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

Mr Hemmati is a national of Iran. He came to this country via Iraq, Turkey and Bulgaria. He claims that he would suffer ill-treatment if returned to Iran. He first claimed asylum in Bulgaria and was fingerprinted there. He entered the UK illegally, circumventing immigration controls. On 11 February 2015 he attended Lewisham police station, where he was arrested as an illegal entrant and, it appears, claimed asylum. A search using the Eurodac system revealed that he had first claimed asylum in the EU in Bulgaria in November 2014. On 14 February 2015 he was released. On 18 February 2015 a formal request was made to Bulgaria under the Dublin III procedure for Bulgaria to take responsibility for Mr Hemmati’s asylum claim, and on 17 April 2015 Bulgaria accepted such responsibility. On 27 April 2015 Mr Hemmati’s claim for leave to remain in the UK was refused by the Secretary of State on the grounds that he could be removed to a safe third country, namely Bulgaria.

On 8 June 2015 Mr Hemmati was detained in order to effect his removal to Bulgaria on the basis that, as explained in the explanatory notice IS.91R handed to him, his removal was “imminent” and he had “not produced satisfactory evidence of [his] identity, nationality or lawful basis to be in the UK”. In the notice, other boxes labelled “You are likely to abscond if given temporary admission or release” and “You have previously absconded or escaped” were not ticked. However, the evidence served by the Secretary of State was that, despite this, Mr Hemmati was assessed to pose a risk of absconding to prevent the implementation of his removal from the UK. This was principally on the basis that he had evaded immigration controls in order to enter the UK and had sought to deceive the UK authorities in relation to his earlier claim for asylum.

On 17 June 2015, removal directions were set for 7 July 2015. These were cancelled when Mr Hemmati gave notice that he had issued judicial review proceedings to challenge his removal to Bulgaria. The claim was brought on the grounds that removal to Bulgaria would give rise to a real risk of violation of his rights under Article 3 of the European Convention on Human Rights (“ECHR”), as incorporated into domestic law by the Human Rights Act 1998 (“HRA”). Eventually, Garnham J dismissed that challenge in another judgment – [2016] EWHC 857 (Admin); and an appeal against that judgment was dismissed: [2017] EWCA Civ 1871.

The Secretary of State considered whether to seek an order for expedition of the judicial review, but decided not to do so. Shortly thereafter, on 17 July 2015, Mr Hemmati was released from detention.

Mr Khalili is a national of Afghanistan. He came to this country via Bulgaria. He claims that he would suffer ill-treatment if returned to Afghanistan. Like Mr Hemmati, he first claimed asylum in Bulgaria and was fingerprinted there. Mr Khalili entered the UK illegally, evading immigration controls, on 18 November 2014. He claimed asylum here on 20 November 2014. He was initially released on temporary admission and told to attend Colindale police station on 2 January 2015, then the Secretary of State’s offices on 6 January 2015 for a screening interview. He complied with these requirements.

Mr Khalili made a formal in-country claim for asylum at that interview and was detained on the basis (as set out in the notice IS.91R handed to him) that his “removal from the United Kingdom [was] imminent”. The box labelled “You have used or attempted to use deception in a way that leads us to consider you may continue to deceive” was also ticked. Again, other boxes labelled “You are likely to abscond if given temporary admission or release” and “You have previously absconded or escaped” were not ticked. However, in an internal document reviewing his detention, prepared on about 13 January 2015, it was stated that it was assessed that there was a risk of absconding, it being noted that he had been a clandestine illegal entrant. The witness statement for the Secretary of State confirmed that Mr Khalili was detained because he was assessed to pose a risk of absconding to avoid removal from the UK.

The Secretary of State made a Eurodac search, discovered that Mr Khalili had made his previous asylum claim in Bulgaria and indicated that it was proposed to request the Bulgarian authorities to take him back pursuant to Dublin III. A formal request to Bulgaria was made on 8 January 2015. On 26 January 2015 Mr Khalili submitted representations seeking temporary admission and challenging the decision to make that request. On 28 January 2015, however, no response having been received from Bulgaria and it therefore being deemed to have accepted that request, the Secretary of State certified Mr Khalili’s asylum claim on the grounds that he could safely be removed to Bulgaria. On 10 February 2015 the Secretary of State set directions for his removal to Bulgaria on 23 February 2015. On 12 February 2015 Bulgaria formally agreed to take back Mr Khalili under Dublin III.

At about this time, Mr Khalili was served with a fresh IS.91R notice to explain that he was detained on the ground that his removal was “imminent”. The box referring to a risk of absconding was not ticked. However, internal review documents noted that he remained an absconder risk. The evidence filed by the Secretary of State confirms that he was detained on the basis that he was assessed to pose a risk of absconding.

In due course, Mr Khalili joined with Mr Hemmati and others in bringing the judicial review challenge regarding the safety of Bulgaria as a destination for removal referred to above. Mr Khalili issued his judicial review claim on 20 February 2015 to prevent the removal scheduled for 23 February 2015. On issue of that claim, the removal directions were cancelled.

On 3 March 2015, Mr Khalili applied to the First-tier Tribunal (“FTT”) for bail. The Secretary of State opposed bail on the grounds that Mr Khalili posed a high risk of absconding as he had shown a disregard for UK immigration policy. In an internal review of his detention at this time it was noted that the plan should be to “complete [the judicial review] and expedite”, and also that “detention should be reviewed in the event that we are unable to expedite [the judicial review]”.

On 9 March 2015 Mr Khalili was released on bail by the FTT.

Mr Abdulkadir is a national of Iraq. He claims he would suffer ill-treatment if removed there. He first claimed asylum in the EU in Austria in July 2015 and was fingerprinted there. He entered the UK clandestinely in the back of a lorry on 18 August 2015. Upon arrival in Kent, he ran from the lorry. The driver called the police, who apprehended Mr Abdulkadir. He was detained and a check was conducted using the Eurodac system, which revealed that he had previously claimed asylum in Austria. He then claimed asylum in the UK.

In the IR.91R notice given to Mr Abdulkadir in respect of his detention, the boxes ticked were “reasonable grounds to suspect directions may be given for [his] removal”; “there is insufficient reliable information to decide on whether to grant you temporary admission or release”; and “you have not produced satisfactory evidence of your identity, nationality or lawful basis to be in the UK.” The box in relation to risk of absconding was not ticked. However, the witness statement filed on behalf of the Secretary of State explains that the boxes ticked indicated that there was a risk of absconding, especially in light of the circumstances in which he was apprehended.

The Secretary of State made a request to Austria under Dublin III for it to accept responsibility to review Mr Abdulkadir’s asylum claim. On 15 September 2015 Austria formally accepted that responsibility.

On 19 September 2015 the Secretary of State certified that Mr Abdulkadir could safely be removed to Austria. On 28 September 2015 removal directions were set, but these were cancelled at the request of Austria for organisational reasons. Mr Abdulkadir made representations to challenge his removal to Austria on human rights grounds and requested temporary admission to the UK. On 7 October 2015 the Secretary of State certified his human rights claim against removal to Austria as “clearly unfounded” and wholly without merit.

On 15 October 2015 the Secretary of State set new removal directions for 23 October 2015. On 22 October 2015 Mr Abdulkadir issued judicial review proceedings to challenge his removal to Austria, alleging that in Austria he would be exposed to a real risk of treatment contrary to Article 3 of the ECHR. Upon receipt of that claim the removal directions were cancelled.

Mr Abdulkadir applied for bail to the FTT, but his application was withdrawn on 5 November 2015. On 10 November 2015 his detention was reviewed and maintained. He made a fresh application for bail, and on 13 November 2015 the FTT refused that application on grounds of the risk of absconding and the likelihood of removal within a short period. On 27 November 2015, permission was granted for Mr Abdulkadir’s judicial review claim, and the Secretary of State was given notice of this on 30 November 2015. Upon review of his detention in the light of this, Mr Abdulkadir was released from detention on 8 December 2015.

Eventually, Mr Adbulkadir’s challenge to his removal to Austria was dismissed by Irwin J in his judgment.

Mr Mohammed is a national of Iraq, of Kurdish Sorani ethnicity. He claims he would suffer ill-treatment at the hands of ISIS if removed there. He first claimed asylum in the EU in Austria in August 2015 and was fingerprinted there. He entered the UK clandestinely on 8 September 2015, hidden in the back of a lorry. He ran off when the lorry doors were opened, but was apprehended by the police the same day. He claimed asylum. He was detained and served with an IS.91R notice, with the box ticked to explain the reason for detention as “there was insufficient evidence to decide whether to grant him temporary admission on release.” The box for risk of absconding was not ticked. Nonetheless, the witness statement for the Secretary of State explains that the combination of illegal entry, the lack of ties in the UK, the failure to provide evidence of identity, the underlying circumstances of his arrival and his initial flight, taken together, indicated that he was assessed to pose a risk of absconding.

On 9 September 2015, a Eurodac search revealed that Mr Mohammed had previously claimed asylum in Austria. Therefore, removal there pursuant to Dublin III was identified as a possibility. A detention review of 14 September 2015 summarised the background and concluded that detention should be maintained as his removal from the UK “is considered to be a realistic prospect.” There was no specific comment about the risk of absconding, but the witness statement for the Secretary of State explains that for the same reasons as before Mr Mohammed was regarded as presenting a high risk of absconding.

On 15 September 2015 a formal request was made to the Austrian authorities for Mr Mohammed to be removed there pursuant to Dublin III. The same day, a detention review recorded that his risk of absconding was assessed to be high and that detention was necessary to prevent it. On 18 September 2015, Austria formally accepted responsibility to deal with Mr Mohammed’s claim for asylum. On 23 September 2015 the Secretary of State set removal directions for 5 October 2015. These were cancelled and re-set for 12 October. However, on 2 October 2015 Mr Mohammed commenced a claim for judicial review, on the ground (similar to that in Mr Abdulkadir’s case) that removal to Austria would violate his rights under Article 3 of the ECHR. The result of this was that the re-set removal directions were cancelled.

On 6 October 2015 a detention review in respect of Mr Mohammed assessed the risk of absconding as high, but by reason of the judicial review proceedings his removal was no longer imminent. The likely progress of the judicial review proceedings was reviewed thereafter, and in the absence of expedition it was decided that he should be released from detention, and instead made subject to reporting conditions. He was released from detention on 4 November 2015.

Eventually, as with Mr Abulkadir, the challenge to removal to Austria was dismissed by Irwin J.

SS is a national of Afghanistan. He claims to fear that he would be ill-treated if removed there. He first claimed asylum within the EU in Bulgaria in May 2015, again in Hungary in June 2015 and once more in Germany in July 2015. On 15 September 2015 he entered the UK clandestinely, hidden in the back of a train. Upon arrival here, he claimed asylum and also said that he was a child, aged only 16. The immigration officer assessed that he was an adult and dealt with him as such when deciding that he should be detained (in his judgment, Mr Howell QC found that SS was indeed an adult when he was detained). A Eurodac search later the same day revealed that SS had previously claimed asylum in Bulgaria, Hungary and Germany.

On 15 September 2015 SS was served with a notice IS96 stating that he was liable to be detained as there was a reasonable suspicion that he was liable to removal from the UK as an illegal entrant. He was also served with a notice IS.91R, with ticks against boxes stating that “There is insufficient reliable information to decide on whether to grant you temporary admission or release”; “You have used or attempted to use deception in a way that leads us to consider you may continue to deceive”; and “You have not produced satisfactory evidence of your identity, nationality or lawful basis to be in the UK”. The boxes referring to likelihood of absconding and to imminent removal from the UK were not ticked. The judge considered the case on the footing that the Secretary of State detained SS because it was assessed to be reasonably likely that he would be accepted by another EU Member State under the Dublin III procedure.

A formal request was made to Bulgaria under Dublin III to accept responsibility for dealing with SS’s asylum claim, but it declined. Therefore on 15 October 2015 formal requests were made to Hungary and Germany. On 27 October 2015 Germany accepted responsibility for dealing with SS’s asylum claim. On 27 November 2015, directions were set for 30 November for the removal of SS to Germany. Those directions were cancelled when SS commenced the present claim for judicial review.

Certain documents from SS’s Home Office file were placed before us. A detention review dated 13 October 2015 sent to SS noted that he had entered the UK clandestinely and stated that it had been decided he should remain in detention to effect his removal from the UK, where there was reason to believe he would fail to comply with any conditions attached to temporary admission or release, since:

“…

•

You have failed to observe the United Kingdom immigration laws by entering by actual clandestine means

•

You have not produced satisfactory evidence of your identity, nationality or lawful basis to remain in the United Kingdom

•

You do not have enough close ties to make it likely that you will stay in one place.”

A further detention review dated 8 December 2015 repeated these points.

The original grounds of claim focussed upon SS’s assertion that he was a child, and as such could not be removed to Germany pursuant to Dublin III and ought not to have been detained, by reason of the Secretary of State’s policy regarding treatment of unaccompanied minors as set out in chapter 55 of the EIG. However, at the hearing, the judge granted permission to SS to take an entirely new point in relation to his detention, based upon Article 28. I explain what happened in more detail below.

In his judgment in relation to Mr Hemmati and Mr Khalili (known as the Khaled judgment, after the name of another of the claimants), Garnham J dealt with a range of issues. They did not include Issues (1) and (2) as they have been formulated for the purposes of this appeal. His decision pre-dated the judgment in Al Chodor.

The issues before Garnham J in relation to Article 28 were whether it had direct effect (in the sense that it conferred actionable rights upon the appellants) and its proper construction. The predecessor Regulation, known as the Dublin II Regulation, had not included a provision equivalent to Article 28 and had generally been held not to have direct effect in this sense. However, in the light of the opinion of Advocate General Sharpston and the judgment of the Grand Chamber of the CJEU in Case C-63/15 Ghezelbash v Staatssecretaris van Veiligheid en Justitie EU:C:2016:186; EU:C:2016:409; [2016] 1 WLR 3969, which established that Article 27 of Dublin III (concerning rights of appeal in relation to a decision to remove an individual pursuant to the arrangements in that Regulation) had direct effect, Garnham J recognised that provisions in Dublin III could potentially have direct effect. Whether or not Article 28 had direct effect in a way which could benefit the appellants in the present case turned on the proper interpretation of that provision.

The argument of Mr Chirico for the appellants was that Article 28 displaces any domestic power of detention under the immigration regime where the Secretary of State intends to remove an individual to another EU Member State pursuant to the Dublin III procedure. Instead, it requires detention in such circumstances to be in conformity with the requirements of Article 28(2), even if that is not the sole reason the individual may be being detained.

Garnham J considered the proper construction and application of Article 28 at [58] to [68]. He rejected Mr Chirico’s argument regarding its effect. The judge held that Mr Hemmati and Mr Khalili could not succeed in their claim for false imprisonment, as they sought to do, simply by pointing to Article 28. In the judge’s view, Article 28(1) meant what it said, namely that a state could not hold a person in detention simply because he was subject to the procedures under Dublin III; but by implication it continued to allow a Member State to detain the person if their detention was justified on other grounds under national law. Article 28(2) qualifies the obligation in Article 28(1) and applies only in cases falling within the scope of that provision (i.e. where a person is being detained for the sole reason that he is subject to the procedure under the Regulation). Mr Hemmati and Mr Khalili could not show that they were detained for that sole reason: see [63]-[66]. As regards the application of Article 28 on the facts of their cases, the judge said this at [63]-[65]:

“63.

In the course of argument … I put to Mr Chirico [for the appellants] … that Article 28(1) appeared to me to mean that a Member State could not hold a person in detention simply because he was subject to the Dublin III procedures. Mr Chirico agreed. I suggested that that carried with it the implication that, if their detention was justified on other grounds under national law, Article 28 did not make that detention unlawful. Mr Chirico agreed that that might flow from Article 28(1) if read alone, but he said such an interpretation could not survive consideration of Article 28(2).

64.

I disagree. Further reflection has confirmed me in the view I expressed as to the natural reading of Article 28(1). And if that is right, I fail to see how Article 28(2) has caused the radical transformation of English domestic law governing the removal of non-UK citizens for which Mr Chirico contends. Article 28(2), in my judgment, applies in the circumstances covered by Article 28(1), namely where the detention is solely for the purpose of a removal under Dublin III and now when the detention is authorised under some free-standing domestic law provision.

65.

The Immigration Act 1971 provides that a person who does not have current valid leave to remain is subject to administrative removal. The Claimants fall into that category. As noted above, the 1971 Act gives powers of detention provided by paragraph 16(2) of Schedule 2. That was the power exercised in the case of these Claimants. The fact that the [Secretary of State] then decided to employ the Dublin III provisions to effect removal to Bulgaria does not affect the legality of the detention. Article 28 governs and conditions the exercise of powers to detain when Dublin III is the source of the power to detain and remove; it does not not abolish the pre-existing power under English domestic law to detain a non-UK citizen with no right to enter or remain in the UK pending their removal by whatever lawful means are available to the [Secretary of State]”.

On that footing, Mr Hemmati and Mr Khalili simply did not fall within the protection provided by Article 28. Accordingly, even if Article 28 had direct effect, which it might well do, it would not avail the appellants in that case. (On the appeal to this court, the Secretary of State concedes that Article 28, which is directly applicable since it is a provision in an EU Regulation, has direct effect in the sense of creating rights on which individuals can rely in the cases to which it applies).

In the part of his judgment dealing with the application of the Hardial Singh principles in relation to Mr Hemmati and Mr Khalili, Garnham J reviewed the evidence and found that the Secretary of State detained them both on the basis of a concern that otherwise they might abscond and evade immigration control. At [119]-[120] the judge found that the Secretary of State intended to deport Mr Hemmati and Mr Khalili; she used her powers of detention for that purpose; in each case the period of detention was modest in length and reasonable in all the circumstances; there was no want of reasonable diligence and expedition in the process of seeking to effect removal; and there was no stage in either case at which it could be said that there was no realistic prospect of removal within a reasonable time. Accordingly their claim based on the Hardial Singh principles failed. The judge observed at [120]:

“… There is no obligation on [the Secretary of State] to release from detention simply because proceedings have been begun. The commencement of proceedings may require [the Secretary of State] to reassess the position but, first, she must have a reasonable period in which to do that and, second, she may nonetheless come to the conclusion, depending on the circumstances of the case and the nature of the challenge, that removal can still be achieved in a reasonable period.”

Irwin J (as he then was) dealt with the cases of Mr Abdulkadir and Mr Mohammed. The evidence from the relevant official of the Secretary of State was that Mr Abdulkadir was detained on the ground that he posed a risk of absconding, the boxes ticked on the form given to him being indicative of such a risk. (At an early stage there was also a concern that Mr Abdulkadir might be of interest to the Secretary of State’s Counter Terrorism Unit, but that concern was disposed of fairly quickly and found not to be relevant). Similarly, the evidence from the relevant official was that Mr Mohammed was detained because he posed a risk of absconding, the boxes ticked on the form given to him being indicative of such a risk.

As explained by Irwin J at [112], Mr Southey, who appeared for Mr Abdulkadir and Mr Mohammed, submitted that Article 28 imposed requirements that there should be a significant risk of absconding; there should be an individualised assessment of the requirement to detain in order to secure the transfer under the Dublin III Regulation arrangements; less coercive arrangements must be ineffective; and any detention must be proportionate. He submitted that there had been a breach of these requirements in the cases before the court. I note that Mr Southey did not advance the argument now maintained on this appeal, to the effect that the Secretary of State was disabled from detaining the appellants by reason of a combination of Article 2(n) and Article 28 and the inadequacy of the form of legal regime in domestic law. Mr Southey also advanced additional submissions based on Chapter 55 of the EIG and the Hardial Singh principles, but these are not in issue on these appeals.

Irwin J made the assumption, for the purposes of argument, that Article 28 applies in respect of detention once there has been a decision to seek to return the individual pursuant to Dublin III. Assessing Mr Southey’s argument based on Article 28, the judge found that there was no important difference between English law (as set out in statute, the Hardial Singh principles and the requirement to adhere to policy) and the requirements of Article 28. The judge found at [124] that in the case of each of Mr Abdulkadir and Mr Mohammed, the facts of their cases gave rise to a significant risk of absconding from the beginning of their detention; that risk was in the minds of those who took the decisions to detain them; although the box for risk of absconding was not ticked in their respective forms as it should have been, it was obvious in each case that it was a major consideration – “In each case they arrived in a clandestine and illegal way and the evidence suggests that each did not voluntarily approach the authorities to claim asylum.” In each case, the risk of absconding in the face of imminent removal justified detention; the matter was kept under review in conformity with the Hardial Singh principles and the Secretary of State’s policy; and detention was both reasonable and proportionate until the time when it became clear that the judicial review proceedings could not be expedited, when the position was reviewed and they were released: [125]-[127].

After Irwin J’s judgment was sent to the parties in draft, under embargo, he was provided with a copy of the Khaled judgment of Garnham J. Irwin J observed that the issue on Article 28 appeared to have been argued in a rather different way than before him, but nonetheless said that he felt fortified in his conclusion that there had been no breach of Article 28.

Mr Howell QC’s judgment in the case of SS was concerned with the legality of his detention, dealing in that regard both with Article 28 and with the issue of SS’s age. The judge decided to treat the hearing before him as confined to the issue of liability. This is the first of the three judgments in issue on these appeals which post-dates and considers the effect of the judgment of the second chamber of the CJEU in Al Chodor. It is the first of the judgments which considers the question of what form domestic law should take in order to satisfy the requirements of Article 28 read in conjunction with Article 2(n). However, that question arose in an indirect way, and was not the subject of any pleaded claim. The judge granted permission in the course of the hearing for SS to amend his grounds for judicial review to include a particular challenge based on Article 28. Since the judge had decided that the trial before him was to be on liability only, he did not consider whether the breach of Article 28 which he identified gave rise to any entitlement to damages.

Although we do not have all the documents in the case, the way in which the new point based on Al Chodor emerged seems to have been as follows. As originally pleaded in November 2015, SS’s claim was a straightforward one: he was an unaccompanied minor who had claimed asylum; according to Article 8(4) of Dublin III, such a minor should have his asylum claim determined in the state where he had lodged his current application; therefore SS could not be removed to Germany, as the Secretary of State proposed, at any rate until a further and better investigation into his age had been carried out. Instead, he should be released from detention into the care of Oxfordshire County Council, the second defendant, which was the relevant local authority with child-care responsibilities. The relief claimed included the quashing of removal directions to Germany and damages for unlawful detention.

Permission to apply for judicial review was refused on the papers. SS proposed to make a renewed oral application for permission, but it seems that the hearing was postponed for some months. Presumably some interim arrangement had been put in place to delay SS’s removal to Germany pending resolution of the issue of permission. In the meantime, SS was released from detention on 10 December 2015 and was subject to further age assessments which again concluded that he was not a child. By a consent order dated 17 August 2016 it was agreed that SS could amend his grounds of claim to take account of these changes in circumstances. SS served amended grounds of claim dated 14 September 2016, which replaced his original grounds of claim. He pleaded that his detention from 14 September 2015 until his release on 10 December 2015 was unlawful, again on the footing that he should have been assessed to be a child at the relevant time who, according to Article 8(4) of Dublin III, could not be removed to Germany. No mention was made of Article 28. As the Secretary of State noted in her detailed grounds of defence, dated 31 December 2016, SS’s argument that his detention was unlawful was “entirely predicated on the basis that he was a minor at the time.”

On the first day of the hearing before Mr Howell QC (2 May 2017), the judge asked the parties these questions: (i) was SS detained under Article 28 of the Dublin III Regulation? (ii) if not, on what basis was he detained? (iii) if so, on what basis is the legality of that detention to be judicially reviewed under Article 9 of Directive 2013/33/EU (the so-called Receptions Directive)? The hearing continued on 3 May, apparently focusing on the age assessment issues without regard to these points. Each party put in short written submissions about these points on 4 May, and there were further oral submissions on 5 May.

In her written submissions of 4 May, Ms Sabic for SS contended that SS became subject to the transfer procedures under Dublin III as soon as the Secretary of State carried out a Eurodac search which revealed that SS had claimed asylum in other Member States and that thereafter, although SS was detained in the exercise of the Secretary of State’s power of detention set out in paragraph 16(2), such exercise was constrained by Article 28, which was directly applicable, in that the detention had to be to secure transfer under the Regulation, the detainee had to pose a significant risk of absconding, as established on the basis of an individual assessment, the detention had to be proportionate and should last for the shortest time possible, and it was necessary that other less coercive measures were not effective. It was submitted that SS was not detained because of a risk of absconding, and no evidence to that effect had been produced (I interpose that this was hardly surprising, because Article 28 and the question whether SS posed a risk of absconding had not up till then been put in issue by SS’s pleaded case). Issues (1) and (2) as formulated for this appeal, regarding Article 2(n), Article 28 and the form of domestic law, were not raised. Nor did Ms Sabic refer to Al Chodor. She contended that the judge should depart from the reasoning of Garnham J in the Khaled case.

Ms Idelbi was counsel for the Secretary of State. In Ms Idelbi’s written submissions of 4 May it was also contended that SS was detained in exercise of the Secretary of State’s power under paragraph 16(2), not under Article 28. Reliance was placed on the reasoning of Garnham J in the Khaled judgment at [58]-[69]. Ms Idelbi referred to the Al Chodor judgment in support of her argument that the power of detention in this case was not given by Article 28; she also argued that the judgment was not applicable in the present case because the nature of the challenge was that to which Issue (1) on this appeal relates, and was not concerned with the issue of direct effect. She submitted that Article 28 did not have direct effect, relying on Khaled in that regard (as noted above, this argument has been abandoned by the Secretary of State on this appeal). Ms Idelbi also objected to SS introducing any new argument based on Article 28 at this stage in the proceedings.

Accordingly, at this point, no one was contending that the Secretary of State was disabled from detaining SS by reason of Article 28 read with Article 2(n) or the Al Chodor judgment, i.e. the argument which is the subject of Issue (1) on this appeal.

On 5 May, Ms Sabic applied to amend SS’s grounds of claim to rely on Article 28. It does not appear that there was any application notice, nor does it appear that any amended pleading was actually produced. So far as one can tell from the documents, the application to amend was to take the point made in Ms Sabic’s written submissions of 4 May, to the effect that SS was not detained because of a risk of absconding and therefore his detention was unlawful by reason of Article 28 (i.e. it does not appear that she applied to amend to take the point now raised under Issue (1) on this appeal). The order eventually drawn up after the hearing stated that permission was granted to SS to amend his grounds for judicial review to include “the challenge on the basis of Article 28” (my emphasis), indicating that there was only one, i.e. that formulated by Ms Sabic in her written submissions. No other text for the amendment in question has been provided to us.

This laxness of procedure is not ideal. It means that the issues arising in respect of Article 28 were not formulated with precision so as to allow this court to follow what was going on and the nature of the argument below. It also means that the Secretary of State did not have an opportunity to put in evidence relevant to or to advance a considered case in relation to what is now Issue (1) on this appeal. Further, although, as I have explained, Ms Sabic’s application to amend was just to take her point on Article 28 (i.e. that detention of SS was unlawful, whether or not he was a child, because he posed no risk of absconding), the judge seems to have taken that as licence to embark upon a completely different analysis based on Article 28, of which the Secretary of State was not on notice before part way through 5 May. This is the analysis which is the subject of Issue (1) on this appeal.

It seems that in the course of her oral submissions in reply in support of her application to amend on 5 May, but unheralded in her written submissions of 4 May, Ms Sabic made an alternative submission that even if SS presented a risk of absconding, “Al Chodor established that his detention would have been unlawful in the absence of objective criteria defined in legislation for establishing such a risk”; and she submitted that to succeed on that argument it was necessary to depart from the reasoning of Garnham J in Khaled (the judge records this alternative submission at [35]; I am uncertain why this submission necessarily involved a departure from the reasoning of Garnham J in Khaled, since he had not addressed this point at all). After the hearing Ms Idelbi applied by email for a stay in relation to the determination of that issue either to await consideration by this court on the appeal from the Khaled judgment or at least to enable the Secretary of State to give attention to the issue and produce considered and refined submissions in relation to it. As the Secretary of State pointed out in her written submissions seeking permission to appeal after Mr Howell QC handed down his judgment containing the analysis which is now the subject of Issue (1) on this appeal, “Insofar as the court has relied on and considered Al Chodor in coming to its conclusion, the claimant did not premise their own arguments on the case, rather dealing with the authority in response.”

In the judgment, Mr Howell QC granted permission for Ms Sabic’s amendment to the grounds of claim. In his view, SS’s new case on Article 28 could be put on the basis of the facts put forward by the Secretary of State, since the judge thought that it was plain that the basis on which SS was detained from 15 September 2015 was in order to secure his transfer pursuant to Dublin III: [14]. The judge discounted the possibility that a more nuanced or fuller statement of the factual position might have been put forward by the Secretary of State had she had notice before trial that reliance was to be placed on Article 28 in line with SS’s proposed amended case: [15]-[16]. The judge refused Ms Idelbi’s application for a stay in relation to argument on Article 28: [18]-[19].

I have dealt with these procedural matters in some detail, both to explain how Issue (1) first arose in these cases despite not being pleaded by any of the appellants and also to deal with a submission made by Ms Lieven for SS on this appeal. At one point in her wide-ranging oral submissions Ms Lieven invited us to find that all officials for the Secretary of State, when deciding whether individuals subject to possible removal under the Dublin III arrangements should be detained, had no regard to the terms of Article 28. This was not a factual matter raised by the pleadings at first instance and accordingly there is no evidence about it. Nor was any finding to that effect made in any of the cases at first instance. I do not consider that we can properly make the finding which Ms Lieven proposes.

Returning to Mr Howell QC’s judgment, at [20]-[31] he set out the relevant provisions of Dublin III and those parts of the Reception Directive (Articles 9, 10 and 11) which are incorporated by reference into Dublin III by virtue of Article 28(4) of that Regulation. Article 9 of the Reception Directive provides that there should be the possibility of a speedy judicial review of the lawfulness of detention. Article 2(n) was not set out in this section of the judgment, but was recited by the judge at [41]. At [32]-[34] the judge set out the framework of domestic authority in paragraph 16(2) and para. 18B of Schedule 2 to the 1971 Act (para. 18B is concerned with the detention of unaccompanied children).

At [35] the judge referred to Ms Sabic’s main submission that SS’s detention was unlawful because he was an unaccompanied child and also set out her primary argument and alternative argument based on Article 28. Her primary argument was that the relevant power of detention stemmed from paragraph 16(2); but that power was constrained by Article 28, which was directly effective in the relevant sense; SS was initially detained under the general power in paragraph 16(2), but Article 28 applied to him as a person subject to the Dublin III procedure as soon as a Eurodac search revealed on 15 September 2015 that he had been recorded as having claimed asylum in Bulgaria, Hungary and Germany; and his detention became unlawful at that point because he was not detained because he presented a risk of absconding. Her alternative argument based on Article 28, with reference to Al Chodor, is referred to above.

At [36]-[37] the judge summarised Ms Idelbi’s submissions in reply, including that SS was detained under paragraph 16(2) of Schedule 2, not Article 28, and the legality of his detention was unaffected by the fact that the Secretary of State decided to try to use the Dublin III procedure to effect SS’s removal, in reliance on the analysis by Garnham J in Khaled. As to Article 9 of the Reception Directive, Ms Idelbi did not accept that it applied, notwithstanding the terms of Article 28(4) of Dublin III (this is another argument not pursued by the Secretary of State on this appeal); however, if it did apply, the standard of review under it of the legality of decisions to detain would involve asking whether there had been a manifest error of assessment.

At [38] Mr Howell QC analysed Garnham J’s judgment in Khaled, which he said contained three relevant conclusions regarding Article 28. The first was that Article 28 only applies where detention is solely for the purpose of a removal under the Dublin III Regulation: it does not apply if the detention is justified on other grounds: Khaled judgment, [63]-[64]. At [39] Mr Howell QC agreed with this conclusion, saying “It would be absurd to suppose that the relevant legislators contemplated that detention would not be permitted under Dublin III given article 28.2, for example, when protection of national security or public order so provides, even if there is no significant risk of him absconding.”

Also at [39], the judge disagreed with the analysis proposed by Ms Sabic for the interaction of paragraph 16(2) and Article 28: in his view,

“Article 28 governs detention in order to secure transfer procedures in accordance with Dublin III and it confers a power to do so (as the references in paragraph 3 of that article to persons ‘detained pursuant to this article’ makes plain). It is not merely concerned to impose limitations on other powers (as Ms Sabic submitted). As the Advocate-General said in paragraph [32] of her Opinion in Al Chodor, ‘article 28(2) of the Dublin III Regulation … authorises Member States to detain applicants, provided that three conditions are met.”

According to Mr Howell QC, the second conclusion of Garnham J, at [64]-[67] of his judgment in Khaled, is that Article 28 is irrelevant “when detention for the same purpose [sc. for the purpose of removal under Dublin III] is under some other power” and so “is authorised under some free-standing domestic law provision” (SS judgment at [38(ii)], referring to Khaled judgment at [64]). Mr Howell QC said that this conclusion could not stand with Al Chodor: [40] and [44].

In my view, however, this involves a misreading of Garnham J’s judgment. Although what Garnham J says is perhaps slightly elliptical, I do not think that he was saying that if the sole reason for detention was to effect removal under Dublin III then it is an answer to the appellants’ case based on Article 28 that the Secretary of State could point to a domestic law power of sufficient ambit contained in paragraph 16(2). Rather, the point he was making, in particular at [65], set out above, was that the appellants could not show that this was the sole basis for their detention. Therefore, Article 28 did not assist them. I return to this in more detail below.

According to Mr Howell QC, the third conclusion of Garnham J, at [68]-[69] of Khaled, is that Article 28 does not provide an individual with a right to challenge administrative detention in any event, because even if it were not possible to conclude with confidence that his interpretation of Article 28 was correct, it was certainly a realistic reading of the provision so that that it could not be said that the article is sufficiently clear and unconditional as to make reliance on it possible by a detained person “in the position of these claimants” (Khaled judgment, [68]). Mr Howell QC seems to have taken Garnham J to be saying that Article 28 could not have direct effect in the relevant sense. Mr Howell QC said that this conclusion also could not stand with Al Chodor: [40] and [44].

Again, however, I think this turns on a misreading of Garnham J’s judgment. Garnham J was careful at [68] to emphasise that he was talking about the effect of Article 28 in relation to “a detained person in the position of these claimants” (i.e. persons who were not being detained solely for the purpose of removal under the Dublin III procedure) and to state his conclusion at [69] in very limited terms: having first stated that Dublin III “is capable of direct effect”, depending on analysis of the particular provision in question, he said that “Article 28 does not provide an individual with a right to challenge administrative detention by the UK in circumstances such as the present” (my emphasis).

At [41]-[53] Mr Howell QC analysed the judgment in Al Chodor and other principles of EU law and made criticisms of the reasoning of Garnham J in the light of them. I analyse Al Chodor in the next section of this judgment. Mr Howell QC said that in Al Chodor it was recognised that Article 28 confers a power to detain in order to secure transfer procedures in accordance with Dublin III, failure to comply with the requirements of which will render such detention unlawful, and that it is irrelevant that national legislation may authorise a domestic authority to detain individuals for that purpose: [43]. He also said that it would be absurd and contrary to principles of EU law to suggest that detention of a person under any domestic legislation for the sole reason that he or she is subject to the Dublin III procedure would be permitted, even though that situation was covered by Article 28: [45]-[48] (his discussion also referred to cases which had not been the subject of argument before him). Therefore, again, what he described as Garnham J’s second conclusion could not stand.

However, with respect, the difficulty with this part of Mr Howell QC’s reasoning is that he has misread what Garnham J was actually saying. Contrary to Mr Howell QC’s reading, I do not think that Garnham J was saying that if the power of detention in paragraph 16(2) was exercised in the case of someone who was being detained solely for purpose of removal pursuant to the Dublin III procedure, then Article 28 would not have effect to make that detention unlawful. That would indeed have been a surprising conclusion. Instead, Garnham J’s point was that the appellants were not in that position on the facts.

In a short section of the judgment, at [54]-[56], Mr Howell QC then turned to consider whether SS’s detention was contrary to Article 28. He held that it was. Rather than address Ms Sabic’s primary submission under Article 28 (i.e. that on the facts SS did not pose a risk of absconding), at [54] the judge said that since SS was detained from 15 September 2015 in order to secure the transfer procedure under Dublin III,

“It makes no difference whether or not it was considered that there was a significant risk of him absconding. If it was considered that there was not such a risk, his detention was unlawful by virtue of Article 28.1. If it was considered that there was, it was still unlawful, as Al Chodor shows, as there were then no objective criteria for determining the risk established in a binding provision of general application.”

At [55] the judge noted that in the light of the judgment in Al Chodor the Secretary of State had made the Transfer for Determination of an Application for International Protection (Detention) (Significant Risk of Absconding Criteria) Regulations 2017 (“the 2017 Regulations”), which had come into effect at midday on 15 March 2017. This seems to have been by way of emphasis that prior to the 2017 Regulations, there had been no relevant legislative provision of general application in place setting out objective criteria for risk of absconding, as required by Article 2(n) and Article 28, as interpreted in Al Chodor.

The judge then turned to deal with other grounds on which SS’s detention was impugned, in particular on grounds of his age assessment and Article 8(4) of Dublin III, and by reference to the implicit standards of judicial review under Article 9 of the Reception Directive. The judge found that at the relevant time the Secretary of State had rationally concluded that SS was an adult (so Article 8(4) did not apply) and that the relevant standard of review under Article 9 of the Reception Directive was to ask whether the Secretary of State had made a manifest error of assessment in that regard. Since the Secretary of State’s conclusion that SS was an adult did not reflect any manifest error of assessment, these other grounds of challenge to the lawfulness of his detention were rejected: see [57]-[63]. The judge then also dismissed the separate challenge to the age assessments by Oxfordshire County Council. There is no appeal in relation to these parts of his judgment.

As a result of the way in which Mr Howell QC determined the claim based on Article 28, we do not have the benefit of any finding by him as to whether SS in fact posed a risk of absconding at the time of his detention. Also, since SS’s Article 28 ground of challenge arguing that SS did not pose a risk of absconding was only raised part way through the hearing and the judge only granted permission for the amendment in respect of it after the hearing, there was no witness statement on behalf of the Secretary of State which dealt with that issue. On the appeal, the parties simply made submissions to us based upon documents taken from SS’s file, referred to above.

In my view, it is clear from the documents that the Secretary of State considered that SS did pose a risk of absconding, even though that particular box was not ticked on the IS91R form which was given to him. The points made by the Secretary of State in the IS91R form and the detention reviews plainly related to the existence of a risk of absconding and the thrust of what was being said was that it had been decided to detain SS in order to effect his removal, since otherwise he might well abscond and not be available for removal. The position in relation to SS is closely similar to that in relation to Mr Abdulkadir and Mr Mohammed, whom Irwin J rightly found were detained by reason of an assessment by the Secretary of State that they posed a risk of absconding, even though the risk of absconding box had not been ticked on the IS91R forms given to them.

Working out the effect of what was said in the AGO and in the judgment, and how the two relate to each other, is not an entirely straightforward exercise.

The facts can be summarised as follows. The claimants were Iraqi nationals of Kurdish origin who travelled via Turkey and Greece to Hungary, where they claimed asylum. However, within days they left Hungary and travelled via the Czech Republic towards Germany, to join relatives there. The claimants were stopped by the Czech police and interviewed. The claimants again claimed that they could not be returned to Iraq on the grounds that they would be at risk of ill-treatment if sent there. A check on Eurodac showed that they had first claimed asylum in Hungary.

The Czech police took the view that there was a serious risk that the claimants would abscond unless detained, since they had no residence permit and no accommodation in the Czech Republic and had already left Hungary without waiting for a decision on their asylum claim. Accordingly, the claimants were placed in detention for 30 days pending their transfer to Hungary pursuant to the Dublin III arrangements. This detention was said to be pursuant to Czech domestic law, in the form of paragraph 129(1) of Law No. 326/1999 on the residence of foreign nationals (“paragraph 129(1)”), read in conjunction with Article 28(2) of Dublin III. Paragraph 129(1) provided in relevant part as follows:

“The police shall detain a foreign national who has entered or stayed in the Czech Republic illegally for the period of time necessarily required in order to secure transfer procedures in accordance … with directly applicable legislation of the European Union.”

The claimants brought an action to quash the decision ordering their detention. The Czech Regional Court annulled the decision, on the grounds that Czech legislation did not lay down objective criteria for assessment of the risk of absconding within the meaning of Article 2(n) of Dublin III. Accordingly, the detention was ruled to be unlawful. The claimants were released and then left the Czech Republic for an unknown destination. Meanwhile, the police appealed on a point of law to the Supreme Administrative Court of the Czech Republic. They submitted that the authority to detain under Article 28(2) was not rendered inapplicable by the mere absence in Czech legislation of objective criteria defining the risk of absconding; rather, the assessment of risk of absconding required by Article 28(2) was only made subject to three conditions, namely an individual assessment of the circumstances of the case, the proportionality of the detention and the impossibility of employing a less coercive measure, all of which had been satisfied in this case.

The Supreme Administrative Court made a reference to the CJEU to ask whether the sole fact that a law has not defined objective criteria for assessment of a significant risk that a foreign national might abscond, within the meaning of Article 2(n), renders detention under Article 28(2) inapplicable (Al Chodor judgment, [23]). The referring court was unsure whether Article 28(2), read in conjunction with Article 2(n), and/or paragraph 129(1), constituted a sufficient legal basis for detention, where the national legislation did not set out such objective criteria (Al Chodor judgment, [20]). The background set out in the reference was summarised by the CJEU at [21]-[22] as follows:

“21.

In that regard, [the referring court] points out that the language versions of article 2(n) of the Dublin III Regulation diverge. While the French- and German-language versions of that provision require a definition, laid down in legislation, of the objective criteria for the purposes of assessing the risk of absconding, other language versions require a definition of those criteria “by law (in the general sense)”, with the result that the meaning of the term “defined by law” does not follow clearly from the wording of that provision. Furthermore, the referring court notes that the European Court of Human Rights interprets the term “law” broadly, in so far as, for that court, that term is not limited solely to legislation, but also includes other sources of law: Kruslin v France (1990) 12 EHRR 547, para 29. In the context of the detention of persons who are staying illegally, it is clear from Mooren v Germany (2009) 50 EHRR 23, paras 76 and 90–97, that it is necessary to assess the quality of the legal basis, in particular in terms of clarity, accessibility and predictability.

22.

Accordingly, the referring court is uncertain whether the recognition by its settled case law of objective criteria on the basis of which the detention of persons pursuant to paragraph 129 of the Law on the residence of foreign nationals may be carried out can meet the requirement of a definition “by law” within the meaning of article 2(n) of the Dublin III Regulation, in so far as that case law confirms a consistent administrative practice of the Foreigners Police section which is characterised by the absence of arbitrary elements, and by predictability and an individual assessment in each case.”

This background is significant. The best that the Czech Republic could say in relation to whether there were “objective criteria defined by law” in the domestic immigration regime, within the meaning of Article 2(n), was that the settled case law of the Supreme Administrative Court referred to and confirmed a consistent administrative practice of the police. Some additional information about this background is set out at paras. 25-26 and 74-78 of the AGO.

On the reference the Czech and UK governments contended that the term “law” in Article 2(n) should be taken to reflect the concept of “law” as used in the ECHR, as interpreted in the case-law of the European Court of Human Rights (“ECtHR”). In the ECHR, the word “law” and its cognates is not confined to statute, but also includes unwritten law, such as a restriction imposed by virtue of the common law: see e.g. the classic judgment of the ECtHR in Sunday Times v United Kingdom (1979-80) 2 EHRR 245, at [47] (in relation to the rules of contempt of court as established at common law), and Kruslin v France, supra, at [29]. In either case, whether the law is in statute or unwritten, to qualify as “law” for the purposes of the ECHR the ECtHR has held that it must satisfy certain substantive requirements, including accessibility and foreseeability of its effects: see, e.g., the Sunday Times judgment at [49] and Mooren v Germany, supra, at [76] and [90]-[97]. Sometimes the list of substantive requirements is said to be “precision, foreseeability and accessibility”.

The Advocate General noted this submission of the Czech and UK governments at para. 40 of the AGO (EU:C:2016:865), where he also noted a further submission made by them: “They have asserted that established case-law and consistent administrative practice possess those qualities [viz. of precision, foreseeability and accessibility] in the present case.” In a footnote to that sentence, the Advocate General wrote:

“I would observe that, in its judgment of 11 April 2013, Firoz Muneer v Belgium (CE:ECHR:2013: 0411) … §59 and 60), the European Court of Human Rights held that established case-law constituted a sufficient legal basis for prolonging the detention of a person under Article 5(1)(f) of the ECHR [Article 5(1)(f) provides that a person may be deprived of liberty in accordance with a procedure prescribed by law in relation to “the lawful arrest or detention of a person … against whom action is being taken with a view to deportation or extradition”]. However, I am unaware of any case-law in which the European Court of Human Rights has regarded consistent administrative practice as a sufficient legal basis for interference with a fundamental right.”

However, for the Advocate General, the ambit of the concept of “law” in the ECHR was not important, because in his view, on proper interpretation of Dublin III, the EU legislator chose to provide more extensive protection of the right of liberty of the claimants than under the ECHR, and the meaning of “defined by law” in Article 2(n) meant that the relevant criteria for assessment had to be laid down in legislation: see paras. 43-45 of the AGO and the discussion which follows. According to the Advocate General, the requirement that the criteria for assessing the risk of absconding be “defined by law” indicates a dual objective: (a) to ensure that those criteria offer sufficient guarantees in terms of legal certainty, which is a general principle of EU law which has to be observed by Member States when taking measures pursuant to a regulation, including when exercising a discretion under a regulation, as is the case with detention under Article 28(2) (AGO, para. 62); and (b) to ensure that the discretion enjoyed by the individual authorities within a Member State responsible for applying the criteria for assessing a risk of absconding “is exercised within a framework of certain pre-determined markers” (AGO, para. 63).

In the Advocate General’s view, these two objectives are dependent on the objective criteria for assessing risk of absconding being defined in a legislative text: para. 71 of the AGO, and the following discussion. As regards objective (a) (ensuring legal certainty), it was neither appropriate nor possible to assess the relative merits of legislation, case-law and administrative practice (AGO, para. 73: i.e. in theory it seems any of these might meet the objective), so the Advocate General focused on “the specific circumstances which characterise [the particular case before him]” (AGO, para. 73). At paras. 74 to 80 of the AGO he gave reasons why in the particular case he considered that the case-law and administrative practice relied upon by the Czech government were inadequate to ensure that the standard of legal certainty was met: the presentation of the criteria in case-law was fragmentary, being scattered across a series of judgments; they had been developed on a case-by-case basis and might develop and change as new cases came forward, so they did not offer sufficient guarantees in terms of foreseeability; since the criteria in the case-law were derived from administrative practice, they might fluctuate as administrative practice develops and so lack the stability necessary to be regarded as “defined”, as required by Article 2(n); the administrative practice relied upon had not been publicised, and the CJEU had held in other cases “that mere administrative practices, which by their nature are alterable at will by the authorities and are not given appropriate publicity, do not have the clarity and precision necessary to satisfy the requirement of legal certainty”; and “it was precisely in order to limit the risk of individual administrative and judicial authorities exercising their discretion arbitrarily that the EU legislature called for that discretion to be circumscribed by statutory criteria”.

As regards objective (b) (circumscribing the administrative and judicial authorities’ discretion), the Advocate General’s analysis was not dependent on the particular characteristics of the administrative practice or case-law in question in the particular case: see footnote 59 to para. 84 of the AGO. At paras. 81-87 of the AGO he gave reasons for his view that this objective indicated that the criteria had to be laid down in legislation: legislation “offers additional assurances in terms of external control of the discretion of the administrative and judicial authorities responsible for assessing the risk of absconding and, where appropriate, ordering the detention of applicants” (para. 81); to guard against arbitrary deprivations of liberty, “it is important that the content of the criteria in the abstract and their actual application in a specific case be determined by institutionally separate authorities” (para. 82); the requirement for pre-defined, objective criteria arising from Article 2(n) “ensures that the discretion of the individual authority concerned [in making an individual assessment in a particular case] is channelled by means of general, abstract criteria that have been determined in advance by a third authority” (para. 83).

The Second Chamber of the CJEU which gave judgment in the case held that “criteria such as those listed in article 2(n) of the Dublin III Regulation require implementation in the national law of each member state” (judgment, [28]) and that “article 2(n) and article 28(2) of the Dublin III Regulation, read in conjunction, must be interpreted as requiring member states to establish, in a binding provision of general application, objective criteria underlying the reasons for believing that an applicant who is subject to a transfer procedure may abscond” (judgment, [47]). The CJEU endorsed some parts of the Advocate General’s reasoning, but it did not adopt his conclusion that the criteria had to be set out in legislation.

The CJEU noted that in some language versions of Dublin III the phrase “defined by law” in Article 2(n) used a word meaning legislation (e.g. “loi” in French), whereas in other language versions, including English, the word “law” or its equivalent could have a wider scope (judgment, [31]). Therefore the interpretation of the provision could not be determined on an exclusively textual basis, but had to have regard to the purpose and general scheme of the rules of which it forms part (judgment, [32]). The CJEU had already held in Case C-63/15, Ghezelbash EU:C:2016:409; [2016] 1 WLR 3696, particularly in light of Recital (9), that one of the objectives of Dublin III was to improve the judicial protection afforded to asylum seekers under that system (judgment, [33]; Ghezelbash was concerned with appeal rights under Article 27 of Dublin III, which were held to be enforceable by individuals). Article 2(n) and Article 28 had not appeared in the Dublin II Regulation and were also directed at improving the protection for individual applicants for asylum with regard to detention (judgment, [34]-[35]).

The CJEU noted that Article 2(n) and Article 28(2), in authorising the detention of an applicant in certain circumstances, limit the fundamental right to liberty enshrined in Article 6 of the EU Charter of Fundamental Rights (judgment, [36]; Article 6 of the Charter corresponds with Article 5 of the ECHR). The following paragraphs in the judgment are rather densely reasoned. In view of their significance for this case I set them out in full:

“37.

In that regard, it is clear from article 52(1) of the Charter that any limitation on the exercise of that right must be provided for by law and must respect the essence of that right and be subject to the principle of proportionality. In so far as the Charter contains rights which correspond to rights guaranteed by the Human Rights Convention, article 52(3) of the Charter provides that the meaning and scope of those rights must be the same as those laid down by that Convention, while specifying that EU law may provide more extensive protection. For the purpose of interpreting article 6 of the Charter, account must therefore be taken of article 5 of the Human Rights Convention as the minimum threshold of protection.

38.

According to the European Court of Human Rights, any deprivation of liberty must be lawful not only in the sense that it must have a legal basis in national law, but also that lawfulness concerns the quality of the law and implies that a national law authorising the deprivation of liberty must be sufficiently accessible, precise and foreseeable in its application in order to avoid all risk of arbitrariness: see Del Río Prada v Spain (2013) 58 EHRR 37, para 125.

39.

Furthermore, according to the case law of the Court of Justice in that regard, it must be noted that the objective of the safeguards relating to liberty, such as those enshrined in both article 6 of the Charter and article 5 of the Human Rights Convention, consists in particular in the protection of the individual against arbitrariness. Thus, if the execution of a measure depriving a person of liberty is to be in keeping with the objective of protecting the individual from arbitrariness, this means, in particular, that there can be no element of bad faith or deception on the part of the authorities: see N's case [2016] 1 WLR 3027, para 81.

40.

It follows from the foregoing that the detention of applicants, constituting a serious interference with those applicants' right to liberty, is subject to compliance with strict safeguards, namely the presence of a legal basis, clarity, predictability, accessibility and protection against arbitrariness.

41.

With regard to the first of those safeguards, it must be recalled that the limitation on the exercise of the right to liberty is based, in the present case, on article 28(2) of the Dublin III Regulation, read in conjunction with article 2(n) thereof, which is a legislative act of the European Union. The latter provision refers, in turn, to national law for the definition of the objective criteria indicating the presence of a risk of absconding. In that context, the question arises as to what type of provision addresses the other safeguards, namely those of clarity, predictability, accessibility and protection against arbitrariness.

42.

In that regard, as was noted by Advocate General Saugmandsgaard Øe in point 63 of his opinion EU:C:2016:865, it is important that the individual discretion enjoyed by the authorities concerned pursuant to article 28(2) of the Dublin III Regulation, read in conjunction with article 2(n) thereof, in relation to the existence of a risk of absconding, should be exercised within a framework of certain predetermined limits. Accordingly, it is essential that the criteria which define the existence of such a risk, which constitute the basis for detention, are defined clearly by an act which is binding and foreseeable in its application.

43.

Taking account of the purpose of the provisions concerned, and in the light of the high level of protection which follows from their context, only a provision of general application could meet the requirements of clarity, predictability, accessibility and, in particular, protection against arbitrariness.

44.

The adoption of rules of general application provides the necessary guarantees in so far as such wording sets out the limits of the flexibility of those authorities in the assessment of the circumstances of each specific case in a manner that is binding and known in advance. Furthermore, as Advocate General Saugmandsgaard Øe noted in points 81 and 82 of his opinion EU:C:2016:865, criteria established by a binding provision are best placed for the external direction of the discretion of those authorities for the purposes of protecting applicants against arbitrary deprivations of liberty.

45 It follows that article 2(n) and article 28(2) of the Dublin III Regulation, read in conjunction, must be interpreted as requiring that the objective criteria underlying the reasons for believing that an applicant may abscond must be established in a binding provision of general application. In any event, settled case law confirming a consistent administrative practice on the part of the Foreigners Police section, such as in the main proceedings in the present case, cannot suffice.

46 In the absence of those criteria in such a provision, as in the main proceedings in the present case, the detention must be declared unlawful, which leads to the inapplicability of article 28(2) of the Dublin III Regulation.”

The CJEU’s conclusion in relation to the specific facts of the case before it was clear: the case-law confirming the relevant consistent administrative practice of the Czech police was not sufficient to comply with Article 2(n): judgment, [45]. However, having used the specific word “legislation” at para. [31] of the judgment, the Court did not use that term when stating in positive terms in [42] what was required by Article 2(n) to be done at the national level, namely that the criteria should be “defined clearly by an act which is binding and foreseeable in its application” (my emphasis; NB the word “act” is not being used in the technical sense of an Act of Parliament or equivalent).

In my judgment, on a correct interpretation of what the CJEU was saying in these paragraphs, it did not go so far as the Advocate General and say that the objective criteria in relation to the risk of absconding had to be laid down in legislation, in the strict sense. Rather, it held that what was required was some “act” by relevant authorities in a Member State (being a concept wider than “legislation”) which would meet the substantive requirements of clarity, predictability, accessibility and protection against arbitrariness, as discussed in the judgment. In other words, in interpreting Article 2(n) by reference to the purpose and general scheme of Dublin III, the CJEU did not adopt either of the ordinary textual meanings of “law” or “loi” referred to in [31], but rather something of a half-way house interpretation of the phrase “defined by law” which focuses on the substantive qualities of the “act” of relevant domestic authorities relied upon in a manner similar, though not identical, to the way in which the ECtHR has interpreted the word “law” in the ECHR in the Sunday Times and other cases.

Immediately after the judgment in Al Chodor was issued, the Secretary of State promulgated the 2017 Regulations. This was on the footing that, if Article 2(n) does require legislation to be in place to define objective criteria in relation to the risk of absconding, the 2017 Regulations would constitute such legislation. This was a prudent step, given the uncertainty about what exactly the CJEU decided in Al Chodor. However, it does not affect the analysis of the position so far as concerns the EIG in this case.

By reason of the unusual procedural position which I have described above, I was uncertain at the beginning of the hearing what exactly we were being asked to decide and on what factual basis. This was clarified in the course of the hearing. Before turning to address the particular issues which the parties identified through this process, it is appropriate to say something generally about the interaction of Article 28 and domestic law regulating the treatment of immigrants and the circumstances in which they can be detained under the usual domestic regime.

An individual from another country might arrive clandestinely in the UK without any right of entry or leave to enter, intending either simply to disappear and hide or to claim asylum here. When first apprehended by the immigration authorities or the police, there will typically be uncertainty about his position and what is to happen to him. The individual might claim asylum without revealing that he has already claimed asylum elsewhere in the EU. The Secretary of State’s powers of detention under Schedule 2 to the 1971 Act may be exercised to detain the individual while checks are carried out, and with a view to possible removal in due course. The detention of an individual for these purposes was found to be lawful and compatible with Article 5(1)(f) of the ECHR in Saadi v United Kingdom (2008) 47 EHRR 17. At this initial stage, it will not be known whether the individual might be removed pursuant to the Dublin III arrangements and Article 28 has no application.

Usually, officials will conduct a search of the Eurodac database to see if the individual made a prior claim for asylum in the EU. If not, the Secretary of State will have to assess the individual’s claim for asylum in the UK. Detention may lawfully be continued under paragraph 16(2) in such a case if there is a real prospect that the individual’s asylum claim might not be accepted, there is a real prospect of removal within the scope of the Hardial Singh principles and it is reasonable and proportionate to detain, so as to ensure that removal occurs and the immigration regime is enforced effectively should the individual’s asylum claim not be upheld after examination. Obviously, if the individual poses a risk of absconding, that will be a material consideration in deciding whether detention is appropriate and proportionate. There might be other good reasons for detaining the individual pending a decision whether to grant asylum or leave to enter, such as if he posed a threat to national security or might commit crimes if released. The basis for detention in each case will be the Secretary of State’s contingent intention to remove the individual if his asylum claim is not upheld.

If the Eurodac search reveals that the individual has made a previous claim for asylum in another EU Member State, the Secretary of State has an option. Despite that previous claim, and albeit this would not be the usual course, he can choose to keep the individual here and assess his asylum claim in the UK. In that case, the Secretary of State might decide to detain (or continue to detain) the individual pending determination of that claim, in the same way as if he had not made a previous asylum claim. On the other hand, the Secretary of State might instead choose to request the other Member State to accept responsibility for examining the individual’s asylum claim pursuant to the Dublin III arrangements and, if the other Member State accepted responsibility, remove the individual to that Member State.

At the point when the Secretary of State makes a request under Dublin III it will not be known whether the requested State will accept responsibility for examination of the individual’s asylum claim. There is no guarantee it will do so. The requested State might deny that it has an obligation under Dublin III to accept responsibility or might simply act in breach of its obligations under the Regulation. This is not a fanciful scenario, as the refusal by Hungary of responsibility for SS’s asylum claim shows. Further, the individual concerned may mount a successful claim that for reasons to do with his Convention rights, such as under Article 3 of the ECHR, he cannot be sent to the EU Member State to which the Secretary of State has sent his request pursuant to Dublin III (this is what Messrs Khalili, Hemmati, Abdulkadir and Mohammed claimed, although eventually these claims failed). In such cases, the Secretary of State will not be able to remove the individual to that requested State and will have to examine the asylum claim himself. Even if a requested State accepts responsibility, it could potentially change its mind at any time until the individual is removed there. Then, again, the Secretary of State will have to examine the asylum claim himself. If that happens, and the asylum claim fails, the Secretary of State will seek to remove the individual to his state of origin.

Given these possible outcomes, even when the Secretary of State makes a request to another Member State pursuant to Dublin III it will usually be the case that his reason for wanting to detain the individual is, first, in the hope that he can be removed to the requested State under the Dublin III arrangements, but, second, if that proves not to be possible, in the expectation that the individual might be removed after the Secretary of State examines the asylum claim himself. The Secretary of State will again have a contingent intention to remove the individual, and he may wish to exercise his powers of detention to ensure that if it becomes appropriate for him to examine the asylum claim and, if that claim is rejected, to remove the individual, he is able to do so. So, again, that contingent intention may form part of the basis why the Secretary of State wishes to detain the individual. In such a case, my provisional view is that it would be difficult to say that the individual is held in detention “for the sole reason that he or she is subject to the procedure established by [Dublin III]”, within the meaning of Article 28(1). This seems to me to be the substance of the reasoning of Garnham J in the Khaled judgment, in particular at [64]-[65], set out above. It is also likely that this is what the Secretary of State would have been able to explain in the case of SS, if given permission to put in further evidence to meet the new case based on Article 28 presented at the hearing for the first time.

I am doubtful that Mr Howell QC is right when, at [39] of his judgment, he rejects the submission of Ms Sabic that Article 28 imposes limitations on other powers. The form of Article 28(1) is negative. As it seems to me, it presupposes that authorities in a Member State may have various grounds in domestic law for imposing immigration detention (which is, of course, the reality), and stipulates a limitation, namely that such powers may not be exercised “for the sole reason” that the individual is subject to the Dublin III procedure. If that is right, the logic of Garnham J’s interpretation of Article 28(2) seems to have force. It does not provide a power of detention applicable in all cases, or in all cases in which the authorities decide to attempt to use the Dublin III procedure – since Article 28 presupposes that such power exists in national domestic law. Rather, Article 28(2), which is in positive terms, provides a power to detain under the Dublin III Regulation itself, which is directly applicable, in that class of case in which the Regulation has disabled the Member State from applying its own domestic law powers of detention, i.e. where the sole reason for detention is that the individual is subject to Dublin III.

On this reading of Article 28, it only has application where the Secretary of State has decided that were it necessary for him to examine an individual’s asylum claim himself there would be no possibility of him removing that individual to his country of origin, e.g. because the Secretary of State has a policy against removal there of any person in the position of the individual in question. Outside that sort of case, the phrase “the sole reason” in Article 28(1), which (as Garnham J held) identifies the class of case in which the obligation in Article 28(2) applies, appears to me to be designed to preserve for Member States their usual sovereign rights to control immigration to them from third countries and associated powers to ensure the implementation of any decision to remove an individual back to a third country which is his country of origin. On this reading, as Garnham J said at [64], Article 28(2) has not “caused the radical transformation of English domestic law governing the removal of non-UK citizens for which [counsel for the appellants] contends.”

I say this is my provisional view because ultimately we did not hear argument on this, despite the ruling of Garnham J in Khaled. After debate at the start of the hearing before us in which the court sought to clarify what issues arose for its decision, Mr Swift told us that in the case of each appellant we should proceed on the assumption that the sole reason he was detained was that the Secretary of State wished to removal him using the Dublin III procedure (i.e. if he was not removed using that procedure, the Secretary of State would simply accept his asylum claim here without further examination and would not seek to remove him to his country of origin). Ms Lieven agreed. We were not told the reason for this concession by the Secretary of State. I assume it was made in order to present the court with a simplified set of cases in order to seek its guidance in relation to a situation in which Article 28 is applicable. It is not usual for a court to examine a case on the basis of assumed facts, but since it became evident that no party or counsel was prepared to argue the case on any other basis we agreed to proceed on the basis of this assumption.

For her part, Ms Lieven said that she accepted that if there was some other reason than risk of absconding why the Secretary of State wished to detain an individual seeking asylum, such as on grounds of national security, that would not be rendered unlawful by Article 28. I am not sure what analysis she would adopt to reach this conclusion.

I can see the Secretary of State could rely on other grounds for detention if there is more than one immigration reason why the Secretary of State wishes to detain, i.e. it is not solely because he wishes to use the Dublin III procedure but also because he might wish to use his powers of detention in support of removal to the individual’s country of origin should that prove to be necessary: see above. On that analysis and Garnham J’s interpretation of Article 28, Article 28(2) simply would not apply in such a case. But it was not clear to me that Ms Lieven did mean this.

Alternatively, if Ms Lieven meant to say that, contrary to the view of Garnham J, Article 28(2) applies in any case where the Secretary of State wishes to employ the Dublin III procedure to remove an individual (even a case where he has a contingent fall-back intention to decide the asylum claim himself, if the Dublin III procedure proves not to be workable for some reason), perhaps she might argue that Article 28(2) only applies to impose limits as regards a decision to detain on grounds of risk of absconding, but allows a Member State to rely on any other grounds it might have for detention pending removal, such as threat to national security. I think this is what Mr Howell QC had in mind at [39] in his judgment in SS’s case, quoted above. But I am doubtful that this is a correct reading of Article 28(2). To my mind, Article 28(2) seems more naturally to be aimed at regulating completely the grounds on which detention is permissible in any case where that provision applies, thereby excluding any power for a Member State to rely on further grounds for detention, not specified in Article 28(2). It is difficult to see why, in cases where Article 28(2) applies, the EU legislature should have been concerned to regulate the power of detention so far as regards the risk of absconding, but not as regards other policy grounds that a Member State might wish to rely upon such as threat to national security or public order, which would appear to pose a still greater problem in terms of protection against arbitrariness - which is central to the CJEU’s interpretation of Article 28(2) read with Article 2(n) in Al Chodor. Again, it was unclear whether, despite this difficulty, Ms Lieven was intending to support such an interpretation of Article 28(2).

It emerges from this discussion that it is likely that these issues will have to be examined more fully in a future case. If that is to happen, it would be highly desirable that this be done on the basis of clear and complete findings of fact. I now turn to examine Issue (1) and Issue (2) on the basis of the factual assumption both sides have invited us to make.

The interpretation of the judgment in Al Chodor set out above is of significance for the legal analysis under Issue (1). In the present case, for the purposes of his submission regarding compliance with Article 28(2) read with Article 2(n), the Secretary of State does not rely upon criteria relating to risk of absconding set out in case-law, nor upon administrative practice of the kind relied upon by the Czech government in Al Chodor. Instead, he relies upon the statement of policy in the EIG, which is said to be the relevant “act” within the sense of that term in para. [42] of the judgment in Al Chodor. Neither the Advocate General nor the Court in Al Chodor addressed this type of situation.

The policy in the EIG has been promulgated and published to the world. The Secretary of State points out that in domestic law a published policy has significant legal effects, according to the doctrine of legitimate expectation and the related obligation of consistency in application of policy as a matter of good government: see e.g. Mandalia v Secretary of State for the Home Department [2015] UKSC 59; [2015] 1 WLR 4546, at [29]-[31]. Published policy operates as a substantial safeguard against arbitrariness in decision-making, over and above the safeguards provided by general rules of public law. This legal framework in domestic law to give substantive legal effect to a statement of policy means that such a statement will constitute a sufficient legal basis capable of qualifying as “law” as that concept is used in the ECHR: see, in particular, R (Munjaz) v Mersey Care NHS Trust [2005] UKHL 58; [2006] 2 AC 148, [34] (Lord Bingham of Cornhill), [91]-[94] (Lord Hope of Craighead) and [103] (Lord Scott of Foscote). That case concerned rules for segregation of mental patients in hospital according to the policy adopted by the NHS Trust and whether segregation of patients under them was “in accordance with the law” for the purposes of Article 8(2) of the ECHR. The House of Lords held that the “in accordance with the law” requirement was satisfied. As Lord Bingham said at [34],

“… The procedure adopted by the Trust does not permit arbitrary or random decision-making. The rules are accessible, foreseeable and predictable. …”

The Secretary of State submits that the EIG satisfies the substantive requirements identified in the Al Chodor judgment so that it qualifies as “law” for the purposes of the phrase, “defined by law”, in Article 2(n).

Issue (1) includes two sub-issues: (a) whether a statement of policy such as the EIG is in principle capable of qualifying as “law” for the purposes of Article 2(n); and (b) if it is so capable, whether in fact the EIG sufficiently set out objective criteria regarding reasons to believe that an individual may abscond so as to satisfy Article 2(n).

As to sub-issue (a), in my judgment a statement of policy such as the EIG is in principle capable of qualifying as “law” for the purposes of Article 2(n).

Usually, statements of the doctrine of legitimate expectation and the related obligation of consistency in application of policy explain that a public authority must comply with its own policy, unless there is good reason not to (see Mandalia, above). It might be argued that this latter qualification means that the list of criteria in a policy statement regarding grounds for detention could never be closed, and would always be vulnerable to future development as a matter of administrative practice – something which was found in Al Chodor to be incompatible with the concept of “law” as used in Article 2(n). However, in a case where Article 28(2) applies, I consider that the obligation of the Secretary of State (and his officials) to comply with that provision would mean that he could not lawfully depart from his own statement of policy. This means that the qualification which usually arises under general domestic public law would have no practical application. Put another way, any attempt to depart from the policy statement could not be for a good reason, since it would be incompatible with the requirements of EU law to introduce further, unstated criteria in relation to the risk of absconding. Accordingly, the list of criteria in the policy statement would be a closed one, as required by Article 2(n) as interpreted in Al Chodor.

A policy statement is thus capable in principle of satisfying the objective of legal certainty which is taken to inform the concept of “law” in Article 2(n). In fact, the doctrine of protection of legitimate expectations is also well-established in EU law, as an aspect of the principle of legal certainty which is inherent in EU law: see the discussion in T. Tridimas, The General Principles of EU Law, 2^nd ed., chapter 6. This supports the view that a policy statement which gives rise to an enforceable legitimate expectation in domestic law is capable of falling within the concept of “law” in Article 2(n), so far as the substantive aspect of legal certainty is concerned.

Therefore, a policy statement such as the EIG, published to the world, is in principle capable of satisfying the requirements of “legal basis”, “clarity”, “predictability” and “accessibility” referred to in the Al Chodor judgment at [40]. This is also relevant to whether it is capable of meeting the other substantive requirement mentioned in that judgment, namely “protection against arbitrariness”. That is because those substantive qualities (if satisfied by the contents of a particular policy statement), when taken in the context of the domestic doctrine of legitimate expectation and the domestic obligation of compliance with policy and having regard to the availability of judicial review, are also a powerful substantive guarantee against arbitrariness in decision-making concerning an individual.

As a matter of principle, a policy statement such as the EIG is capable of setting out “a framework of certain predetermined limits” regarding the criteria relating to the risk of absconding, and defining them clearly in a way which is binding and foreseeable in its application: see Al Chodor judgment, [42]. Lord Bingham’s speech in Munjaz provides strong support for this view. Such a policy statement is “a provision of general application” which “sets out the limits on the flexibility of [the] authorities in the assessment of the circumstances of each specific case in a manner that is binding and known in advance”: see the Al Chodor judgment, [43]-[44]. It is true that a policy statement can be changed, but such changes will only apply prospectively. So an individual who wishes to make representations regarding whether he poses a risk of absconding will know what the relevant current policy statement is, as will the official who makes the decision regarding detention in his case. As explained above, the official will be bound to act in accordance with the policy statement.

In the Al Chodor judgment, at [44], the CJEU also made the point, with reference to paras. 81 and 82 of the AGO, that criteria established by a binding provision are “best placed for the external direction of the discretion” of the relevant national authorities. A policy statement is not an instrument which is established externally, in the sense of by a body external to and separate from the Secretary of State. Obviously it is not primary legislation laid down by Parliament.

However, I do not consider that this is fatal to the Secretary of State’s case. In a real sense, which is the relevant one for present purposes, a published and binding policy statement does provide “external direction” for the exercise of discretion by the relevant authorities who decide on the question of detention. In all (or almost all) cases the relevant authorities will be officials who will not have been responsible for drafting the policy statement, and are simply required to follow and apply it. As a matter of the Carltona doctrine (Carltona Ltd v Commissioners of Works [1943] 2 All ER 560), in constitutional theory in domestic law the decisions of the Secretary of State’s officials count as his. But such a principle of domestic law does not govern the meaning or application of a rule of EU law such as that at issue in this case. Still more importantly, even if the Secretary of State were himself the decision-maker, his own policy statement would be an instrument which was binding upon him as a matter of domestic law. It would be an instrument setting out a statement of the applicable criteria in advance of his decision which would control his own exercise of discretion in the same way, in effect, as legislation would do. In the substantive sense intended in the Al Chodor judgment at [44], therefore, the policy statement would provide “external direction of the discretion” of the Secretary of State.

That this substantive sense is the relevant sense to be given to this part of the Al Chodor judgment is also supported by consideration of the different forms of legislation which exist. As I have said, the CJEU did not say that the relevant criteria had to be specified in legislation, let alone primary legislation. Most if not all Member States have forms of subordinate or delegated legislation, as indeed exist in the scheme of EU law as well. In English law, relevant subordinate legislation may be made by a Secretary of State and is fully binding upon that Secretary of State if he acts within its scope, just as primary legislation would be. Indeed, even in relation to primary legislation, in the UK as in the EU and other Member States, it will often be the relevant government department which proposes and drafts the legislative instrument. Given this background, I do not think that in para. [44] in the Al Chodor judgment the CJEU can plausibly be said to have laid down a principle regarding the formalities by which a relevant “act” is promulgated to the effect that the relevant national authority responsible for taking detention decisions must have played no part in the promulgation of the instrument or “act” in question.

As regards sub-issue (b) above, the critical section of the EIG is para. 55.3.1 (“Factors influencing a decision to detain”) set out above. It begins by saying that “All relevant factors must be taken into account” and sets out what appears to be a non-exhaustive list introduced by the word “including”. However, in my view, in context, the reference to “relevant factors” is simply a reminder to officials to bear in mind the general principles discussed previously in the EIG – in particular in para. 55.3 – and is not a licence to have regard to other specific factors which have not been listed in para. 55.3.1.

In my view, the factors set out in the first seven bullet points in para. 55.3.1 are specific objective criteria relevant to the formation of a belief in an individual case as to whether an individual may abscond. If they stood as a discrete list, I consider that they would satisfy Article 2(n). The question, then, is whether the fact that the list of factors in the first seven bullet points appear in a list with other factors means that a breach of Article 2(n) arises.

The eighth bullet point relates to the risk of offending or harm to the public. In my opinion, this is primarily directed to general public policy reasons (apart from any risk of absconding) why it might be considered appropriate to exercise a power of immigration detention pending removal of an individual from the UK: see R (Lumba) v Secretary of State for the Home Department [2011] UKSC 12; [2012] 1 AC 245, [106]-[108] and [121] (Lord Dyson JSC). However, it also has a bearing on the risk of absconding, as explained in Lumba at [109] by Lord Dyson, since if a person offends there is a risk that he will abscond so as to evade arrest.

The factors in the ninth, tenth and eleventh bullet points (respectively, is the individual under 18? does he have a history of torture? and does he have a history or physical or mental ill health?) do not relate to risk of absconding. In context, it is clear that they are included as reminders that there is a separate paragraph in the EIG dealing with persons considered unsuitable for detention (i.e. para. 55.10, to which cross-reference is made in para. 55.3.1), which covers these classes of case and to which regard should be had if someone is found to fall within any of these bullet points.

The list of factors in para. 55.3.1 of the EIG is clearly not focused exclusively on criteria relevant to risk of absconding for cases covered by Article 28, but functions as a more compendious list relevant both to Article 28 cases and to other cases in which it may be appropriate to consider detention for immigration purposes. In my view, this is not incompatible with Article 2(n). Article 2(n) does not say that the objective criteria have to be set out in a stand-alone document or format, focused exclusively on Article 28, nor that there has to be express reference to that provision. Moreover, neither the Advocate General nor the CJEU in Al Chodor suggested that this was the case, even though, if correct, that would have provided a simple and straightforward answer to the question referred to them. An individual who falls within the scope of Article 28 and Article 2(n) can readily identify which objective criteria in the list in para. 55.3.1 are relevant to the question which arises where Article 28(2) applies, namely whether he will be assessed to pose a risk of absconding. The objectives of legal certainty and predictability are met. Similarly, the discretion regarding detention exercised by an official addressing a case which falls within Article 28(2) will be constrained by the factors in the list which are relevant to the assessment of a risk of absconding. Accordingly, the objective of constraint of discretion is met.

It is also noteworthy that although this part of the EIG has been in place for many years, no one has brought a claim to challenge the lawfulness of the guidance in para. 55.3.1 on the grounds that it is so poorly formulated that it is apt to mislead as to the true legal position. This is a well-established basis for judicial review of policy statements: see e.g. Gillick v West Norfolk and Wisbech Area Health Authority [1986] AC 112; R (Suppiah) v Secretary of State for the Home Department [2011] EWHC 2 at [137] per Wyn Williams J (“… a policy which is in principle capable of being implemented lawfully but which nonetheless gives rise to an unacceptable risk of unlawful decision-making is itself an unlawful policy”). Given the thousands of cases to which chapter 55 of the EIG is applied each year, if para. 55.3.1 were thought to be capable of being misleading one would have expected such a challenge to have been brought at some point. So far as I am aware, none has been. In my opinion, this reflects the reality of the position, that para. 55.3.1 is not misleading and that no one practising in the immigration field has seriously considered that it is. Paragraph 55.3.1 of the EIG is not apt to mislead an official who seeks to apply Article 28. It provides that official with appropriate and clear guidance as to the relevant criteria to be applied in respect of the risk of absconding.

It follows from the analysis above that I would dismiss the appeals by Mr Hemmati and Mr Khalili (subject to the discussion below regarding their appeal based on the Hardial Singh principles) and the appeals by Mr Abdulkadir and Mr Mohammed and would allow the appeal by the Secretary of State in the case of SS. In each case, the Secretary of State assessed that the individual posed a risk of absconding, as he was entitled to do, and I would hold that there is no incompatibility in respect of his detention by reason of Article 2(n). However, as my lords disagree with this, it is necessary to proceed to consider the question of damages under Issue (2).

I address this section of the judgment on the footing that the EIG is not an instrument which satisfies the requirements of Article 2(n), with the result that the Secretary of State is unable to show that Article 28(2) is applicable. I also accept in that regard that Article 28, read with Article 2(n), creates rights which are directly applicable and are enforceable by individuals. We have been invited to address Issue (2) by both sides, even though none of the courts below did so.

Neither Article 28 nor any other provision in Dublin III stipulates that the remedy for breach of Article 28(2) read with Article 2(n) should be damages. Mr Southey QC for the appellants made it clear that he did not suggest that the EU law principles of effectiveness and non-discrimination in relation to remedies lead to the conclusion that substantive damages should be awarded to the appellants. Mr Southey’s submission is that the effect of Article 28, read with Article 2(n), disables the Secretary of State from relying on any justification for the detention of the applicants, with the result that the detention of the appellants is unjustified for the purposes of domestic law and substantive damages are payable under the domestic law of false imprisonment in respect of such detention in each case. Mr Southey emphasises that the domestic “tort of false imprisonment has two ingredients: the fact of imprisonment and the absence of lawful authority to justify it”: see Lumba at [40] per Lord Hope.

Against this, Mr Swift submits that the proper approach in respect of damages is to ask whether the relevant criteria in EU law for an award of damages in respect of a breach of EU law have been satisfied, in particular whether the relevant breach of EU law was “sufficiently serious” within the meaning of the leading judgment of the ECJ in the Factortame case at [55]: “the decisive test for finding that a breach of Community law is sufficiently serious is whether the member state … concerned manifestly and gravely disregarded the limits on its discretion.” In this case, each of the appellants was detained on the basis of a legitimate individualised assessment by the Secretary of State that he posed a risk of absconding, which was proportionate and lawful so far as paragraph 16(2) was concerned, so it could not be said that there has been any violation of the substance of what Article 28 requires (see, e.g., the assessment by Irwin J in his judgment below, in relation to which there is no appeal). The appellants’ claim turns purely on the alleged failure of the UK to adopt a particular form of law when implementing Article 2(n). Under Article 2(n), and as confirmed in the judgment in Al Chodor at [28], it is a matter for the discretion of a Member State to decide what “objective criteria” should be stipulated in its national law and, subject to certain limits, how that should be achieved. Mr Swift says that the failure of the UK to set out the list of “objective criteria” regarding the risk whether an individual may abscond (Article 2(n)) in legislation in the strict sense (rather than in a binding statement of policy) or in a discrete list relating only to that risk, rather than covering other factors as well, does not qualify as a “sufficiently serious breach” within the test laid down in Factortame. Therefore, no award of damages is justified. Further and in the alternative, Mr Swift submits that if the UK had promulgated legislation in the strict sense with a list of objective criteria in relation to risk of absconding as set out in para. 55.3.1 of the EIG, but in a discrete list referable only to risk of absconding, it is clear that each of the appellants would have been detained on the basis of such a risk. Accordingly, again, Mr Swift says, no substantive damages are payable in respect of their detention.

So far as Mr Swift’s first argument is concerned, the question is how principles of the domestic law of false imprisonment relate to and are affected by EU law. I have not found this question easy to answer. Mr Southey says that although the Secretary of State appeared to be clothed by paragraph 16(2) with the power to detain, Article 28 disabled him from relying on that power because the form in which the “objective criteria” regarding risk of absconding were set out did not satisfy Article 2(n). According to this argument, the appellants are entitled to rely on EU law as a foundation for a claim for substantive damages under domestic law, and that is so even if it the relevant norm of EU law in Article 2(n) was unclear and it cannot be said that the state manifestly and gravely disregarded the limits of its discretion under that provision as to the manner in which it could lay down “objective criteria” so that they would be “defined by law”. Against this, however, it is argued that the critical legal ingredient which is said to make the detention unlawful is EU law, and EU law itself stipulates the relevant test for an award of substantive damages in such a case, namely the “sufficiently serious breach” test.

In my judgment, in these circumstances it is the “sufficiently serious breach” test set out in EU law which governs whether substantive damages are payable in relation to the detention of each of the appellants. In my opinion, Mr Southey’s argument is too simplistic.

The judgment of the ECJ in Factortame is of fundamental importance here. It recognises that in many cases the precise effect of EU law may be unclear until that effect is authoritatively established by a judgment of the ECJ (now CJEU). The “sufficiently serious breach” test for liability in damages was laid down in recognition of the reality of this situation. It is not only individuals which are entitled to the benefit of legal certainty and predictability in the application of EU law, but also Member States which are subject to directions laid down in EU law with which they are required to comply. The risk of liability to pay damages by reason of a violation of EU law should not have a “chill effect” in relation to the choices a Member State makes as to how it will carry out its legislative functions in cases other than those where there has been a manifest and grave disregard of the limits on the exercise of its powers, as laid down in EU law: see Factortame at [45]-[47]. Putting it another way, the fair balance between the general public interest represented by the state and the interests of individuals as regards the proper implementation of EU law takes account of the reality, in many cases, of a degree of lack of clarity in EU legislation, so that an individual will only be entitled to damages where the relevant rule of EU law is in fact clear. As well as individuals, Member States are entitled to the protection of rule of law values inherent in EU law, and this is reflected in the test of “sufficiently serious breach” laid down in the Factortame judgment.

In my view, the problem with Mr Southey’s argument is that he seeks to circumvent this balance of interests which is inherent in the approach adopted by EU law to the award of damages, by using a breach of (unclear) EU law to ground a substantive claim to damages in domestic law. The domestic law of false imprisonment imposes liability on a strict basis, which is justified because liability is generally imposed on the basis of a much clearer delineation of rights and obligations which domestic law sets out, as compared with some EU legislation. The domestic law of false imprisonment has been framed without reference to the particular problems to which EU legislation gives rise and it ignores the fair balance of interests which EU law aims to achieve. In my opinion, Mr Southey’s argument gives disproportionate effect to Article 2(n), in seeking to base a claim for substantive damages for false imprisonment upon it. In my judgment, it is not appropriate or fair to resort to the strict approach to the imposition in domestic law of an obligation to pay substantive damages for false imprisonment on the basis of the later clarification by the CJEU in Al Chodor of the unclear norm of EU legislation contained in Article 2(n) in this case.

In effect, a question of choice of law arises so far as any entitlement to substantive damages is concerned. In my view, the appropriate law to govern that question, which arises solely because of a breach of EU law, is EU law. Judged by reference to ordinary domestic law, the power of detention under paragraph 16(2) is to be taken as having been lawfully exercised in these cases and there was no false imprisonment.

In R v Secretary of State for Transport, ex p. Factortame Ltd [1990] 2 AC 85, at 140, Lord Bridge explained the effect of section 2(1) of the European Communities Act 1972 in relation to directly enforceable Community (now EU) rights. In line with that explanation, paragraph 16(2) is to be taken to be qualified by the impact of EU law as if a further provision were incorporated in the 1971 Act alongside it, which in terms enacted that the power of detention in paragraph 16(2) was to be without prejudice to the directly enforceable Community rights of individuals under Article 28, read in conjunction with Article 2(n). The principles of EU law according to which damages are payable in respect of such directly enforceable rights are themselves an inherent part or dimension of those directly enforceable rights. In my view, section 2(1) of the 1972 Act has the effect that an individual should have all the rights and remedies afforded to him by EU law in relation to such rights, but does not require more than this.

Therefore, I consider that in this case the power of detention in paragraph 16(2) is qualified in the following way. An individual within the scope of Article 28(2) who has been detained in purported exercise of the power in paragraph 16(2) can claim before a court that he should be released, by reason of the applicability of Article 28(2) and the UK’s non-compliance with Article 2(n). If the court determines that the case falls within Article 28(2) and that there has been non-compliance with Article 2(n), it should order his release - and that would be so even if the claim were brought at a time when the true effect of Article 2(n) remained unclear and the relevant clarification would only be provided by the court’s own ruling, perhaps with guidance from the CJEU after a reference. However, in so far as the individual claimed damages for a period of detention up until release pursuant to the order of the court, the authorisation under paragraph 16(2) for that period of detention would only be taken to have been qualified to the extent that the individual could show that there has been a “sufficiently serious” breach of Article 2(n). As regards the damages claim, this is the relevant aspect of the individual’s directly effective EU rights which qualifies the previous exercise of the power under paragraph 16(2) and in respect of which the national court would grant relief pursuant to section 2(1) of the 1972 Act. Section 2(1) is in this way apt to accommodate the modulated response of EU law to the type of situation which has arisen in this case, and does not commit a domestic court to give the disproportionate effect to Article 2(n) for which Mr Southey contends.

Depending on the view one takes about the interaction of paragraph 16(2) and Article 28 in this sort of case, an alternative analysis would be as follows. In the present cases, on the factual assumptions we are invited to make, it might be said that Article 28(1) excludes any application of the domestic law power of detention in paragraph 16(2). In each case, the individual’s detention is therefore completely governed by EU law, as set out in Article 28(2) read in conjunction with Article 2(n). Since his detention is governed by EU law, he is only entitled to claim damages if he has a good claim for damages under EU law, which depends upon whether there has been a “sufficiently serious” breach of Article 2(n).

I think it is significant that both modes of analysis lead to the same result. This underlines the fact that from an EU law perspective the substantive conclusion is that the individuals should only be entitled to damages if the “sufficiently serious breach” test is satisfied. Modification of domestic law is only required to give effect to rights in EU law to the extent that EU law requires those rights to be respected, including in that concept the extent to which they are required to be respected by an award of damages to vindicate them in relation to past unlawful conduct by the State. EU law does not require that Member States should do more than this, and there is no inexorable logic of English law which requires damages to be awarded on a more generous basis in relation to a breach of EU law.

The disconnection in principle between EU law as set out in Article 28 and Article 2(n) and the domestic law of false imprisonment in this area is also well illustrated in Mr Abdulkadir’s case. As noted above, on 13 November 2015 Mr Abdulkadir applied to the FTT for bail, and his application was refused. From that point, since his continued detention was the result of a court order, it would appear that he had no claim for damages in domestic law for false imprisonment (I should mention that this was another dimension of the case which was not addressed in argument, as a result of the procedural position described above). Yet despite the FTT having made an order, he would have a claim for damages for breach of EU law if that breach was “sufficiently serious”: see Case C-224/01 Köbler v Austria [2003] ECR I-10239.

In my opinion, the position as regards the award of damages in domestic law for breach of Convention rights under the HRA supports the conclusion above. In R (Greenfield) v Secretary of State for the Home Department [2005] UKHL 14; [2005] 1 WLR 673 a prison governor conducted the hearing of a disciplinary charge against a prisoner, found him guilty and ordered the prisoner to serve 21 additional days of imprisonment. The prisoner brought a claim to contend that the hearing before the governor infringed his right under Article 6 of the ECHR to a fair hearing before an independent tribunal in respect of a criminal charge. The prisoner claimed damages for violation of Article 6. The House of Lords found that there had been a breach of Article 6 as alleged, but held that the finding of violation was sufficient just satisfaction and that no substantive damages equivalent to damages for false imprisonment should be awarded. The House of Lords rejected the submission of the prisoner that damages should be awarded as just satisfaction under section 8 of the HRA assessed by reference to levels of damages for false imprisonment in domestic law: see in particular [19] per Lord Bingham of Cornhill. It could have been said that there was no proper justification for the imposition of the additional 21 days imprisonment, because it was imposed by an official who ought not to have been involved, according to the requirements of Article 6. However, the balance of interests between individuals and the state under the ECHR regime as regards the award of damages is different from that under the ordinary domestic law of false imprisonment, and the latter could not be taken to govern the former. Similarly, in the present case, the balance of interests under EU law is different from that under ordinary domestic law, and the latter should not be taken to govern the former.

In my judgment, any breach of Article 2(n) by the UK in this case does not satisfy the “sufficiently serious breach” test. In advance of the clarification provided by the judgment in Al Chodor, it was not unreasonable for the Secretary of State to take the view that the EIG sufficiently complied with the requirements of Article 2(n). In Al Chodor, the CJEU itself found it difficult to spell out precisely what the requirements of Article 2(n) were. Even after that judgment, the position in relation to a list of objective criteria in a statement of policy such as the EIG remained uncertain. Further, I think it is very revealing that none of the eminent counsel experienced in the field of immigration law who have appeared for the appellants in these three cases considered that there was any relevant argument available to them on the basis of Article 2(n) in respect of their clients’ detention until after the judgment in Al Chodor was handed down. Indeed, until then, it does not appear that anyone thought that there might be a problem in the UK as regards compliance with Article 2(n). We have been shown no academic commentary in the period since 2014 which would indicate that there was any appreciation that any problem existed. The EU Commission did not make any complaint about the UK’s compliance with Dublin III so far as Article 2(n) was concerned.

Mr Southey sought to suggest that the Secretary of State is not entitled to the protection of the “sufficiently serious test” in Factortame because he had not promulgated the EIG as a response to Dublin III. He referred to Joined Cases C-178, 179 and 188-190/94 Dillenkofer v Federal Republic of Germany [1997] QB 259, at [29].I do not accept this submission. The application of the Factortame test does not depend upon whether a Member State happens to change its law in response to new EU legislation or simply assesses that its existing legal regime is already sufficient to comply with that legislation so that no change is required. In my view, Dillenkofer, which concerned a situation in which a Member State had no relevant law in place and completely failed to do anything in respect of implementation of a Directive, has no bearing on this point.

Accordingly, even if the appellants succeed on Issue (1) (contrary to my view above), I would hold that they are not entitled to substantive damages in respect of their detention.

I do not find Mr Swift’s alternative argument persuasive. If Mr Southey is right that the question of damages is to be approached purely as a matter of domestic law, I do not think that Mr Swift’s alternative argument assists the Secretary of State. The Secretary of State would have had no proper justification for the detention on the state of the law at the time, and in my view could not avoid liability by speculating that he might have had a proper justification had the legal framework been different. Similarly, if the question of damages is to be addressed as a matter of EU law and the breach is “sufficiently serious”, the Secretary of State could not avoid liability in damages by speculating that in a hypothetical world there would have been no “sufficiently serious” breach.

For the reasons given above, even if the appellants succeed on Issue (1), as regards Issue (2) I would hold that they are not entitled to an award of damages by reason of any breach of Article 2(n) and Article 28.

Finally, I turn to the discrete points based on Hardial Singh advanced by Mr Chirico on behalf of Mr Hemmati and Mr Khalili. In my judgment, there is nothing in them. I agree with Garnham J, for the reasons given by him, that the detention of these appellants in no way infringed the Hardial Singh principles. The judge correctly directed himself regarding the law and reached a conclusion which he was fully entitled to reach. It would be contrary to principle for this court to interfere with his judgment: see R (Muqtaar) v Secretary of State for the Home Department [2013] 1 WLR 649, [46]; and Fardous v Secretary of State for the Home Department [2015] EWCA Civ 931, [34].

In any event, I agree with the judge’s assessment. Mr Hemmati and Mr Khalili were detained for proper reasons; each of them was assessed to pose a risk of absconding, which assessment was a rational and justified one; and throughout the period of their detention there remained a real prospect that they could eventually be removed.

Mr Chirico made three criticisms of the judge’s assessment. First, he said that the judge erred by failing to treat the Hardial Singh principles as informed by the requirements of Article 28 and Article 2(n). However, in my view there was no error here. The Hardial Singh principles are well settled principles of domestic law arising as a matter of statutory interpretation of the powers of detention in the 1971 Act (specifically, as concerns the present case, paragraph 16(2)). I can see no basis for saying that these principles of domestic law are qualified by Article 28 and Article 2(n).

Secondly, Mr Chirico submitted that Hardial Singh requires an individualised assessment whether detention is justified in the particular case. I note that this is entirely uncontroversial. He then says that there was no individualised assessment in the cases of Mr Hemmati and Mr Khalili. In particular, he says that the Secretary of State did not properly factor into the assessment in each case that there was a degree of compliance with immigration conditions when they were not in detention. However, in my view there were strong positive features of their cases to indicate that they did indeed pose a risk of absconding, on which the Secretary of State was entitled to rely and did rely. There is nothing to indicate that the Secretary of State failed to have regard to certain limited positive aspects of their cases which were clear on the file when reaching the overall assessment that they did indeed pose a risk of absconding. It is untenable to suggest that there was no individualised assessment in their cases.

Thirdly, Mr Chirico submitted that Garnham J wrongly assumed in a generic fashion that the Secretary of State was entitled to continue to detain Mr Hemmati and Mr Khalili for a period after they had issued their judicial review proceedings, after permission was granted for those proceedings and even when it was decided that no application would be made for expedition of those proceedings. Again, however, in my view there was no error by the judge.

Contrary to Mr Chirico’s contention, the judge made no generic assumption. He rightly held that the detention of the appellants was justified in law in conventional fashion under the usual Hardial Singh principles. The fact that the appellants had brought judicial review proceedings to challenge their removal to Bulgaria pursuant to Dublin III, which might take some time to resolve, did not mean that their immigration detention became unlawful. There remained a real prospect that they would be removed to Bulgaria, since they could fail in their judicial review challenge (as in fact they did). Even if the judicial review to challenge removal to Bulgaria were successful, there was also a real prospect that they could be removed to their respective countries of origin if their asylum claims were rejected in the UK after examination by the Secretary of State. Even though the judicial review proceedings might take some time to resolve, the timetable for prospective removal did not exceed what was reasonable in the circumstances: compare, e.g., R (MH) v Secretary of State for the Home Department [2010] EWCA Civ 1112, when a far longer period of detention was found to be lawful for so long as there remained a real prospect of removal (see [66]).

In the present case, the Secretary of State decided to release the appellants a short time after it emerged that there would be no expedition of the judicial review proceedings. Mr Chirico suggested that the inference from this is that when it became clear that there would be no order for expedition, the detention became unlawful. In my view, this is not correct.

The Secretary of State was not under any obligation to apply for expedition in relation to the judicial review claim. Her decision not to do so seems to have been based on a realistic assessment that the court would be unlikely to order expedition. However, it does not follow that just because expedition was not going to be ordered that the detention of the appellants became unlawful. As I have said, there remained a real prospect of removal of the appellants and in my view the Secretary of State would have been entitled to continue to detain them while the judicial review proceedings remained on foot. On the other hand, the Secretary of State was not obliged to do that. There might have been a number of reasons for deciding to release the appellants at this stage. For example, the Secretary of State would have been entitled to make an assessment that, on balance, the overall public interest pointed in favour of release after it became clear there would be no order for expedition, having regard to the cost to the public of detention or out of a desire to avoid litigation on the issue. It does not follow from the fact that the Secretary of State decided to release the appellants that it would have been unlawful to keep them in detention. In any event, the Secretary of State was entitled to a period of time to assess the position once it was clear that there would be no application for expedition: compare the position in the MH case, at [115] in my judgment at first instance ([2009] EWHC 2506 (Admin)), recited by the Court of Appeal at [53] and endorsed at [69]. The short period of time to review the position in the present cases did not exceed what was reasonable in the circumstances.

In my view, the Secretary of State should succeed on Issue (1) and on Issue (2) in relation to all the appeals, and on the Hardial Singh issue in relation to Mr Hemmati and Mr Khalili. I would dismiss the appeals of Mr Hemmati, Mr Khalili, Mr Abdulkadir and Mr Mohammed and I would allow the appeal of the Secretary of State in the case of SS. However, since the Master of the Rolls and Peter Jackson LJ take a different view from me on Issue (1) and Issue (2), the outcome of the appeals will be that set out in their conclusion below.

These proceedings and the appeals have proceeded on the footing that the only ground for detaining the appellants was that specified in Article 28(2) – a significant risk of absconding.

Sales LJ has identified the two matters in issue:

We respectfully do not agree with his conclusions on either of those issues.

A starting point for the CJEU in Al Chodor was that Dublin III was intended to improve the protection afforded to applicants under the Dublin system: paras 33-35, and Case C-63/15 Ghezelbash v Staatssecretaris van Veiligheid en Justitie [2016] 1 WLR 3969 at para. 52. Another was that Article 28 and Article 2(n) provide for a limitation on the exercise of the fundamental right to liberty enshrined in Article 6 of the EU Charter of Fundamental Rights: paras. 36 and 37.

The factual context for the decision of the CJEU in Al Chodor was a “consistent administrative practice” of the Czech Foreigners Police section, which was confirmed by Czech case law, and which was “characterised by the absence of arbitrary elements, and by predictability and an individual assessment in each case”: para. 22.

The legal touchstone applied by the CJEU for assessing compliance with Article 28(2) and Article 2(n) was whether the provisions relied upon for detention had the requisite legal basis, and the safeguards of clarity, predictability, accessibility and protection against arbitrariness within a framework of certain predetermined limits: paras. 40, 41 and 42.

The decision of the CJEU was that settled case law confirming the consistent administrative practice on the part of the Czech Foreigners Police Section was insufficient to satisfy that legal touchstone: para. 45

We consider that, in the light of the reasoning and the decision of the CJEU in Al Chodor, it is clear that neither the Hardial Singh principles nor the Secretary of State’s published policy in Chapter 55 of the EIG satisfied the requirements of Article 28 and Article 2(n) in relation to the period of detention of the appellants.

By the end of the oral hearing of the appeal, we did not understand it to be in dispute that the Hardial Singh principles are insufficient to satisfy the requirements of Article 2(n). That is in any event our view.

As summarized in R (I) v SSHD [2003] INLR 196 at [46] (and affirmed in R (Lumba) v SSHD [2011] UKSC 121, AC 245, at [171]), the Hardial Singh principles are as follows:

“(1)

The Secretary of State must intend to deport the person and can only use the power to detain for that purpose.

(2)

The deportee may only be detained for a period that is reasonable in all the circumstances.

(3)

If, before the expiry of the reasonable period, it becomes apparent that the Secretary of State will not be able to effect deportation within that reasonable period, he should not seek to exercise the power of detention.

(4)

The Secretary of State should act with reasonable diligence and expedition to effect removal.”

In substance, the Hardial Singh principles impose a reasonableness test. As Lord Dyson JSC said in Lumba at [30], the principles require that the power to detain be exercised reasonably and for the prescribed purpose of facilitating deportation. As Lord Justice Carnwath said in R (Dritain Krasniqi) v SSHD [2011] EWCA Civ 1549 at [12], they are no more than applications of two elementary propositions of English law: first, that compulsory detention must be properly justified, and, secondly, that statutory powers must be used for the purposes for which they are given. They are certainly not criteria defined by law for the purposes of Article 2(n). They do not specify criteria for deprivation of liberty which have the clarity, predictability, accessibility and protection against arbitrariness “within a framework of certain predetermined limits” required by Al Chodor.

Nor were those requirements satisfied by the EIG at the relevant time. The EIG set out the Home Office’s administrative policy. They contained no reference at all to Dublin III. They contained no direction that in an Article 28 case the sole ground for detention was the existence of a significant risk of absconding or that such a case was any different from any other asylum case. They gave no direction as to the need for proportionality in accordance with the requirement in Article 28(2).

The EIG did not give any direction or provide any clarification as to what “significant risk of absconding” might be. They did give general advice about the exercise of the power to detain. Paragraph 55.1.1. said, for example, that detention must only be based on one of the statutory powers and accord with the limitations implied by domestic and Strasbourg case law and with stated policy, and that detention must be for the shortest possible period of time. Paragraph 55.3 said that there must be strong grounds for believing that a person would not comply with conditions of temporary admission or temporary release for detention to be justified. None of those general statements about detention in immigration cases generally could possibly have amounted to objective criteria having the qualities of clarity, predictability and accessibility within a framework of certain predetermined limits as required by Al Chodor.

Paragraph 55.3.1 of the EIG was headed: “Factors influencing a decision to detain”. It stated that all relevant factors must be taken into account when considering the need for initial or continued detention. It said that those factors included eleven specified matters. One of them was: “Is there any evidence of previous absconding?” That was the only factor which referred to absconding. Some of the other specified factors could have been relevant to a risk of absconding but on the face of it others, such as the age of the applicant, or whether they had a history of torture or whether they had a history of physical or mental ill health, would not. Some of the factors might have been relevant to absconding but might equally have been relevant to a quite different policy consideration irrelevant to Article 28(2), for example the stated factor: “Is there a risk of offending or harm to the public (this requires consideration of the likelihood of harm and the seriousness of the harm if the person does offend)?”. It is significant that this factor was omitted entirely from the 2017 Regulations.

Paragraph 55.3.2.5 of the EIG was headed, and specifically concerned, “Risk of absconding”. This paragraph described some matters relevant to an assessment of risk of absconding but it was not expressed to contain, and did not purport to contain, all the criteria relevant to that issue. Nor did it say which of the factors specified in paragraph 55.3.1 were relevant and which were not relevant to that issue. None of that is surprising since Chapter 55 was intended to cover all immigration detainees and not merely those subject to Dublin III.

A list of criteria, some of which are relevant to absconding and others of which are not or may not be relevant, do not satisfy the Al Chodor requirements for the criteria for assessment of the risk of absconding to be set out in a legally binding instrument and to be clear, predictable and accessible. They do not enable the applicant to know which criteria will be applied specifically to meet the requirements of Article 28 and 2(n). Critically, they are not a “framework of certain predetermined limits” as required by Al Chodor for deprivation of the fundamental right to liberty under Article 6 of the Charter.

It was recorded at paragraph [53] in the judgment of Lambert J in R (Omar) v SSHD [2018] EWHC 689, which concerned the lawfulness of the 2017 Regulations, that it was common ground between counsel for the claimants and counsel for the Secretary of State that the criteria in the 2017 Regulations had to be “both mandatory and exhaustive”, otherwise the necessary protections of certainty, predictability and accessibility would not be conferred nor the protection against arbitrariness. As Lambert J said in Omar at [44]:

“The purpose of the criteria is to limit the basis upon which the determination of risk may be made; their existence provides sufficient guarantee in terms of legal certainty and ensures that the discretion enjoyed by the individual authorities responsible for applying the criteria for assessing the abscond risk is exercised within a framework of certain pre-determined markers.”

We consider that the agreement of the Secretary of State in that case and the statement of Lambert J were correct.

It was suggested by Mr Swift, on behalf of the Secretary of State on the present appeals, that it must be assumed that the case workers were aware of the law and so would have applied the relevant criteria with knowledge of the requirements of Article 28(2) and Article 2(n) and Al Chodor. It is impossible to see how such an assumption could lead to either the EIG or the criteria being clear, predictable and accessible if the instrument itself did not have those qualities.

There is no evidence before the court that case workers at the relevant time were told anything about Dublin III. Indeed, as Ms Lieven pointed out in her oral submissions, the suggestion that the case workers would have been told the implications of Article 28 and Article 2(n), as interpreted in Al Chodor, is entirely inconsistent with the argument of the Secretary of State at first instance before Garnham J and Mr John Howell QC that Article 28 added nothing and could not be relied upon directly in the UK.

In any event, the evidence of what actually happened in the case of the appellants undermines any assumption that the case workers knew and understood the implications of Article 28 and Article 2(n). In the case of each appellant, Form IS.91R (Notice to Detainee of Reasons for Detention and Bail Rights) was completed by an immigration officer. The form contained a box side-marked “You are likely to abscond if given temporary admission or release”. That box was not ticked in the case of any of the appellants. By way of further example, in the case of SS, another box side-marked “Your removal from the United Kingdom is imminent”, which corresponded to one of the factors in paragraph 55.3.1 of the EIG (“What is the likelihood of the person being removed and, if so, after what timescale?”), was, however, ticked. Monthly progress reports were subsequently sent to SS by the caseworker. Two of them dated 13 October 2015 and 8 December 2015 respectively were shown to us. They did not contain any express statement that SS continued to be detained because of a risk of absconding.

For those reasons the EIG at the relevant time did not satisfy the requirements of Article 28(2) and Article 2(n) and so the detention of the appellants was unlawful.

It is not necessary, therefore, to consider two further matters that were in issue on the appeal on the first question, namely whether the EIG, even though not legislation and even though capable of being changed at any time by the Secretary of State without being subject to Parliamentary scrutiny or consultation, (1) provided the necessary predictability required by Al Chodor, and (2) constituted the binding and foreseeable “act” specified in paragraph [42] of the decision of the CJEU in Al Chodor and the defining “law” required by Article 2(n).

Mr Swift argued, and Sales LJ would hold, that the common law doctrine of legitimate expectation and the remedy of judicial review are sufficient to satisfy those requirements of Article 28 and Article 2(n). As we have said, it is not necessary to decide this issue. Sales LJ has powerfully articulated one side of the argument but there are points to the contrary. In the first place, it may be said that retrospective endorsement of Executive action based on the past conduct of the Executive, which requires an intensive assessment of the facts, is no more compatible with the requirements of Article 2(n) than the “settled case law” confirming the “consistent administrative practice of the Foreigners Police section … characterised by the absence of arbitrary elements, and by predictability” which was considered insufficient by the CJEU in Al Chodor.

Mr Swift invoked R (Munjaz) v Mersey Care NHS Trust [2005] UKHL 58, [2006] 2 AC 148 in support of his argument on this point. Sales LJ considers he was correct to do so. It is not obvious, however, that Munjaz assists. It certainly is not directly applicable to Article 28 or Article 2(n). The case was about whether a policy on seclusion drawn up by Ashworth Hospital, which provided high security hospital accommodation and services for persons liable to be detained under the Mental Health Act 1983, was lawful under the domestic law of England and Wales and compatible with the ECHR. Relevant provisions of the ECHR considered by the House of Lords were Articles 3, 5 and 8. One of the issues was whether seclusion under the policy was “in accordance with the law” under Article 8 (2) of the ECHR because it was not prescribed by a binding general law. It was held that there was no violation of any of the ECHR Articles under consideration. It was held that the rules were accessible, foreseeable and predictable. Lord Hope said (at [91]) that the phrases “in accordance with the law” in Article 8(2) of the ECHR has the same meaning as the expression “prescribed by law” in Articles 9, 10 and 11. He said that “law” in that context is not limited to statutory enactment or to measures that have their base in a statute. It includes the common law. But the measure must be formulated with sufficient precision and be sufficiently accessible to satisfy the criterion for foreseeability. The present cases and appeals are not under the ECHR. They are brought under Dublin III, specifically with regard to the requirements of Article 28 and Article 2(n) as interpreted and applied in Al Chodor. The same point applies to the judgment of the European Court of Human Rights in The Sunday Times v The UK Application no. 6538/74 (judgment 27 April 1979) – which was also mentioned by Mr Swift and is cited by Sales LJ - in its consideration of the expression “prescribed by law” in various provisions of the ECHR.

Secondly, the appellants argued that it follows from the decision of the CJEU in Al Chodor that the requisite binding and foreseeable “act” specified in paragraph [42] of the decision will only comply with Article 2(n) if is made in legislation, that is by an entity different from the Executive or, at any event, if it is made by a different person or entity than the Secretary of State, who is also responsible for the implementation of the detention policy.

This point arises as a result of the express endorsement by the CJEU of paragraphs 81 and 82 of the opinion of the Advocate General and the statement of the CJEU in paragraph 44 of its decision that “criteria established by a binding provision are best placed for the external direction of the discretion of those authorities [in the circumstances of each specific case] for the purposes of protecting applicants against arbitrary deprivations of liberty”. Paragraphs 81 and 82 of the opinion of the Advocate General were as follows:

“81.

The adoption of legislation, in addition to providing advantages in terms of legal certainty, offers additional assurances in terms of external control of the discretion of the administrative and judicial authorities responsible for assessing the risk of absconding and, where appropriate, ordering the detention of applicants.”

82.

Given the particularly serious nature of the interference with the fundamental right to liberty represented by a detention order and the EU legislature’s desire to limit that interference to exceptional circumstances, the discretion of those authorities should be circumscribed in such a way as will best guard applicants against arbitrary deprivations of liberty. From that perspective, I believe it is important that the content of the criteria in the abstract and their actual application in a specific case be determined by institutionally separate authorities.”

On the one hand, it is a notable that, despite its endorsement of those paragraphs, the CJEU did not state that the criteria had to be defined by legislation but, instead, required them (in paragraph 42) to be defined by “an act which is binding and foreseeable in its application”. Furthermore, the language of the CJEU in paragraph 44 of its decision, which uses the expression “best placed”, is less emphatic than the clear statement of the Advocate General that “it is important that the content of the criteria in the abstract and their actual application in a specific case be determined by institutionally separate authorities”.

On the other hand, the extent to which the CJEU was adopting or not adopting paragraphs 81 and 82 of the opinion of the Advocate General is not entirely clear. That is reflected in the apparently inconsistent position taken on the point by the Secretary of State. In Omar, which concerned the lawfulness of the 2017 Regulations, counsel for the Secretary of State is recorded at [7] of the judgment of Lambert J as accepting that:

“Following Al Chodor the requirement is for objective criteria which form the basis of a decision concerning the absconding risk to be implemented in domestic legislation”.

While no such concession is made by Mr Swift in the present case, and Mr Swift informed us that the concession by counsel for the Secretary of State in Omar was without prejudice to the present cases and appeals, the importance of the point, the lack of clarity on it in Al Chodor and the apparent inconsistent position taken by the Secretary of State indicate that we should not decide the point unless we have to.

Sales LJ has concluded that it is the “sufficiently serious breach” test set out in Factortame which governs whether damages are payable in relation to the detention of each of the appellants and not the domestic law of false imprisonment. We respectfully disagree. We consider that Mr Southey’s argument, which Sales LJ has dismissed as “too simplistic”, is correct.

There is no doubt that all the necessary ingredients for the common law cause of action for false imprisonment are satisfied in the case of each of the appellants. As Lord Dyson said in Lumba at [65]:

“All that a claimant has to prove in order to establish false imprisonment is that he was directly and intentionally imprisoned by the defendant, whereupon the burden shifts to the defendant to show that there was lawful justification for doing so. As Lord Bridge of Harwich said in R v Deputy Governor of Parkhurst Prison, ex p. Hague [1992] 1 AC 58, 162C-D: “The tort of false imprisonment has two ingredients: the fact of imprisonment and the absence of lawful authority to justify it.””

The proper analysis here is that paragraph 16(2) of Schedule 2 to the Immigration Act 1971 confers a discretion to detain persons liable to be removed from the UK pending a decision whether or not to give directions for removal and pending removal in pursuance of such directions, and it was that discretion which was purportedly exercised in respect of all the appellants. The detention of the appellants was unlawful because it was purportedly pursuant to the policy in the EIG which was itself unlawful insofar as it failed to give effect to Article 28(2) and Article 2(n): see AlChodor at [46]. Although the CJEU (at paragraphs 36 and 41) described the effect of Article 28(2) as a limitation on the fundamental right to liberty, its direct effect in the UK operated as a limitation on the exercise of the statutory discretion to detain pursuant to paragraph 16(2) of Schedule 2 to the 1971 Act.

In R (Kambadzi) v SSHD [2011] UKSC 32, [2011] 1 WLR 1299, at [40] Lord Hope said: “Where there is an executive discretion to detain someone without limit of time, the right to liberty demands that the cause of action [for false imprisonment] should be available if the discretion has not been lawfully exercised.” The person responsible for the unlawful detention is liable even if he or she acted in good faith and without any negligence: [40] (Lord Hope), [63] (Baroness Hale).

The Factortame principle has no relevance because the individual right of each human being to liberty exists save insofar as it is legitimately cut down by law. The right to liberty does not exist because of the EU and its Charter of Fundamental Rights any more than it is exists because of the Council of Europe and the Convention. As Article 1 of the 1948 Universal Declaration of Human Rights states: “All human beings are born free”.

Our common law has provided the remedy for an infringement of that inherent fundamental human right, whether it be habeas corpus (now governed by statute) or damages. Unlike Factortame these appeals do not concern infringement of rights which are to be found only in EU law. That is also why it differs from Nabadda v City of Westminster [2001] 3 CMLR 39, on which Mr Swift relied, in which the claims of the student applicants, who were Swedish nationals, to damages in respect of the calculation of their mandatory education fees awards from UK local authorities were based on EU freedom of movement provisions and the Treaty of Rome.

The assessment of the amount of damages will, of course, be a matter to be determined for by court to which the cases are remitted.

As these appeals have proceeded on the footing that the only ground for detaining the appellants was that specified in Article 28(2) and, on that basis, the detention of the appellants was unlawful for the reasons we have given, it is not necessary to address the application of the Hardial Singh principles in respect of the detention of Mr Hemmati and Mr Khalili.

In paragraphs [101] to [111] of his judgment Sales LJ considers various matters which it is not necessary to decide on these appeals, and many of which were not the subject of submissions on the hearing of the appeals. We prefer to express no view on them.

For the reasons stated above, we would allow the appeals of Mr Hemmati, Mr Khalili, Mr Abdulkadir and Mr Mohammed and dismiss the appeal of the Secretary of State.