Personal Immigration
Business Immigration

Court of Appeal Finds Rwanda Policy Unlawful

In a judgment published on 29 June 2023, the Court of Appeal allowed by a majority the appeal against the decision of the Divisional Court in the judicial review challenge to the Rwanda policy. The appeal was against the Divisional Court’s rejection of the generic challenges to the policy. Most of the Appellants (then Claimants) had been successful in challenging the lawfulness of the decisions certifying their human rights claims on individual grounds and the Secretary of State for the Home Department (“SSHD”) did not appeal against the Divisional Court’s decision on this matter.


Sir Geoffry Vos, Master of the Rolls (“MoR”) and Underhill LJ both concluded that the Divisional Court had erred in its approach to the consideration of whether there were substantial grounds for believing that the asylum seekers removed to Rwanda would be subjected to a real risk of treatment contrary to article 3 of the European Convention on Human Rights, i.e. the test established by the Strasbourg Court in Soering v. United Kingdom (1989) 11 E.H.R.R. 439 (Soering). The Lord Burnett of Maldon, Lord Chief Justice of England and Wales (“LCJ”) dissented, finding no error in the Divisional Court’s decision. In view of the majority finding, the Court of Appeal proceeded to consider the matter for themselves, with reference to the relevant evidence. Underhill LJ and the MoR held by majority that there were substantial grounds for believing there were real risks of article 3 ECHR breaches in accordance with the Soering test. The LCJ reached the opposite conclusion. 

The Court of Appeal further considered the Divisional Court’s decision on the remaining generic grounds that had been raised by the Claimants at the judicial review stage. Given their finding on the Soering issue, it was considered unnecessary by the majority to determine the issues relating to the rationality of the SSHD’s decision on whether Rwanda was a safe country, or the adequacy of the SSHD’s enquiries on this matter by reference to the procedural requirement of  Ilias v. Hungary (2020) 71 E.H.R.R. 6 (Ilias) and the test in Secretary of State for Education and Science v. Tameside Metropolitan Borough Council [1977] AC 1014 (Tameside). Whilst the MoR thought it unnecessary to determine this matter, he stated he would have agreed with Underhill LJ’s judgment that the Rwanda policy was unlawful per the test in Gillick v. West Norfolk and Wisbech AHA [1986] AC 112 (Gillick), as it positively authorised or approved removals that would involve a breach of article 3 ECHR. In view of his dissenting judgment on the Soering issue, the LCJ held that the Rwanda policy was not unlawful in Gillick terms and that the Divisional Court had been right to conclude that the SSHD had complied with her Ilias (investigative) and Tameside duties.

Similarly, having found that Rwanda was not a safe country, MoR and Underhill LJ held that the certifications of the asylum and human rights claims under paragraphs 17(c) and 19(c) of  Schedule 3 to the Asylum and Immigration (Treatment of Claimants etc) Act 2004 (“2004 Act”) were unlawful, separately to the individual findings of unlawfulness by the Divisional Court, which had not been challenged by the SSHD. The MoR noted that the certification test is stricter than the (“substantial grounds”) Soering test. Despite his dissent to the majority finding that Rwanda was not a safe third country, the LCJ agreed that the asylum and human rights claims should not have been certified, given that the question of safety is contestable.

The Court of Appeal unanimously rejected the Appellants’ remaining grounds. It was held that the Divisional Court had correctly decided that the removal of asylum seekers to Rwanda was not inconsistent with the Refugee Convention; that the Procedures Directive no longer forms part of UK law, and therefore the inconsistency between its requirements and the Rwanda policy does not render the latter unlawful; that there was no presumption of Rwanda’s safety by the SSHD in circumvention of the statutory framework of the 2004 Act; and that any alleged breaches of the UK GDPR would not, even if established, have invalidated the SSHD’s decisions made on the basis of the Rwanda policy. The last issue was considered for permission to appeal on a rolled-up basis and permission was refused for lack of any real prospect of success.

Finally, it was also held unanimously that the Divisional Court was correct in concluding that the procedure leading to the decisions that an individual should be relocated to Rwanda was not procedurally unfair in a systemic way. Whilst Underhill LJ disagreed with the reasoning of the Divisional Court that fairness did not require providing asylum claimants with an opportunity to make effective representations about the safety of Rwanda, or that it did not require in any circumstances access to legal assistance to enable these representations, this did not render its overall conclusion wrong. The imposition of an adequate, in principle, “base-line” seven-day timeframe coupled with the additional flexibility of the SSHD’s policy to grant extensions where appropriate did not justify a declaration that the process was inherently unfair.

The Divisional Court’s consideration of the Soering question

In view of its appellate jurisdiction, the Court of Appeal had to consider if the Divisional Court had erred in its consideration of the Soering question. Both the MoR and Underhill LJ held that the Divisional Court did not ask itself the right question, with the latter also concluding that, even if it had, its answer was wrong. It was held that an objective reading of the judgment showed that the Divisional Court only considered the Soering test in relation to conditions in Rwanda generally, at [73]-[77] of its decision. Conversely, in relation to other matters such as the adequacy of the asylum system in Rwanda and, by extension, the risk of refoulement, the Divisional Court had instead asked whether the SSHD had been entitled to reach the conclusions she did. In view of section 6 of the Human Rights Act 1998, the Soering question requires the court to assess for itself whether there are substantial grounds for believing that there was a real risk of article 3 mistreatment to asylum seekers sent to Rwanda, whether due to the conditions there or the risk of refoulement, rather than to review the SSHD’s assessment on this matter on public law grounds.

The LCJ disagreed. He considered that the duty identified in Ilias, which the Divisional Court turned to at [62] of its judgment, is not solely procedural and does not simply depend upon a government forming a lawful view, but a correct view on the safety of the third country in question. Therefore, and based on how the grounds had been argued before it, the Divisional Court was considering both the article 3 considerations and conventional public law challenges in tandem, under the rubric of whether the SSHD’s decisions were “legally flawed”, which covered both elements. Essentially, the LCJ considered that the SSHD’s decision would only have been lawful if a real risk of article 3 breaches did not in fact exist; as such, the criticism of the wording used by the Divisional Court in some paragraphs, taken together with the instances where it clearly identified the correct approach, albeit somewhat confusing, does not justify a finding that the wrong test was applied.

As the majority concluded that the Divisional Court had erred in its approach to the Soering question (or in fact answered it wrongly, per Underhill LJ), the Court of Appeal had to consider for itself the Soering question. This included the LCJ, despite his dissenting judgment on the Divisional Court’s approach.

The Court of Appeal’s consideration of the Soering question

In considering this matter, the MoR explained that great weight had to be given to the UK Government’s view of Rwanda’s diplomatic assurances. With that in mind, following Othman v. United Kingdom (2012) 55 E.H.R.R. 1 (Othman), the quality of the assurances given and whether, in light of Rwanda’s practices, they can be relied on, had to be assessed. Particular importance should be attached to the evidence and opinions from UNHCR, although the evidence is not pre-eminent. The MoR held that the fact that the structural institutions that gave rise to past violations remain in Rwanda did not lend support to the SSHD’s view that the assurances can obviate the real risk of article 3 violations. He concluded that, upon giving appropriate but not overriding weight to the UNHCR’s evidence and weighing it against the Rwandan assurances and the SSHD’s view, substantial grounds for believing that there is a real risk of article 3 breaches existed.

Underhill LJ first made three preliminary observations, before proceeding to consider the criticisms directed at the 5 particular stages of the Rwandan asylum system and then 3 overall issues with it: access to legal assistance/representation; availability of interpreter services; and training. The preliminary observations related to the need to consider both how the Rwandan asylum system has operated thus far and the nature and likelihood of implementation of any proposed changes (also observed by the MoR); the fact that the bulk of asylum grants in Rwanda had not involved individual assessment of the claimants and the 2020 UNHCR review, which was complimentary about Rwanda’s compliance with the Refugee Convention, was not directed to the “refugee status determination” (“RSD”) process; and that an overview of the totality of the evidence was required to resolve factual disputes.

Following his assessment of the evidence in the aforementioned structure, Underhill LJ stated that, despite Rwanda’s assurances, there were several defects that should be regarded as a whole. The main areas of concern were the evidence of the way in which asylum interviews are conducted, of the absence of any opportunity for a claimant to present their case through a lawyer, of the lack of sufficient skills and experience to make reliable decisions, of any NGOs being unlikely to have sufficient capacity to provide legal assistance, and the fact that the appeal process to the High Court is wholly untested, coupled with concerns about the independence  of the Rwandan judiciary. He stated that whilst those defects could be resolved by making changes to the process, by capacity building and by effective training, these steps have not been taken to the necessary extent.

Underhill LJ considered that the Divisional Court’s reasoning could not be supported as it did not seek to engage with the details of the UNHCR’s criticisms of the RSD process. This was due to the weight it attached to the UK Government’s view that the Government of Rwanda would honour its assurances; he held that although great weight should be given to that view, the real issue was the ability of the Government of Rwanda to deliver on its assurances in the light of the present state of the Rwandan asylum system. He held that the asylum system in Rwanda has serious deficiencies that had not been corrected, nor were likely to be in the short term.

Whilst an argument was advanced by the SSHD that the inadequacy of the system would not necessarily lead to a real risk of refoulement, as failed asylum seekers would be considered for a grant of legal residence on other grounds, this was not accepted.

Underhill LJ considered that it was not necessary to decide whether the Divisional Court’s conclusion that there was no real risk of article 3 breaches due to the reception conditions in Rwanda was wrong, given his conclusion on the refoulement issue. In view of his assessment, he concluded, in agreement with the MoR, 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.” 

Conversely, the LCJ held that there were no substantial grounds for believing that there is a real risk that deficiencies in the asylum system will lead to wrong decisions and, thus, in refoulement. He considered that, in view of the terms of the agreement, the Government of Rwanda’s incentives to deliver its promises, the scrutiny that will be placed on decision-making and the strong monitoring arrangements in place, meant that the risk of wrong or perverse decisions, and by extension, the risk of refoulement was low. He also held that there were no substantial grounds for believing that there is a real risk that asylum seekers will face treatment prohibited by article 3 ECHR in Rwanda.


It is interesting to note the difference of approach between the majority, the MoR and Underhill LJ, on one hand, and the LCJ, on the other, in considering whether the Divisional Court had erred in its approach to the Soering question. The former relied on an objective reading of the Divisional Court’s judgment, with reference to several passages, before concluding that it had erred. They did not suggest that the Divisional Court had not understood what it was required to consider under the Soering test but that, on analysis of its judgment, it had for the most part failed to apply the correct test (which was perhaps explained by the confusing way in which the issues were presented). The LCJ, conversely, considered that it would have been remarkable for the Divisional Court to have made what he considered the most basic error in an ECHR-based claim, although he confirmed he also relied on a reading of the judgment as a whole to reach the opposite conclusion to that of the majority. At [76], the MoR appears to essentially respond to the LCJ’s reasoning at [486-489], which was that the SSHD could not have lawfully reached the conclusion that there was no risk of article 3 mistreatment unless there was in fact no such real risk. The MoR held it was not appropriate to express the Soering question in terms of whether the SSHD had lawfully reached that conclusion in any event, as the court is required to reach its own conclusion on the matter. Underhill LJ caveated that, had it been clear that the (wrong) language used by the Divisional Court did not reflect its true reasoning, it would not have necessarily resulted in a finding of error, but that was not the case. As such, whilst the majority gave the objective reading of the judgment its due importance, it cannot be said that they took an overly linguistic and not objective approach.

The difference of approach in considering the Soering test itself is further notable. Whilst the LCJ confirms early in his judgment that the consideration of whether there is a real risk of article 3 breaches requires rigorous analysis of the evidence, it is clear that he, similarly to the Divisional Court, assessed that the assurances between the UK and Rwandan governments and the monitoring arrangements were sufficient to obviate any risk, despite the past and present issues demonstrated by the UNHCR’s evidence. Whilst he relied on Othman to state that monitoring compliance with the UK-Rwanda agreement was an important factor, the MoR, also in reliance on Othman, held that the court should take a more pragmatic approach and consider, whilst giving great weight to the UK’s view of Rwanda’s assurances, whether they can be relied upon in light of the latter’s practices. In the same way, Underhill LJ considered that the Divisional Court had not engaged with the details of the UNHCR’s criticisms of Rwanda’s asylum system, due to having given undue weight to the good faith of the Government of Rwanda at a political level, rather than on its ability to deliver on its assurances in view of the evidence of the present circumstances.

The LCJ referred to the fact that a court is not institutionally well-equipped to make a judgment on the worth of a diplomatic assurance, with reference to domestic case-law; the executive has the expertise and access to advice which a court does not have. However, the cases to which he referred (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) did not, at least for the material part, involve an ECHR element. Whilst in ECHR cases great weight ought to be given to a state government’s view formed within its margin of appreciation, it is for the court to assess any relevant matters, including the diplomatic assurances, per the guidance of the Strasbourg Court.

This is reflective of two distinct approaches in judicial-making: a narrow, deferential approach to the executive’s decisions, and a more pragmatic and objective approach. The question is which approach is correct to a particular jurisdictional situation. In cases involving an evaluation of risk to national security or other matters to which the executive is institutionally and constitutionally competent and where the legal framework limits the courts’ jurisdiction, as the Supreme Court explained in Begum, the former approach is correct. In ECHR cases, it is the latter approach that is correct: the court has to make its own assessment whilst giving great weight to the state’s view. 

In the event the Supreme Court disagreed with the Court of Appeal’s majority judgment and considered that the Divisional Court did not err in its approach to the Soering question, or was not plainly wrong in its reasoning, it would likely endorse its decision in view of the appellate jurisdiction. If the matter then reached it, the Strasbourg Court would not have such jurisdictional limitations. Even if the Court of Appeal’s judgment remains judicially undisturbed, given that the Rwanda policy was only found unlawful on the basis that Rwanda is not (currently) a safe third country, it is in principle possible, as Underhill LJ confirmed, that in the future the real risk of article 3 breaches, and by extension the unlawfulness, will be remedied. In that case, if the Rwanda policy is sought to be implemented and judicially challenged again, the conclusion to the challenge may be different. Even then, the Court of Appeal’s unanimous decision that, in view of the stricter test and contestable matter, the certifications of the asylum and human rights claims were generally unlawful likely means that asylum seekers, whom the SSHD decides should be relocated to Rwanda, should generally be afforded a right of appeal to the First-tier Tribunal. We will have to see whether that will remain the case if the Illegal Migration Bill is endorsed by Parliament.

Contact our Immigration Barristers

For advice and assistance with immigration applications, appeals and judicial review claims, contact our immigration barristers on 0203 617 9173 or complete our enquiry form below.


To arrange an initial consultation meeting, call our immigration barristers on 0203 617 9173 or fill out the form below.

    Attach a file if it supports your enquiry. Only .doc or .pdf files.


    Expert advice & representation from immigration barristers that you can rely on.

    Google+ - Five Stars

    Read the 600+ five out of five star Google reviews of our immigration barristers.