feature is present. The reference to section 6 suggests that the Divisional Court was approaching the issue on the correct basis.

Having set out the factual and legal background the court sought to identify the issues it was required to determine. It was faced with a plethora of disparate and overlong pleadings and skeleton arguments. The parties had themselves been unable to agree a coherent list of issues. At [39] the court distilled the issues to 12. The first issue began:

“The Home Secretary’s conclusion that Rwanda is a safe third country is legally flawed. The Claimants’ primary contention is that this assessment is contrary to article 3 of the ECHR. This rests on: (a) the decision of the [Strasbourg Court] in [Ilias] that a state cannot remove an individual asylum-seeker without determining his asylum claim unless it is established that there are adequate procedures in place in the country to which he is to be removed that will ensure that the individual’s asylum claim is properly determined and he does not face a risk of refoulement to his country of origin; (b) the submission that the removal of the individual Claimants to Rwanda will put them at real risk of article 3 ill-treatment (in breach of the principle recognised in Soering…) and (c) the contention that, systemically, it is inevitable that the policy to remove asylum claimants will lead to occasions when a person will be subjected to article 3 ill-treatment.”

The article 3 issue is correctly stated in points (a) and (b). The third point is the Gillick issue. The court then added that the same points were argued on conventional judicial review grounds. It continued, in summarising issue 2, to note that the central contention was that the asylum claims would not be determined effectively thereby running the risk of refoulement, directly or indirectly. It recorded the way in which the case was put, namely that the Home Secretary could not have confidence that Rwanda would comply with the agreement reached or abide by its assurances.

The court returned to these issues at [43] to [45]. At [43] it repeated the claimants’ primary submission that the Home Secretary’s conclusion that Rwanda is a safe country was legally flawed. That was put by the claimants in several ways. First, it “amounts to a breach of article 3 … for the reasons explained” in Ilias “namely that the asylum claims … would not be effectively determined in Rwanda and the asylum claimants run a risk that they will be refouled directly or indirectly…”; secondly, that the Home Secretary failed to comply with the Tameside duty and made her decision on material errors of fact; thirdly that the decision to treat Rwanda as a safe third country was irrational; and fourthly that the Rwanda policy was unlawful in the sense explained in Gillick “in that it positively authorises or approves removals that would be in breach of article 3 … (i.e. exposes persons to a real risk of article 3 ill-treatment).”

In this paragraph the court is referring to two different types of argument - one relying upon article 3 and the other upon conventional public law principles. In [44] the court noted that it was also argued that the claimants would face a real risk of ill-treatment in Rwanda and that to remove them there “would be in breach of article 3 … in the sense of the Soering principle because there are reasonable grounds for believing that if a person is removed to Rwanda that will expose him to a real risk of article 3 ill-treatment because of the conditions in Rwanda.” In [45] the court repeated the formula whether the “Home Secretary could lawfully reach the conclusion that the arrangement governing relocation to Rwanda would not give rise to a real risk of refoulement or other ill-treatment contrary to article 3.”

The way in which the case was argued before the Divisional Court, as indeed it was before us, focused on Ilias not simply for the proposition that if there were substantial grounds for believing that there would be a real risk of refoulement then removal to Rwanda would breach article 3. Ilias is also relied upon to support the submission that there is a free-standing procedural obligation of investigation under article 3 ECHR which, if not satisfied, would render removal unlawful in article 3 terms even if it could be shown that such reasonable grounds for belief did not in fact exist.

The court dealt with questions of investigation, whether under article 3 or Tameside, risk of refoulement and adequacy of the Rwandan asylum system (in the light of the Migration and Economic Development Partnership) under a single heading: Was the assessment that Rwanda is a safe third country legally flawed? That entailed resolving issues both by reference to the ECHR and section 6 of the Human Rights Act as well as applying conventional public law principles. Having considered the question of investigation and inquiries ([46] to [61]) the court turned to the adequacy of the Rwandan asylum system. The first sentence in [62] is criticised:

“Next we consider whether the Home Secretary was entitled to conclude that there were sufficient guarantees to ensure that asylum seekers relocated to Rwanda would have their asylum claims properly determined there and did not run a risk of refoulement in accordance with the obligations in Ilias and that Rwanda was a safe third country in accordance with the criteria in paragraph 345B (ii) to (iv) of the Immigration Rules.”

That long sentence is capable of being read in different ways with the words “entitled to conclude” governing all that follows or only part of it. Additional punctuation would have assisted. But the obligations on the state identified in Ilias do not depend upon a Government forming a tenable view, but a correct view. They have an objective element. Moreover, Rule 345B, which is designed to ensure compliance with the ECHR and Refugee Convention, is couched in terms of the objective establishment of various criteria, and not qualified by “if the Secretary of State is of the opinion” or similar words. In reading this part of the judgment it should not be forgotten that the court was considering both article 3 and conventional public law challenges in tandem because that is the way they were argued. The term “legally flawed” covered both. Moreover, this paragraph must be read in the context of what has gone before. The Divisional Court then stated its conclusion at [71]. The court discussed the status of evidence from the UNHCR and continued:

“We must consider it together with all the evidence before us and decide whether, on the totality of that evidence, the Home Secretary's opinion is undermined to the extent that it can be said to be legally flawed. For the reasons we have already given, the Home Secretary did not act unlawfully when reaching the conclusion that the assurances provided by Rwanda in the MOU and Notes Verbales could be relied on. That being so the conclusion that, for the purposes of the criteria at paragraph 345B (ii) to (iv) of the Immigration Rules, Rwanda is a safe third country, was neither irrational nor a breach of article 3 of the ECHR in the sense explained in Ilias.”

The relevance of the assurances (along with monitoring arrangements) was that they were said by the Home Secretary to mitigate such risk as there was that there might be refoulement. Her case was that substantial grounds for believing that there was a real risk were not present. The adequacy of the assurances was attacked by the claimants and the UNHCR. Her conclusion would be legally flawed if such a real risk was present despite the assurances. The criteria in rule 345B referred to in [71] are: (ii) that the principle of non-refoulement will be respected in accordance with the Refugee Convention; (iii) that the prohibition on removal, in violation of the right to freedom from torture and cruel, inhuman, or degrading treatment … is respected in that country; and (iv) that the possibility exists to request refugee status and to receive protection in accordance with the Refugee Convention. The breach of article 3 explained in Ilias was removal to a third country in which there were substantial grounds for believing that the asylum seeker would face a real risk of ill-treatment or of being subjected to refoulement. There would be a breach of article 3 if either real risk existed on the evidence. The Divisional Court had identified the relevant refoulement risk at [9] of its own judgment and referred to [134] in the judgment of the Strasbourg Court in Ilias:

“The Court would add that in all cases of removal of an asylum seeker from a Contracting State to a third intermediary country without examination of the asylum requests on the merits, regardless of whether the receiving third country is an EU Member State or not or whether it is a State Party to the Convention or not, it is the duty of the removing State to examine thoroughly the question whether or not there is a real risk of the asylum seeker being denied access, in the receiving third country, to an adequate asylum procedure, protecting him or her against refoulement. If it is established that the existing guarantees in this regard are insufficient, Article 3 implies a duty that the asylum seekers should not be removed to the third country concerned.”

The Home Secretary’s decision could only be lawful if such a real risk did not exist. I would add that in [72], when dealing with the Gillick issue, the Divisional Court added that if the relevant criteria under rule 345A to C were met “removal to that country will not, applying the principles in Ilias (themselves a particular application of the principle in Soering), give rise to a breach of article 3 of the ECHR.”

The court went on to consider “conditions in Rwanda generally” from [73] which it described as the “wider Soering submissions, that persons removed to Rwanda … are exposed to a real risk of article 3 ill-treatment not for any reason connected with the handling of their asylum claim but by reason of conditions in Rwanda, generally.” This part of the judgment straightforwardly considers the evidence and competing arguments and concludes that there is no such real risk. That approach reinforces the reality that in considering the central article 3 issue, both by reference to the risk of refoulement and treatment in Rwanda itself, the Divisional Court applied the right test. With respect to the Divisional Court, I accept that its use of language (“entitled to conclude” etc. e.g. in [64]) in a discussion of issues that raise both ECHR and public law points of law has generated some confusion but when read as a whole, I am not persuaded that this criticism of the judgment is made out.

This issue is more than a technical one. It feeds into the role of an appellate court in a case where the first instance court has made an assessment on the basis of all the evidence on the question of whether the action under scrutiny would breach the ECHR and be unlawful by virtue of section 6 of the Human Rights Act 1998. We have been presented with thousands of pages of materials (leaving aside the superabundance of authorities – most not referred to) said to be of relevance to the question whether Rwanda is a safe third country for article 3 purposes. If the Divisional Court applied the wrong test, it would be open to this court to allow the appeal on that basis and remit the matter for fresh consideration at first instance. Alternatively, this court could undertake the evaluation. I have the misfortune to disagree with the Master of the Rolls and Underhill LJ on this issue. Inordinate delay would be caused by remitting the refoulement article 3 issue to the High Court and so I shall proceed to consider the issue as if sitting at first instance, having considered for myself all the evidence. I am grateful to Underhill LJ for his comprehensive review of the evidence which informs this issue and will express my conclusions relatively briefly.

The Strasbourg Court has explained that when considering whether substantial grounds have been shown that an individual would face a real risk of treatment contrary to article 3 that the analysis of the evidence said to support that conclusion must be “rigorous”: see e.g. Chahal v. United Kingdom (1997) 23 EHRR 413 at [96]; Saadi v. Italy (2009) 49 EHRR 30 at [128]; Sufi v. United Kingdom (2012) 54 EHRR 9 at [214]. For example, in cases where the argument rests on assertions of general violence in a country the court has made clear that not every situation of general violence will give rise to such a risk. A general situation of violence would only be of sufficient intensity to create such a risk “in the most extreme cases” where there was a real risk of ill-treatment simply by virtue of an individual being exposed to such violence on return: NA v. United Kingdom (2009) 48 E.H.R.R. 15 at [115]. The assessment of risk in an article 3 exercise requires careful consideration of all the evidence in the context of the position in which removed persons will find themselves.

Ilias establishes that in cases where an ECHR state removes an asylum seeker to another state without considering the merits of an application, the removing state has a duty to examine thoroughly the adequacy of the procedures in the receiving country to determine whether they protect against refoulement. The Strasbourg Court was concerned to determine whether the applicant was adequately protected against removal, directly or indirectly, to his or her country of origin in circumstances where article 3 risks had not been properly evaluated: see [130] to [138].

The article 3 question boils down to whether substantial grounds have been shown for believing that there is a real risk of refoulement, directly or indirectly, to a country in which the applicant in fact needs protection, because of deficiencies in the asylum processes in the third country, here Rwanda.

To answer that question, it is helpful to consider what would happen to individuals identified for removal by the Home Office and accepted by Rwanda. There is no reason to suppose that the practical and purely administrative steps agreed between the governments will not be followed.

Those removed will arrive at Kigali and be accommodated at the Hope Hostel where they will be free to come and go. That has a capacity of 100. The evidence describes plans for further sites to be identified. According to the reception and accommodation Note Verbale they will be provided with mobile telephones with internet access. They will be given a temporary residence permit for three months on arrival. That will be extended if the asylum claim is not completed within three months. Those removed to Rwanda will be provided with financial support by the British Government in Rwanda at the same level they would receive in the United Kingdom during the asylum process, and thereafter, for a total of five years if granted asylum and up to three years if not. Those who make an asylum claim will be interviewed by the Directorate General of Immigration and Emigration (“DGIE”). The asylum claim would then be considered by the Refugee Status Determination Committee (“RSDC”). If the claim is refused there is a right of appeal to the relevant ministry (“MINEMA”). If that appeal fails, there is a further right of appeal to the High Court and from there to the Appeal Court. Only on the hypothesis that an asylum claim has failed at all four stages will the question of removal arise.

In the event of a refusal of protection under the 1951 Convention the agreement requires Rwanda to consider whether there are other humanitarian grounds which preclude removal to the person’s country of origin. It also requires Rwanda to consider any application from a failed asylum seeker to remain in Rwanda on any other basis. Only then does removal become a possibility.

Para 10.4 of the MoU provides that “Rwanda will only remove such a person to a country in which they have a right to reside. If there is no prospect of such removal occurring for any reason Rwanda will regularise that person’s immigration status in Rwanda.” These are important provisions which significantly reduce the prospects of refoulement. There is other evidence about the prospects of removal. In practical terms forcible removal from Rwanda to a failed asylum seeker’s country of origin is possible only if that country is willing to accept such returns. Therein lies a difficulty for all governments. The unequivocal evidence is that Rwanda has no agreements for return with any country material for these purposes. There is other evidence suggesting that removal is unlikely which is set out in the judgment of Underhill LJ. I do not ignore the evidence from the UNHCR that there have been instances of “airport refoulement” (which is not a risk in these cases). Rwanda immediately returned a Syrian to Turkey and an Afghan to Dubai. The evidence suggests that from there they were sent to their countries of origin. Nor do I overlook the evidence that, in different contexts, people have been pushed over the border by the DGIE into Tanzania or Uganda. The circumstances were very different.

Objection is taken that the practical likelihood of refoulement was not addressed as a separate issue by the Divisional Court and referred to before us only in the skeleton argument of the respondent without oral elaboration by Sir James Eadie KC. The way in which it was dealt with by the court below and in argument should not lead to the conclusion that a relevant consideration in the overall evaluation of the risk of refoulement should artificially be left to one side. In the scheme of determining whether a proposed course of action amounts to a breach of section 6 of the Human Rights Act (and therefore the ECHR itself) the court making the decision must consider all the evidence before it. The Divisional Court reposed confidence in the MoU (it quoted para 10.4) and the monitoring arrangements. The evidence of the practicality of removal, with which para 10.4 is concerned, was before it. Part of the Secretary of State’s case before the Divisional Court was that there will be little practical chance of removal in any of these cases. In agreement with Underhill LJ, I would not ignore this evidence despite there being no respondent’s notice in respect of it. My conclusion, having regard to all the evidence, is that the risk of refoulement of a failed asylum seeker sent by the United Kingdom to Rwanda is low and that the assessment of this evidence is relevant to determining the overall evaluation of whether substantial grounds for believing there is a real risk of refoulement have been established.

The UNHCR has provided cogent evidence that various individuals or groups of people have been denied access to the Rwandan asylum process at the entry stage. Individuals have been turned away at the airport; families in Rwanda have been denied the opportunity to make a claim; and large numbers who travelled from Israel to Rwanda under an agreement (the details of which are unknown) did not have their claims properly assessed. They were either pushed across the border into neighbouring countries or left Rwanda and travelled to Europe. None of this, troubling though it is, suggests that anyone sent from the United Kingdom to Rwanda is at real risk of similar treatment. The passage of each individual would be agreed in advance. They would be met on arrival at Kigali and would be expected to make asylum claims. Their journey through the asylum process and beyond would be monitored. Underhill LJ has analysed the evidence relating to the nature of the interview that can be expected to be conducted by the DGIE, the involvement of an eligibility officer and the early stages of engagement in the asylum process in Rwanda. The RSDC acting on the fruits of the interview, further country information and possibly personal appearance of the person in question will make its decision. I share the concerns identified by the UNHCR about whether those involved in the RSDC have sufficient training and expertise to deal appropriately with asylum claims and also whether what are reported as ingrained attitudes of scepticism towards claims made by Middle Eastern nationals will be influential. There is certainly evidence of poor practice. There will, no doubt, be changes in respect of those considered under the agreement with the United Kingdom. But the question is whether the system as a whole can be relied upon to deliver appropriate outcomes.

To my mind an important factor in answering that question is whether the monitoring arrangements, both formal and informal, provide sufficient protection to drive good decision making and thus to reduce the risk of refoulement below the level that would give rise to a breach of article 3 by the United Kingdom in sending people to Rwanda.

The Strasbourg Court has recognised the importance of effective monitoring arrangements when considering assurances in support of the removal of a named individual to a country where, in the absence of the assurances, there is every reason to suppose that article 3 standards will not be met there. The principles were drawn together in Othman v. United Kingdom (2012) 55 EHRR 1 at [89]. Othman (Abu Qatada) had been convicted of terrorism in Jordan in absentia and was to be removed to Jordan on the strength of assurances from the Jordanian Government that he would not be ill-treated and would get a fair retrial. The approach of the court does not read over precisely to generic, rather than person-specific, assurances. Nonetheless, the purpose of the list of factors set out by the Strasbourg Court was to focus attention on whether the assurances would be effective. Monitoring compliance with the agreement is an important factor.

It is reasonable to assume that individuals who find themselves in Rwanda would be familiar with the agreement reached between the governments. With the assistance of lawyers in England, those unwilling to be removed to Rwanda would have been engaged in resisting on all available grounds. In referring to “informal” monitoring I have in mind 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. It is probable that they will be in contact with family and friends, their English lawyers, the British High Commission and, indeed, the UNHCR in Rwanda. The UNHCR would be expected to pay close attention to what was happening on the ground despite having no formal role in the monitoring arrangements. It has done that in respect of refugees generally in Rwanda beyond the aspects where it has its own agreements with the government. Those sent from the United Kingdom will be housed with other asylum seekers. One way or another, shortcomings in the provision of interviews, transcripts, interpreters, lawyers, reasons for decisions etc. in accordance with the agreement would readily come to light with a good chance of their being dealt with.

Importantly, the formal monitoring provided both by the agreement and by arrangements put in place in the British High Commission in Kigali would also do so.

The first formal part of the monitoring arrangements involves the British High Commission in Kigali with Home Office officials embedded there to monitor compliance with the agreement. Finnlo Crellin was posted to the High Commission in Kigali as Home Office liaison officer. The role, which will be a permanent one for the duration of the agreement, involves developing relationships with players in all parts of the system in Rwanda to flag concerns and to make the agreement work. As he puts it, “the extent of these relationships between the respective Governments is key to the strength and collaborative nature of the [agreement], allowing both sides to assess progress, discuss specific issues or flag any concerns – including around implementation of the assurances in the MoU and [Notes Verbale] – and to resolve these effectively.” Kristian Armstrong, a senior Home Office official, explains that the Home Office officials in Kigali have the right under the MoU to observe any stage of the asylum process. This enables the United Kingdom to monitor, on a constant basis, that the assurances are being met and the system in Rwanda is working. It also provides accountability by the Government of Rwanda to the United Kingdom for the assurances given under the agreement. He adds that there is an agreement that Rwanda will provide a quarterly report to the United Kingdom on the outcome of each asylum claim and appeal, the status of each relocated individual in Rwanda and details of anyone who has left or been removed from Rwanda. The agreement makes provision for complaints which adds another safeguard.

The agreement also makes provision for independent monitoring. The MoU provides for an independent Monitoring Committee to which each government nominates four members, operating independently of the governments. Those nominations were made and the Terms of Reference of the Monitoring Committee agreed. The MoU ensures unfettered access by the Monitoring Committee to relevant records, officials and facilities. The Monitoring Committee is designed to ensure that there are frequent independent and authoritative reports on how all parts of the system are performing in Rwanda. It will provide reassurance to the British Government that relocation of individuals to Rwanda is compatible with the 1951 Convention and the ECHR. In my view this monitoring arrangement will provide significant assurance that the agreement is being abided by, pick up problems and enable any that develop to be dealt with.

The next level of monitoring which the agreement establishes is the Joint Committee. The MoU provides for the agreement to be closely managed. The Joint Committee is made up of senior officials from the United Kingdom and Rwanda. It first met in Kigali on 31 May 2022 to discuss preparations for the initial flight and the work to ensure that assurances contained in the agreement are implemented. Its role is to provide direction and manage the implementation of the commitments set out in the MoU. The United Kingdom representatives on this committee are the British High Commissioner in Rwanda, a Senior Operational Director from Immigration Enforcement, a senior member of Home Office Legal Advisors and Mr Armstrong. The Rwandan members include members from the DGIE, MINEMA and the Foreign Ministry. Its terms of reference have been agreed.

A function of the Joint Committee is to discuss plans for the number of individuals to be relocated to Rwanda over the forthcoming year, with a particular focus on the immediately following quarter. This plan will be a joint effort produced with a view to ensuring that the numbers are realistic on both sides having regard to capacity to deal with the individuals and asylum claims in accordance with the agreement.

These multiple levels of protective monitoring provide powerful reassurance that the terms of the agreement will be honoured and that if there are problems they will be picked up and ameliorated. I understand the concerns of UNHCR but do not consider that the organisation is giving sufficient weight in its assessment to the strong interest of both governments to make this arrangement work, the detailed monitoring arrangements which will pick up any problems and the ability to sort them out if they arise. If significant problems arose giving rise to the relevant risk of refoulement the British Government would be unable to continue lawfully to send people to Rwanda. The reputation of Rwanda in this very high-profile public agreement is at stake. More prosaically, there are powerful financial incentives at work, described in general terms by Mr Armstrong. Not only is every cost associated with the reception and processing of asylum seekers being met by the United Kingdom and those removed to Rwanda provided with an income, but substantial sums of future aid support depend upon Rwanda’s compliance with the agreement. These factors operate in an environment of a deepening relationship between the United Kingdom and Rwanda in recent years explained in the evidence from Simon Mustard of the Foreign Commonwealth and Development Office. He does not seek to avoid confronting some of the profound human rights concerns that remain, particularly concerning the lack of tolerance for political opposition to the government of President Kagame. He expresses the confidence of the Foreign Office that the Government of Rwanda will honour its commitments.

The focus of concern of the UNHCR set out in Mr Bottinick’s evidence is on the administrative stages of the asylum process: DGIE interview; RSDC consideration of the claim and then the appeal to MINEMA. Those concerns arise particularly if Rwanda deals with substantial numbers of claims. He says very little about the Rwandan Courts beyond observing that the right of appeal to the High Court, introduced some years ago, does not yet appear to have been utilised and, so far as the UNHCR are concerned, the jurisdiction and procedures that will be applied are unclear.

The Note Verbale records that “the court will be able to conduct a full re-examination of the Relocated Individual’s claim in fact and law in accordance with Rwandan rules of court procedure.” There will be legal representation. Proceedings will be in public and an adverse decision may be appealed. The UNHCR does not suggest that the judges of the High Court and Appeal Court in Rwanda will not deal with cases that reach the courts properly. Before the Divisional Court the appellants had developed a nascent argument that the Rwandan Courts generally lack independence which was expanded before us. The submission, in short, is that the Rwandan judiciary cannot be expected to disagree with the conclusion of MINEMA, a government department.

Mr Husain relied upon the decision of the Divisional Court (Irwin LJ and Foskett J) in Government of Rwanda v. Nteziryayo [2017] EWHC 1912 (Admin) which concerned a request for extradition of a Hutu for genocide arising out of the events of 1994. The request was made in 2013, an earlier request from 2006 eventually having been unsuccessful in April 2009. The argument before the District Judge was that the requested person would not get a fair trial in Rwanda. The judgments of both the Senior District Judge at first instance and the Divisional Court explored in detail questions of judicial independence in Rwanda. They expressed concerns of a lack of independence in some types of politically charged case. Nonetheless, their conclusions did not rest on the simple proposition that the judiciary was not independent and would thus deliver a result desired by the government: see [365] to [384] of the judgment of the Divisional Court and Annex 3 citing from the judgment of the Senior District Judge. The pivotal issue was the effectiveness of the legal profession in criminal cases of that sort.

“377 …The closer we have read the evidence in this case, the firmer has become our agreement with the judge below that defence capacity is the vital element, the capstone, of the case. Whilst in the context of our court system adequate representation is of course important, other safeguards in the system such as responsible unbiased prosecution, witness protection, unchallenged and complete judicial independence taken together, mean that inadequate defence may be compensated for and a reasonable quality of justice delivered overall. Even in that context, it is well established that miscarriages of justice will occur, where defence representation is inadequate. However, in the context of Rwanda, with the difficulties and weaknesses we have identified, the presence or absence of effective defence is absolutely central. We are completely of one mind with the judge below on that point.”

The lack of independence of the Rwandan judiciary in “politically sensitive cases” was also called into question by Human Rights Watch in a letter dated 11 June 2022 to the Home Secretary, a view essentially accepted by the Foreign Office. The question, as it seems to me, is whether the government would put pressure on the courts to adhere to the administrative view and whether the courts would be influenced by that pressure or, it might be said, themselves go along with the decision come what may, without any pressure being exerted. That question must be answered in the context of the agreement between the governments which rests upon a desire, indeed determination, of Rwanda to demonstrate the integrity of its asylum system including the independence and efficacy of its High Court and Court of Appeal. The Rwandan judiciary, on the hypothesis that appeals reach the High Court or beyond, will be under detailed scrutiny, indeed international scrutiny.

Sir James Eadie KC submitted that the context of these possible appeals is very far removed from prosecutions in which the state has an interest in securing a conviction for genocide or for offences alleged against political opponents. A particular terrorism trial was referred to in evidence. He submitted that an asylum appeal against an administrative decision is of a different character and bears no obvious political dimension which would give rise to the risk of manipulation. Those are points well-made. He also referred to what I would regard as a circularity at the heart of the appellants’ submission on this point. 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. There is 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.

I have indicated my conclusion (see [501] above) that the risk of refoulement at the end of the process (including administrative and legal appeals) in cases involving individuals sent to Rwanda from the United Kingdom pursuant to the agreement is low. I am also satisfied 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. My evaluation of all the evidence, only a part of which has been referred to in the three judgments we are delivering, results in the conclusion that substantial grounds for believing that there is a real risk that deficiencies in the asylum system will lead to wrong decisions and refoulement have not been established.

The Divisional Court dealt with this issue between [73] and [79]. It identified two bases upon which it was suggested that persons removed to Rwanda would face a real risk of treatment prohibited by article 3 ECHR. The first relates to a general intolerance of political criticism and the second to events which occurred in Kiziba refugee camp in 2018 when protests about the conditions in the camp resulted in disturbances which were violently supressed by the Rwandan police and resulted in deaths. As to the second, the Divisional Court, correctly in my view, considered that it was unlikely anything similar could happen to those sent to Rwanda under the agreement. They will not be in a refugee camp. As it said at [74]:

“the treatment of transferred persons, both prior to and after determination of their asylum claims is provided for in the MoU … and in the Support NV. For the reasons already given, we consider the Rwandan authorities will abide by the terms set out in these documents … The Support NV includes (at paragraph 17) that a mechanism is to be established to allow complaints about accommodation and support provided under the MoU to be raised and addressed. Provision for those arrangements is strong support for the conclusion that the possibility of complaint on such matters, made by persons transferred under the [agreement] does not give rise to any real risk that the consequence of complaint will be Article 3 ill treatment.”

The court dealt with the wider submission concerning whether those transferred to Rwanda would be at real risk of article 3 ill-treatment because of the way the Rwandan authorities might respond to expressions of opinion adverse to them or acts of political protest. Between paragraphs [76] and [77] the Divisional Court rejected that submission for reasons with which I agree:

"76.

There is no suggestion that any of the individual Claimants … holds any political or other opinion that is adverse to the Rwandan authorities. If there were such evidence it would fall to considered under paragraph 345B(i) of the Immigration Rules. A proper application of that criterion would be sufficient to ensure that were a person to face a real risk of article 3 ill-treatment, he would not be transferred. That being so, the Claimants' case comes to the proposition that, following removal to Rwanda, it is possible that one or more of those transferred might come to hold opinions critical of the Rwandan authorities, and that possibility means that now, the Soering threshold is passed.

77.

There is evidence that opportunities for political opposition in Rwanda are very limited and closely regulated. The position is set out in the "General Human Rights in Rwanda" assessment document, one of the documents published by the Home Secretary on 9 May 2022. There are restrictions on the right of peaceful assembly, freedom of the press and freedom of speech. The Claimants submitted that this state of affairs might mean that any transfer to Rwanda would entail a breach of article 15 of the Refugee Convention (which provides that refugees must be accorded the most favourable treatment accorded to nationals in respect of non-political and non-profit-making associations and trade unions). However, we do not consider there is any force in this submission at all. Putting to one side the fact that article 15 does not extend to all rights of association, it is, in any event, a non-discrimination provision - i.e., persons protected under the Refugee Convention must not be less favourably treated than the receiving country's own citizens. There is no evidence to that effect in this case. Returning to the material covered in the Home Secretary's assessment document, there is also evidence (from a US State Department report of 2020) that political opponents have been detained in "unofficial" detention centres and that persons so detained have been subjected to torture and article 3 ill-treatment short of torture. Further, there is evidence that prisons in Rwanda are over-crowded and the conditions are very poor. Nevertheless, the Claimants' submission is speculative. It does not rest on any evidence of any presently-held opinion. There is no suggestion that any of the individual Claimants would be required to conceal presently-held political or other views. The Claimants' submission also assumes that the response of the Rwandan authorities to any opinion that may in future be held by any transferred person would (or might) involve article 3 ill-treatment. Given that the person concerned would have been transferred under the terms of the [agreement] that possibility is not a real risk. It is to be expected that the treatment to be afforded to those transferred will be kept under the review by the Monitoring Committee and the Joint Committee (each established under the MOU). Further, the advantages that accrue to the Rwandan authorities from the [agreement] provide a real incentive against any mis-treatment (whether or not reaching the standard of article 3 ill-treatment) of any transferred person.”

In my view the appellants fall short of establishing that there are substantial grounds for believing that there is a real risk that they will face treatment prohibited by article 3 ECHR in Rwanda.

Ilias concerned the removal of two asylum seekers by Hungary to Serbia without any examination of the merits of the claims. Serbia was deemed by Hungarian law to be a “safe third country” on the basis that it was a candidate member to join the European Union and was required to satisfy EU standards when considering asylum applications. The Hungarian authorities did not explore the realities of the position at all and had no special arrangements with the Serbian authorities. Between [139] and [141] the court explained that a state applying the “safe third country” concept must conduct a thorough examination of the relevant conditions and reliability of the asylum system in the third country concerned. It must carry out of its own motion an up-to-date assessment of the accessibility and functioning of the receiving country’s asylum system and the safeguards it affords in practice. The expelling State cannot merely assume that the asylum seeker will be treated in the receiving third country in conformity with the Convention standards but must first verify how the authorities of that country apply their legislation on asylum in practice: M.S.S. v. Belgium and Greece (App. no. 30696/09) at [359].

These paragraphs were relied upon by the appellants in support of the submission that in safe third country cases there is a free-standing procedural obligation to examine the receiving state’s asylum system, which the Home Secretary failed adequately to do. In consequence, submits Mr Husain KC, the proposed removals are unlawful in article 3 terms on procedural grounds.

Like the Divisional Court I accept that this case bears little resemblance to the circumstances which obtained in Ilias where the Hungarian authorities made no investigations at all into the systems to which the applicants would be exposed in Serbia. On the contrary, the British Government has explored extensively the realities for asylum seekers on the ground in Rwanda. To the extent that deficiencies in the general system of considering asylum claims have been identified, the agreement between the governments has sought to remedy them. It was submitted that further inquiries of various sorts should have been made and, in particular, the British Government should have explored the terms and effectiveness of an agreement between the Governments of Rwanda and Israel by which individuals were moved from Israel to Rwanda over a number of years.

Even if further inquiries might have been made on this or other matters, there is no question here of the British authorities simply assuming that the Rwandan asylum system was adequate. On the contrary, the realities were explored and perceived difficulties addressed. I agree with the Divisional Court the Ilias investigative duty was complied with. It is unnecessary to consider whether the Strasbourg Court was creating a truly free-standing investigative or procedural duty. Moreover, I agree with the Divisional Court that the Tameside duty was complied with essentially for the reasons it gave.

I indicated at [479] that because my conclusion is that Rwanda is a “safe third country” for article 3 purposes it follows that the various policies that enable the Home Secretary to send migrants there are not unlawful in Gillick terms. That was the view taken by the Divisional Court. At [72] the court set out how that conclusion runs against each of the various policy documents which make up the Rwanda policy:

“The next matter under this heading is the Claimants' submission that the policy by which persons whose asylum claims are held to be inadmissible may be returned to Rwanda, is Gillick unlawful.  The meaning of the judgment of the House of Lords in Gillick has been considered recently by the Supreme Court in R(A) v Secretary of State for the Home Department [2021] 1 WLR 3931. The Supreme Court emphasised that the relevant question is whether the policy under consideration positively authorises or approves unlawful conduct (in the present context, a removal decision in breach of ECHR article 3). Against this standard the Inadmissibility Policy, which includes the possibility of removal to a safe third country, is not unlawful.  Removal decisions depend on the application of paragraph 345B of the Immigration Rules, and the conclusion reached against the criteria in that paragraph that the country concerned is a "safe third country for the particular applicant".  If the relevant criteria are met (see above at paragraph 11), removal to that country will not, applying the principles in Ilias (themselves, a particular application of the principle in Soering), give rise to a breach of article 3 of the ECHR.  Even if the scope of the policy for this purpose is extended to cover the general conclusion in the 9 May 2022 assessment documents and the conclusion reached following consideration of the further evidence filed in these proceedings by the UNHCR, the position remains the same. The conclusion, based on all that material, that generally, asylum claims made in Rwanda by persons transferred pursuant to the terms of the MOU would be entertained and effectively determined was a lawful conclusion.  And, in any event the final decision on removal would also have to take account of the asylum claimant's personal circumstances - i.e., the criterion at paragraph 345B(i) of the Immigration Rules.”

The Gillick question, reaffirmed by the Supreme Court in the A case referred to by the Divisional Court, requires that where the question is whether a policy is unlawful, “that issue must be addressed looking at whether the policy can be operated in a lawful way or whether it imposes requirements which mean that it can be seen at the outset that a material and identifiable number of cases will be dealt with in an unlawful way” (A at [63]). The policy (reflected in the materials to which the Divisional Court referred) seeks to ensure that article 3 ECHR is not violated when individuals are removed to Rwanda. Taken as a whole that is the aim of the policy. As Underhill LJ explains, if the policy in fact exposes individuals to the material risk of ill-treatment it fails not only because the decisions taken under it would be unlawful for article 3 reasons; it would also be unlawful in Gillick terms as (necessarily) authorising unlawful action. But if I am right in my conclusion on the central issue that is not the case. For completeness I should add that the decision of the House of Lords in R (Munjaz) v. Mersey Care NHS Trust [2006] 2 AC 148, to which Ms Naik KC drew our attention, leads to no different conclusion: see the A case at [79].

The central question in these appeals is whether there are substantial grounds for believing that removal of these appellants and any individual to Rwanda pursuant to the agreement with the Government of Rwanda will give rise to a real risk of treatment contrary to article 3 ECHR either (a) as a result of deficiencies in the asylum system with a consequent real risk of refoulement or (b) in Rwanda itself. My conclusion accords with that of the Divisional Court. The evidence taken as a whole does not support such a real risk in either case. I agree with Underhill LJ and Sir Geoffrey Vos MR that the other grounds fail. Our different conclusions on the central issue deliver a different conclusion on the Gillick issue (which adds nothing to the central question). I also agree with the Divisional Court on the procedural issues (Ilias and Tameside). In the result, I would dismiss the appeals.