R (on the application of EM (Eritrea)) v Secretary of State for the Home Department

Is an asylum seeker or refugee who resists his or her return from the United Kingdom to Italy (the country in which she or he first sought or was granted asylum) required to establish that there are in Italy “systemic deficiencies in the asylum procedure and in the reception conditions of asylum seekers ... [which] amount to substantial grounds for believing that the asylum seeker would face a real risk of being subjected to inhuman or degrading treatment...” (emphasis added). This formulation is taken from para 94 of R (NS) (Afghanistan) v Secretary of State for the Home Department[2013] QB 102. The mooted requirement that there be a systemic deficiency lies at the heart of this appeal.

That is the first and principal issue. It also constitutes the critical finding of the Court of Appeal. But, somewhat unusually, it is an issue on which there is no significant dispute between the parties. The appellants, the interveners (UNHCR), and the respondent all assert and agree that the Court of Appeal was wrong to hold that “… the sole ground on which a second state is required to exercise its power under article 3(2) of Regulation 343/2003 to entertain a re-application for asylum or humanitarian protection, and to refrain from returning the applicant to the state of first arrival, is that the source of risk to the applicant is a systemic deficiency, known to the former, in the latter's asylum or reception procedures” (emphasis added) – [2012] EWCA Civ 1336; [2013] 1 WLR 576, para 62.

The parties are also agreed that the test laid down in Soering v United Kingdom(1989) 11 EHRR 439 on this issue continues to hold the field. That case had established that the removal of a person from a member state of the Council of Europe to another country is contrary to the European Convention on Human Rights (ECHR) “where substantial grounds have been shown for believing that the person concerned ... faces a real risk [in the country to which he or she is to be removed] of being subjected to [treatment contrary to article 3 of the Convention]” – para 91 of Soering.

Council Regulation 343/2003 is commonly known as the Dublin II Regulation. In certain circumstances it provides that asylum claims must be processed and acted on by the member state of the European Union in which an asylum seeker first arrives. Asylum seekers and those who have been granted asylum (refugees) may therefore be returned to the first member state by any other member state of the EU in which asylum seekers and refugees subsequently arrive.

But where a person claims that his removal from the United Kingdom would expose him to the risk of breach of his human rights and/or article 3 ill-treatment within the member state to which it is proposed to return him, he has a statutory right of appeal under section 82 of the Nationality, Immigration and Asylum Act 2002 against a decision to remove him. This right is exercisable from within the United Kingdom unless the Secretary of State certifies the claim to be “clearly unfounded”. By virtue of section 92(4)(a) of the 2002 Act and of para 5(4) in Part 2 of Schedule 3 to the Asylum and Immigration (Treatment of Claimants etc) Act 2004, claims concerning removals to a listed country (of which Italy is one) are to be certified as clearly unfounded unless the Home Secretary is satisfied that they are not.

Such a certificate can be issued if "on any legitimate view" the claimant’s assertion that his enforced return would constitute a violation of his human rights would fail on appeal: R (Yogathas) v Secretary of State for the Home Department[2003] 1 AC 920, by Lord Hope at para 34; R (L) v Secretary of State for the Home Department[2003] EWCA Civ 25; [2003] 1 WLR 1230 and ZT (Kosovo)[2009] 1 WLR 348.

The Home Secretary in each of these appeals has decided that the contention that Italy is in systemic breach of its material international obligations is clearly unfounded, and that there is no separate reason to abstain from removal. Certification that the claims are clearly unfounded has the effect of prohibiting any appeal while the applicant remains in the United Kingdom.

Sir Stephen Sedley, who delivered the judgment of the court in the Court of Appeal, summarised the accounts given by the appellants in paras 13 to 28 of that judgment. The brief description of their circumstances which follows is drawn mainly from that synopsis. By way of preamble Sir Stephen correctly observed that, when deciding whether an asylum claim is capable of succeeding, it is customary to take the facts at their highest in the claimant’s favour. That is the approach that I intend to follow in my consideration of these cases. Where, therefore, it is stated that a particular event took place or that a certain factual proposition is established, this is for the purposes of considering the appellants’ cases at their reasonable height. It does not betoken any final finding or conclusion.

EH is an Iranian national aged 32. He arrived in Italy on 11 November 2010 or thereabouts. It is recorded that his fingerprints were taken on that date. A short time later he left Italy and made his way to the United Kingdom. On 11 March 2011 he applied for asylum in this country on the ground that he had been tortured while a political detainee in Iran. When it became clear that he had first claimed asylum in Italy, the Italian authorities were contacted about EH. They failed to respond within the time stipulated in Dublin II and they were deemed to have accepted responsibility for his claim. (It appears that the Italian authorities subsequently accepted responsibility for the claim.) EH’s claim was certified as being clearly unfounded. Removal directions were set. EH launched judicial review proceedings to challenge both the decision to certify and the removal directions. He claimed that there was a real risk that he would be subjected in Italy to inhuman and degrading conditions. He relied not on his own experience of reception in Italy, which was brief, but on that of others.

There is an abundance of evidence that EH is now severely disturbed and suffering from PTSD and depression, both of which require treatment. The Court of Appeal found that there was a real risk that EH, whether as an asylum-seeker or as an accepted refugee, will be homeless if returned to Italy. For the purposes of the present appeal that finding cannot be challenged.

EM is an Eritrean national. It is believed that he was born on 8 January 1989. He is an Orthodox Pentecostal Christian. His father was of the same faith and had been arrested by the Eritrean authorities for having arranged prayer meetings at the family home. His uncle was concerned that EM would also be arrested on suspicion of following his father’s faith and made arrangements for him to leave Eritrea. EM arrived in Italy at Lampedusa, and was first recorded as being there on 21 August 2008. He was fingerprinted and placed in a hotel in Badia Tedalda in the Arezzo province. After about 2 months he and the other asylum seekers there were told that they must each pay €120 for further processing of their applications. Having no money, he and other asylum seekers, who were likewise without funds, were given train tickets to Milan. For some three weeks after he arrived there he was himself homeless and destitute, living among other asylum-seekers in similar circumstances.

A fellow asylum seeker helped him to travel clandestinely to the United Kingdom, where he claimed asylum on 11 November 2008. His fingerprints were found to correspond with fingerprints on record in Italy. On 18 November 2008 Italy was asked to accept responsibility for his claim and, having failed to respond, was deemed to have accepted responsibility. Removal directions were set, but were challenged by an application for judicial review. On 1 June 2010 the Home Secretary certified EM's asylum claim as clearly unfounded. This was also challenged in the judicial review proceedings.

AE fled from Eritrea because she and her husband had been ill treated by the authorities after their arrest on suspicion that her husband was helping people to leave the country illegally. She arrived in Italy in August 2008 and was screened. After this she was placed in a hotel at Bibbiano in the north of Italy in the Emilia-Romagna region. She was accommodated there for some three months and about halfway through her stay she was interviewed about her asylum claim. At the end of that period, AE was recognised as a refugee and granted a five year residence permit. At about the same time she and other inhabitants of the hotel were told that it was too expensive to house them there and they were sent to a place that she knew as Aruso but was probably Arezzo. She was given accommodation in crowded and insanitary premises which she was obliged to share with other women and with men. Vouchers which she was given for food ran out after two weeks and she depended on charities for food after that.

After three months they were told that they had to leave. AE and a friend went back to Bibbiano. They were refused accommodation but managed to contact a friend who let them stay with him for a month, sharing a room with three men. They left after one of the men tried to rape AE. She and her friend managed to get train tickets to France and she then secretly boarded a lorry which took her to the United Kingdom, arriving here on 19 January 2010. Following unsuccessful judicial review proceedings she was returned to Italy on 15 October 2010.

She then found herself homeless and destitute in Milan. In desperation she was forced to live in a squat where she was repeatedly raped by a number of men who threatened her with reprisals if she reported them. Finally, with €100 borrowed from a fellow Eritrean, she made her way back to this country, where she was detained on arrival. A decision was made to remove her again to Italy. Her claim that to do so would violate her human rights was certified by the Home Secretary as clearly unfounded, and an application for permission to seek judicial review of the certificate was dismissed.

Psychiatric evidence was submitted to the Home Secretary to the effect that AE was traumatised as a result of her experiences in Italy and suicidal at the prospect of being returned there. It was contended that to return her to Italy would violate her rights under article 3 of ECHR. The Home Secretary rejected an application to use her discretionary power to transfer AE’s refugee status to the United Kingdom and confirmed the decision to remove her to Italy. In response to a Rule 39 indication issued by the European Court of Human Rights (ECtHR), removal of AE has been stayed. On 10 November 2011 her renewed application for permission to apply for judicial review was refused by the Administrative Court. Her challenge to the refusal to transfer her refugee status to this country was not pursued but the challenge to the certification of her claim remains.

MA is an Eritrean woman who reached Italy in 2005 and in April 2006 was accorded refugee status there on the ground of fear of persecution as a Pentecostal Christian. In January 2008 an agent brought her three children to Italy to join her: M, D and Y.

MA's evidence is that the family, despite being recognised as refugees, had to live on the streets, sleeping under bridges, lighting fires for warmth when rain permitted and relying on charitable hand-outs for food. After three months MA brought her children covertly to the United Kingdom. In the course of embarking in a lorry at Calais in the dark, she lost Y, whose whereabouts are still not known. The other two are now settled in secondary and tertiary education here and are both doing well.

Because of their failure to respond to the UK's request, the Italian authorities in July 2008 were deemed under Dublin II to have accepted responsibility for MA and her children. Removal directions were set but were cancelled because the Italian police considered that they had been given inconsistent details about the children and would not accept them. MA would not cooperate with attempts to interview her about this. Instead she sought to oppose removal by reliance on medical evidence that she was HIV positive. By July 2009 Italy had accepted responsibility and fresh removal directions were set. They were cancelled because of a new application for judicial review, which was later withdrawn. They were re-set for July 2010, but the family failed to check in for their departure to Italy. MA then made further allegations about her treatment both in Eritrea and in Italy.

In August 2010 the Home Secretary certified MA’s claim as clearly unfounded. She refused to transfer MA's refugee status to the United Kingdom and re-set removal directions. These were cancelled when the present proceedings were brought.

The Court of Appeal found that MA had displayed considerable deviousness. She had lacerated her fingertips to prevent identification on arrival here and had used a different name from that which she used in Italy. It was only after a third set of removal directions was given that, for the first time, she gave an account of being serially raped in both Italy and Eritrea. As the court found, however, her late accounts of rape do not necessarily make them incredible. Moreover, MA's account of the effects of her experiences is now supported by what appears to be cogent medical evidence.

As to MA's two children, M, although now legally an adult, continues to form part of the mother's human rights claim. She is taking a course at an educational establishment, and staff there speak highly of her. D is at a school which has reported favourably on both his behaviour and his academic progress. Neither child has any desire to be returned to Italy, with its associations of misery and hardship. MA is reportedly suicidal at the prospect of enforced return.

The Court of Appeal sat as a first instance court in two of the cases (AE and EH) and in its appellate jurisdiction in the cases of EM and MA [2013] 1 WLR 576. This came about because permission to allow AE and EH to apply for judicial review was refused at first instance and granted on application to the Court of Appeal which then conducted the substantive hearing in those cases. In the cases of EM and MA, appeals against substantive decisions by, respectively, Kenneth Parker J and Langstaff J were heard by the Court of Appeal in a conjoined hearing with EM and MA.

In para 30 of its judgment the Court of Appeal summarised the evidence that had been proffered by the Secretary of State:

The challenge which the appellants presented to the claims contained in this passage was set out in para 31 of the court’s judgment:

The Court of Appeal held that if “the matter stopped [t]here” they would be bound to conclude that there was a triable issue in all four cases as to whether return to Italy entailed a real risk to exposing the appellants to inhuman or degrading treatment contrary to article 3 of ECHR. This is clearly in keeping with well-established jurisprudence in the area. For instance, in ZT (Kosovo) v Secretary of State for the Home Department [2009] 1 WLR 348, para 23 Lord Phillips said, “If any reasonable doubt exists as to whether the claim may succeed then it is not clearly unfounded”. Plainly, therefore, the Court of Appeal considered that if it could have regard to the evidence presented on behalf of the appellants, their claims could not be characterised as “clearly unfounded”. The Home Secretary’s certificates would therefore have been of no effect and the appellants would have to be afforded an in-country appeal against removal. But the court found itself deflected from giving effect to this preliminary view because of what it understood to be the Home Secretary’s argument that access to article 3 and the assertion of a right of appeal could only be countenanced if it was shown that Italy was “in systemic rather than sporadic breach of its international obligations” and the case made on behalf of the appellants fell well short of establishing that.

The Court of Appeal felt driven to this conclusion by its analysis of recent jurisprudence from ECtHR and the Court of Justice of the European Union (CJEU), particularly the trilogy of cases, KRS v United Kingdom (2008) 48 EHRR SE 129, MSS v Belgium and Greece (2011) 53 EHRR 28, and NS (Afghanistan) v Secretary of State for the Home DepartmentCases C-411/10 and C-493/10, [2013] QB 102.

In the first of these cases, KRS, the Fourth Section of ECtHR found the applicant’s case to be inadmissible. He was an Iranian asylum seeker who had entered Greece before seeking asylum in the UK. Adverse reports on Greece’s treatment of asylum seekers were noted by the Fourth Section but it concluded that Greece’s international commitment to the European asylum system and (it was to be presumed) her compliance with that system provided a comprehensive answer to the applicant’s claim. Although UNHCR had advised member states to suspend returns to Greece under Dublin II, this had not displaced the presumption that Greece would abide by her obligations.

In the second case, MSS, a Grand Chamber decision, ECtHR noted UNHCR’s claim (in a letter to the Belgian government in April 2009) that the Fourth Section in KRS had apparently overlooked some of the criticisms that it had made of Greece. No reference had been made to whether conditions of reception conformed to regional and international standards of human rights protection or whether asylum seekers had access to fair consideration of their asylum applications or if they were able to exercise their rights under the Geneva Convention.

The Grand Chamber reviewed the numerous reports and materials that had been generated about the situation in Greece since the KRS decision. It observed that these all agreed about the deficiencies of the asylum procedure in Greece. The court therefore concluded that the situation in Greece was known to the Belgian authorities; that seeking assurances from the Greek government that the applicant faced no risk of treatment contrary to ECHR was not sufficient to ensure adequate protection against the risk where reliable sources had reported practices that were tolerated by the authorities and which were manifestly contrary to the principles of the Convention; and that the Aliens Office of the Belgian government “systematically applied the Dublin Regulation … without so much as considering the possibility of making an exception” – (para 352). The Grand Chamber therefore held that there had been a violation by Belgium of article 3 of EHCR because by sending the applicant back to Greece, the Belgian authorities exposed him to detention and living conditions there which were in breach of that article.

The Court of Appeal said of this decision that “the assessment of risk on return is seen by the Strasbourg court as depending on a combination of personal experience and systemic shortcomings which in total may suffice to rebut the presumption of compliance” – (para 39). It is clear that the court felt that the personal experience of the appellants in these cases, taken in combination with documented shortcomings in the manner in which asylum-seekers are dealt with in Italy, would have at least raised a case to be tried as to whether their enforced return to that country would have violated their article 3 rights (see, in particular, paras 32 and 61 of the court’s judgment). But the court decided that raising an arguable case was not enough. It reached that conclusion principally because of its view as to the effect of the CJEU decision in NS.

Notably, in introducing his discussion of that case, Sir Stephen Sedley said (at para 43) that, but for the fact that the decision of CJEU was binding on courts of this country, the Court of Appeal might have had to confront the problem of conflicting decisions of ECtHR and CJEU. This observation seems clearly to signify that, but for the effect of the NS case, the Court of Appeal would have come to a different conclusion from that which it felt compelled to reach. Resonances of this conflict appear later in the judgment of the Court of Appeal and will be touched on in my consideration of the NS decision.

The NS case was concerned with the question whether, in deciding if it should exercise the power under article 3(2) of the Dublin II Regulation (that is the power to examine a claim which is the responsibility of another state), a member state is required to presume conclusively that the other state’s arrangements are compliant with its international obligations. Alternatively, is the member state which is contemplating recourse to the article 3(2) power obliged to examine whether transfer would bring a risk of violation of Charter rights or of the EU's minimum standards?

CJEU decided that there was a presumption that member states would comply with their international obligations but that this was rebuttable. At para 86 of its judgment the court said:

Building on that finding CJEU said this at para 94 of its judgment:

The Court of Appeal considered that CJEU had addressed in its judgment the question of what amounted to systemic deficiencies in paras 81 and 82 for, at para 46 of the Court of Appeal’s judgment, Sir Stephen said that CJEU had taken care in those paragraphs to draw a distinction between a true systemic deficiency and operational problems even if such problems created a substantial risk that asylum seekers would be treated in a manner incompatible with their fundamental rights.

It will be necessary in due course to look at the relevant paragraphs of CJEU’s judgment in order to examine whether that conclusion can be upheld. For the present, it is, perhaps, sufficient to consider its implications. A person applying for asylum in a member state might be able to establish conclusively that he would be at substantial risk of being treated in a manner incompatible with his fundamental rights if returned to a listed country but because that risk did not arise from so-called systemic deficiencies it could not operate to prevent his enforced return to that country. That would be, to say the least, a remarkable conclusion.

In any event, the Court of Appeal decided that proof of “a systemic deficiency in the system of refugee protection” had been elevated by NS into a “sine qua non of intervention” – para 47. The court said:

It is clear that the Court of Appeal considered that NS had changed the landscape in relation to the requirements of proof of possible violation of fundamental rights from that which had hitherto obtained. At para 61 the Court of Appeal, having reviewed the evidence that had been presented on behalf of the appellants about conditions in Italy, said this:

The need for a workable system to implement Dublin II is obvious. To allow asylum seekers the opportunity to move about various member states, applying successively in each of them for refugee status, in the hope of finding a more benevolent approach to their claims, could not be countenanced. This is the essential underpinning of Dublin II. Therefore, that the first state in which asylum is claimed should normally be required to deal with the application and, where the application is successful, to cater for the refugee’s needs is not only obvious, it is fundamental to an effective and comprehensive system of refugee protection. Asylum seeking is now a world-wide phenomenon. It must be tackled on a co-operative, international basis. The recognition of a presumption that members of an alliance of states such as those which comprise the European Union will comply with their international obligations reflects not only principle but pragmatic considerations. A system whereby a state which is asked to confer refugee status on someone who has already applied for that elsewhere should be obliged, in every instance, to conduct an intense examination of avowed failings of the first state would lead to disarray.

It is entirely right, however, that a presumption that the first state will comply with its obligations should not extinguish the need to examine whether in fact those obligations will be fulfilled when evidence is presented that it is unlikely that they will be. There can be little doubt that the existence of a presumption is necessary to produce a workable system but it is the nature of a presumption that it can, in appropriate circumstances, be displaced. The debate must centre, therefore, on how the presumption should operate. Its essential purpose must be kept clearly in mind. It is to set the context for consideration of whether an individual applicant will be subject to violation of his fundamental rights if he is returned to the listed country. The presumption should not operate to stifle the presentation and consideration of evidence that this will be the consequence of enforced return. Nor should it be required that, in order to rebut it, it must be shown, as a first and indispensable requirement, that there is a systemic deficiency in the procedure and reception conditions provided for the asylum seeker.

Violation of article 3 does not require (or, at least, does not necessarily require) that the complained of conditions said to constitute inhuman or degrading conditions are the product of systemic shortcomings. It is self-evident that a violation of article 3 rights is not intrinsically dependent on the failure of a system. If this requirement is grafted on to the presumption it will unquestionably make its rebuttal more difficult. And it means that those who would suffer breach of their article 3 rights other than as a result of a systemic deficiency in the procedure and reception conditions provided for the asylum seeker will be unable to avail of those rights in order to prevent their enforced return to a listed country where such violation would occur. That this should be the result of the decision of CJEU in NS would be, as I have said, remarkable.

More significantly, if the Court of Appeal’s interpretation of NS was correct, it would give rise to an inevitable tension between the Home Secretary’s obligation to abide by EU law, as pronounced by CJEU, and her duty as a public authority under section 6 of the Human Rights Act 1998. On the Court of Appeal’s analysis, the Secretary of State would be bound under Dublin II to return an asylum-seeker or refugee to the first country in which that person had claimed or been granted asylum unless he or she could show that the anticipated breach of their article 3 rights had as its source a systemic deficiency in the asylum procedure and reception conditions. Thus, even if it could be proved conclusively that an article 3 violation was likely to occur, the return of the individual would have to take place. Such an enforced return would involve the Secretary of State in a failure to comply with the duty under section 6 of the 1998 Act not to act in a way that is incompatible with a Convention right.

It may well be that, confronted by such a dilemma, the Secretary of State would have to resort to her powers under article 3(2) of the Dublin Regulation which permits each member state to examine an application for asylum lodged with it by a third-country national, even if such examination is not its responsibility under the criteria laid down in the Regulation. In the event, I do not believe that it is necessary to reach a view on this because I do not consider that NS has the effect which the Court of Appeal considered it to have.

NS was an Afghan national who challenged his removal under the Dublin II Regulation to Greece by the Secretary of State. He relied on material concerning the general situation in Greece for asylum seekers. A series of questions were referred to the CJEU. These raised queries about the Charter of Fundamental Rights of the European Union (the Charter) and the relationship between fundamental rights and returns under the Dublin II Regulation. In the present appeals, of course, the issue of importance from NS is the court’s decision about the circumstances in which a member state must desist from transferring an asylum applicant to the state with responsibility under the Regulation.

In paras 76-80 of its judgment, CJEU sets out the background to the need for mutual confidence among member states about the obligation of those states that participate in the Common European Asylum System to comply with fundamental rights including those based on the Convention relating to the Status of Refugees (the 1951 Convention) ((1951) Cmd 9171) and its 1967 Protocol ((1967) Cmnd 3906). In these paras the court also dealt with the assumption that needed to be made that the states will be prepared to fully comply. These twin considerations (the importance of the obligations and the assumption that they will be fulfilled) underpin the system – a system designed to “avoid blockages … as a result of the obligation on state authorities to examine multiple claims by the same applicant, and … to increase legal certainty with regard to the determination of the state responsible for examining the asylum claim and thus to avoid forum shopping, it being the principal objective …to speed up the handling of claims in the interests both of asylum seekers and the participating member states.” – para 79.

The aspirational aspect of this approach is readily understandable. If the system is going to work properly, if administrative delays and forum shopping are to be eliminated and if bureaucratic quagmires are to be avoided, participating states must live up to their commitments and they must inspire trust in the other participants and, in turn, repose trust in the willingness and capacity of the other participants to likewise fulfil their obligations. CJEU was therefore anxious to ensure that there was no significant compromise on the smooth operation of the inter-state return of asylum-seekers to the country where they first claimed asylum. The critical question is whether it sought to achieve that effective process by permitting challenges to a decision to return under Dublin II only in those cases where there is a systemic failure in the asylum procedure and reception conditions in the state to which the transfer is to take place.

Before examining what CJEU said on this issue, it can be observed that an exclusionary rule based only on systemic failures would be arbitrary both in conception and in practice. There is nothing intrinsically significant about a systemic failure which marks it out as one where the violation of fundamental rights is more grievous or more deserving of protection. And, as a matter of practical experience, gross violations of article 3 rights can occur without there being any systemic failure whatsoever.

One must be careful, therefore, to determine whether CJEU referred to systemic failures in order merely to distinguish these from trivial infringements of the various European asylum directives or whether it consciously decided to create a new and difficult-to-fulfil pre-condition for asylum-seekers who seek to have recourse to their article 3 rights to prevent their return to a country where it can be shown that those rights will be violated. For there can be little doubt that such a condition would indeed be difficult to fulfil. Some of the facts in the present cases exemplify the truth of that proposition. For instance, the Court of Appeal held that there was a real risk that EH, now severely disturbed and suffering from PTSD and depression, both of which require treatment, will be homeless if returned to Italy. But that is not enough to prevent his enforced return. The appalling degradation suffered by AE and the awful but distinct possibility that something of the same will happen again if she is returned to that country are not sufficient to satisfy the stringent standard which the Court of Appeal has decided must now be met.

Because of the narrowly defined (by the Court of Appeal) category of systemic failures in asylum procedures and reception conditions, which these appellants have been deemed not to inhabit, they are prohibited from challenging the validity of their enforced return to a country where, if their claims are right, they will suffer breach of their article 3 rights. The unacceptable artificiality of that situation is that if a systemic failure could be demonstrated, even though the consequences were far less terrible than those which, it is anticipated, will befall these appellants, the enforced return could be resisted.

With these concerns in mind, I turn to consider the critical paragraphs in the judgment of CJEU in NS. At para 80, the court said that it must be assumed that the treatment of asylum seekers in all member states complies with the requirements of the Charter, the Geneva Convention and the ECHR. Para 81 is pivotal to the court’s reasoning:

The system referred to in this para is the system of the “treatment of asylum seekers in all member states” (see para 80). What is contemplated in para 81 is that this system may experience major operational problems in a particular member state. The circumstance that the general system may experience major operational problems in specific settings is not the same as the system having intrinsic deficiencies. The Court of Appeal in para 46 of its judgment suggested that CJEU had taken care to distinguish “a true systemic deficiency” from “operational problems”. With respect, I do not agree. What the CJEU recognised was that any system, however free from inherent deficiency, might experience operational difficulties which would cause a substantial risk that asylum seekers would be treated in a manner incompatible with their fundamental rights. The source of the risk was not systemic deficiencies (in the sense of the deficiencies deriving from intrinsic weaknesses in the system) but rather, major operational problems in a given member state.

I therefore take a different view from that of the Court of Appeal in its analysis of paras 80 and 81 of the CJEU judgment. I do so on two grounds. First, I do not believe that “the system” (as that expression was used by CJEU in these paras) was the system in a particular member state. I consider that the words “that system” in para 81 are a reference back to the system of “treatment of asylum seekers in all member states” in para 80. Secondly, I am of the view that the source of the risk of asylum seekers being treated in a manner incompatible with their fundamental rights, which CJEU identified in these paras, is not a deficiency in the overall system but operational problems experienced in “a given member state”. See also in this context paras 75 and 78 of NS.

Now, it is true that at a later point in the judgment, CJEU turns to refer to systemic flaws in the asylum procedure and reception conditions in Greece. At para 86 the court said:

It is perhaps unfortunate that the expression “systemic deficiency” was employed in two different contexts to describe what are clearly distinctly different phenomena because this creates the potential for confusion. But I believe that, even in the later context, CJEU did not intend to stipulate that an anticipated violation of article 3 could only be a basis for resisting a transfer to a listed state if it could be shown that this was the result of a systemic deficiency in that country’s asylum procedures and reception conditions. Indeed, it is clear from para 89 of the court’s judgment that it considered that the infringement of fundamental rights provided evidence of the systemic deficiency rather than that a systemic deficiency had to be demonstrated before violation of a fundamental right could operate to prevent the transfer. In that para the court said:

The important central feature of MSS and NS is that systemic deficiencies were found to be present in the asylum procedures and reception conditions in Greece. The debate in those cases therefore focused on the question of what, given that systemic deficiencies were present, the effect of those deficiencies was on the application of the presumption of compliance. There was no occasion to address the question whether systemic deficiencies had to be present before the interdict on transferring asylum seekers to the member state responsible. This is how, in my opinion, para 94 of the court’s judgment in NS should be read. In that para the court said:

The focus here is on the member states’ awareness of systemic deficiencies which provide substantial grounds for believing that there is a real risk of inhuman or degrading treatment. In other words, does the member state proposing to transfer an asylum seeker have grounds for believing that the consequence for the person transferred will be inhuman or degrading treatment? As it happened, in both those cases the existence of systemic deficiencies which had been extensively reported on by, among others, UNHCR was the means by which the transferring states were deemed to have that knowledge but there is nothing in the reasoning of CJEU nor is there, I believe, any reason in logic to suggest that, if the transferring state acquires the same knowledge through a different medium, that it should not have the same effect.

I consider that the Court of Appeal’s conclusion that only systemic deficiencies in the listed country’s asylum procedures and reception conditions will constitute a basis for resisting transfer to the listed country cannot be upheld. The critical test remains that articulated in Soering v United Kingdom (1989) 11 EHRR 439. The removal of a person from a member state of the Council of Europe to another country is forbidden if it is shown that there is a real risk that the person transferred will suffer treatment contrary to article 3 of ECHR.

Article 13(1) of Council Directive 2003/9/EC (the Reception Directive) requires that member states provide “material reception conditions” for applicants for asylum. Article 13(2) stipulates that these conditions should be such as “to ensure a standard of living adequate for the health of applicants and capable of ensuring their subsistence”. The Dublin Regulation and the Reception Directive must be interpreted and applied in conformity with fundamental rights: Case C-106/89 Marleasing SA v La Comercial Internacional de Alimentación SA [1990] ECR I-4135; JoinedCases C-402/05P and 415/05P Kadi v Council of the European Union [2009] AC 1225.

The preamble to Council Directive 2004/83/EC (the Qualification Directive) emphasises that, in contrast to the Reception Directive (which identifies minimum standards), the key objective is to ensure that those granted refugee status are not discriminated against in terms of access to welfare support, accommodation etc. Recital 33 is in these terms:

Articles 26-29 of the Qualification Directive requires member states to provide refugees with equivalent access to that enjoyed by nationals of the member state in areas such as employment, education, social welfare and medical treatment. Article 31 requires that they be given equivalent rights as regards accommodation and article 33 calls for member states to provide appropriate integration programmes.

These duties coalesce with the positive obligations on members of the Council of Europe who are also member states of the European Union. Under the EU Charter of Fundamental Rights, article 4 contains a human rights protection in equivalent language to article 3 of ECHR. The UK, as an EU member state, is obliged to observe and promote the application of the Charter whenever implementing an instrument of EU law (see article 51 of the Charter). It is common case that the positive obligations under article 3 of ECHR include the duty to protect asylum seekers from deliberate harm by being exposed to living conditions (for which the state bears responsibility) which cause ill treatment – see MSS at [221]. And in R (Limbuela) v Secretary of State for the Home Department[2006] 1 AC 396 the House of Lords held that article 3 could be engaged where asylum seekers were “by the deliberate action of the state, denied shelter, food or the most basic necessities of life” –per Lord Bingham at para 7.

Where, therefore, it can be shown that the conditions in which an asylum seeker will be required to live if returned under Dublin II are such that there is a real risk that he will be subjected to inhuman or degrading treatment, his removal to that state is forbidden. When one is in the realm of positive obligations (which is what is involved in the claim that the state has not ensured that satisfactory living conditions are available to the asylum seeker) the evidence is more likely to partake of systemic failings but the search for such failings is by way of a route to establish that there is a real risk of article 3 breach, rather than a hurdle to be surmounted.

There is, however, what Sales J described in R (Elayathamby) v Secretary of State for the Home Department [2011] EWHC 2182 (Admin), at para 42(i) as “a significant evidential presumption” that listed states will comply with their Convention obligations in relation to asylum procedures and reception conditions for asylum seekers within their territory. It is against the backdrop of that presumption that any claim that there is a real risk of breach of article 3 rights falls to be addressed.

In his first judgment in EM[2011] EWHC 3012 Admin, delivered on 18 November 2011, Kenneth Parker J referred approvingly to the statement in R v Home Secretary Ex p Adan [1999] 3 WLR 1274 to the effect that a system which will, if it operates as it usually does, provide the required standard protection for the asylum seeker will not be found to be deficient because of aberrations. He then said this at para 12:

‘Systematic’ is defined as “arranged or conducted according to a system, plan, or organised method” whereas the definition of the word ‘systemic’ is “of or pertaining to a system”. Taken in context, I believe that Kenneth Parker J’s statement that it had to be shown that there was a systematic and significant failure to comply with international obligations meant that the omissions were on a widespread and substantial scale. His approach is rather different from that of the Court of Appeal, therefore, in that it does not appear to suggest that it needed to be shown that there were inherent deficiencies in the system, merely that there were substantial operational problems. This approximates (at least) to what I consider is the true import of the decision in NS. On one view, therefore, Kenneth Parker J’s decision is in keeping with the correct test and his decision should stand.

For two reasons, however, I have decided that this would not be the correct disposal. In the first place the Court of Appeal took a different view from that of Kenneth Parker J as to the effect of the evidence. As I pointed out, (in paras 26 and 31 above) the court indicated that, but for the effect of NS, it would have been bound to conclude that there was a triable issue in all four cases as to whether return to Italy entailed a real risk to exposing the appellants to inhuman or degrading treatment contrary to article 3 of ECHR. Secondly, there is an issue as to whether Kenneth Parker J’s approach accords precisely with that in Soering. In that case ECtHR had said that an extraditing contracting state will incur liability under the Convention if it takes action “which has as a direct consequence the exposure of an individual to proscribed ill-treatment”. In order to rebut the presumption a claimant will have to produce sufficient evidence to show that it would be unsafe for the court to rely on it. On proper analysis, it may well be that Kenneth Parker J was not suggesting that there was a requirement that a person subject to an enforced return must show that his or her risk of suffering ill-treatment contrary to article 3 of EHCR was the result of a significant and systematic omission of the receiving state to comply with its international obligations. It seems to me, however, that, to impose such an obligation in every instance would go beyond the Soering requirement. Since there was no reference to Soering in Kenneth Parker J’s judgment and in light of this court’s re-assertion of the test articulated in that case, I consider that it would be sensible to have the matter revisited.

In MA, Langstaff J (whose judgment is reported at [2012] EWHC 56 Admin) said (at para 62) that it could not realistically be argued that Italy systematically breaches the rights of refugees so as to involve a violation of article 3. At para 63 he rejected the argument that to rely on an absence of systematic breach avoided dealing “with the practical realities of life in Italy”. Langstaff J said that such realities might need to be considered if the return was to “some less developed country in which the generality was for there to be such difficulties”. By implication, this approach suggests that a breach of article 3, sufficient to prevent a return, could only arise where there had been systematic breach of the rights of refugees. For the reasons given earlier, I consider that a more open-ended approach to the question of the risk of breach of article 3 is required. Although one starts with a significant evidential presumption that listed states will comply with their international obligations, a claim that such a risk is present is not to be halted in limine solely because it does not constitute a systemic or systematic breach of the rights of refugees or asylum seekers. Moreover, practical realities lie at the heart of the inquiry; evidence of what happens on the ground must be capable of rebutting the presumption if it shows sufficiently clearly that there is a real risk of article 3 ill-treatment if there is an enforced return.

I would therefore remit all four cases to the Administrative Court so that an examination of the evidence may take place to determine whether in each case it is established that there is a real possibility that, if returned to Italy, the claimant would be subject to treatment in violation of the Convention.

That examination can only be conducted properly if there is an assessment of the situation in the receiving country. In appropriate circumstances, this calls for a rigorous assessment – see Chahal v United Kingdom(1997) 23 EHRR 413 at para 96 and Vilvarajah v United Kingdom(1991) 14 EHRR 248 at para 108. The 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, including his or her previous experience – see Vilvarajah at para 108 and Saadi v Italy (2009) 49 EHRR 30 at para 130. This approach has been followed by decisions of ECtHR subsequent to MSS – Hussein v Netherlands Application No 27725/10 at paras 69 and 78 and Daytbegova v Austria Application No 6198/12 at paras 61 and 67-69.

The Court of Appeal recognised that particular importance should attach to the views of UNHCR and noted that ECtHR in MSS had treated UNHCR’s judgment as “pre-eminent and possibly decisive”. At para 41 Sir Stephen Sedley said this:

I fully agree with this assessment. In a recent decision of this court, the unique and unrivalled expertise of UNHCR in the field of asylum and refugee law was acknowledged. In IA (Iran) v Secretary of State for the Home Department [2014] UKSC 6; [2014] 1 WLR 384, this court said at para 44:

It is of course the case that UNHCR’s criticisms of the situation in Greece in its interventions in KRS and particularly MSS were more pointed and direct than they have been in the present appeal in relation to Italy. In a report of July 2012 containing recommendations in relation to Italy, UNHCR did not call for a halt to all Dublin transfers to Italy. But, as Mr Fordham QC, for UNHCR, submitted, this does not mean that the organisation considered that there were no legal obstacles to particular transfers taking place or that UNHCR had given Italy a “clean bill of health”.

The recommendations contained in UNHCR’s report of July 2012 and its more recent report of July 2013 will doubtless be examined carefully by the Administrative Court. While, because of their more muted contents, they do not partake of the “pre-eminent and possibly decisive” quality of the reports on Greece, they nevertheless contain useful information which the court will wish to judiciously consider. Assumptions should not be made about any lack of recommendations concerning general suspension of returns under Dublin II to Italy but it is of obvious significance that UNHCR did not make any such proposal. The UNHCR material should form part of the overall examination of the particular circumstances of each of the appellant’s cases, no more and no less.

Of the four appellants, two are asylum seekers (EH and EM), and two are refugees (AE and MA). AE and MA submit that their transfer to Italy is not governed by Dublin II and is not within the scope of EU law because they are refugees. The Treaty provision under which the Dublin Regulation was adopted, article 63(1) of the Treaty on European Union makes it clear that the Regulation is directed to determine which member state is responsible for considering an asylum application. Accordingly, the appellants say, the return of refugee appellants is governed exclusively by national law.

The respondent, whilst agreeing that refugee appellants are not returned to member states under Dublin II, takes a rather different approach to the question whether asylum seekers and refugees should be treated similarly. It is argued that ECtHR has consistently recognised that asylum seekers are an “underprivileged and vulnerable population group requiring special protection in the form of basic reception facilities” whereas refugees are “on a par, as regards rights and obligations … with the general population” – see Hassan and others v Netherlands and Italy 40524/10 (27 August 2013) para 179.

The Court of Appeal noted that questions had been raised in the course of argument as to whether the return to Italy of a claimant already granted refugee status there would fall under Dublin II but decided that the reasoning of the CJEU in NS required them to adopt a “uniform approach” to all of the present appeals – see para 48.

It seems to me that the relevant matter is not whether Dublin II treats refugees and asylum seekers differently or the same, but that it relates to anyone who has applied for asylum in the country from which he might be transferred, whether or not he has previously been recognised as a refugee in the country to which it is proposed he be transferred. This reflects the nature of Dublin II as a chiefly procedural instrument. ‘Refugee’ is defined, but referred to only once, obliquely, in article 7:

An applicant or asylum seeker is defined in article 2(d) of Dublin II as “a third country national who has made an application for asylum in respect of which a final decision has not yet been taken”. A third country national is defined in para (a) of the same article as “anyone who is not a citizen of the Union within the meaning of article 17(1) of the Treaty establishing the European Community”. The appellants meet these criteria and all are subject, therefore, to the provisions of Dublin II. Whether their respective positions as asylum seekers who have previously been granted refugee status and asylum seekers who have not been granted that status will make it more or less likely that they will be at risk of violation of their article 3 rights if returned to a listed country will depend on an examination of the particular circumstances of their individual cases. One can anticipate an argument that those who have refugee status in Italy are less likely to suffer such a violation because they can assert their rights under the Qualification Directive but whether such an argument would prevail must depend on the evaluation of the evidence which is presented on that issue.