R v Secretary of State for the Home Department

This appeal by the Secretary of State for the Home Department concerns five individuals, the respondents, who arrived in the United Kingdom illegally and claimed asylum. Inquiries revealed that they had travelled to the United Kingdom via at least one other member state of the European Union in which they had already claimed asylum and so the Secretary of State requested those states to take responsibility for examining the asylum claims pursuant to Parliament and Council Regulation (EU) No 604/2013 of 26 June 2013 (“the Dublin III Regulation”, “Dublin III” or “the Regulation”). Ultimately each such state agreed to take the relevant respondent back for that purpose.

The respondents were all detained for a period of time pending their removal pursuant to paragraph 16(2) of Schedule 2 to the Immigration Act 1971. The Secretary of State had published her policy in relation to detention pending removal in Chapter 55 of her Enforcement Instructions and Guidance (23 October 2015) (“the EIG”). This appeal now gives rise to important questions concerning the requirements imposed on member states by the Dublin III Regulation, whether the policy in Chapter 55 of the EIG meets those requirements and, so far as it fails to do so, the consequences of that failure. The particular questions which must be decided are:

whether the detention of each respondent was lawful given that article 28 of the Dublin III Regulation permits detention where there is a “significant risk of absconding”, “risk of absconding” being defined in article 2(n) as the existence of reasons in an individual case, based on objective criteria defined by law, to believe that the person might abscond; and, if the detention was not lawful,

whether damages are payable either under domestic law for false imprisonment or pursuant to what is known as the Factortame principle established in Brasserie du Pecheur SA v Federal Republic of Germany; R v Secretary of State for Transport; Ex p Factortame Ltd No 4 (Joined Cases C-46/93 and C-48/93) [1996] QB 404.

Ms Hoda Hemmati, the first respondent, is a national of Iran and arrived in the United Kingdom illegally by lorry. On 11 February 2015 she presented herself to the authorities and claimed asylum. A check revealed that she had already claimed asylum in Bulgaria. The United Kingdom proceeded formally to request Bulgaria under the Dublin III procedure to take responsibility for the asylum claim and on 17 April 2015 Bulgaria agreed to do so. On 8 June 2015 she was detained in order to effect her removal to Bulgaria and, according to the evidence of the Secretary of State, on the basis she posed a risk of absconding. Removal directions were set for 7 July 2015. These were cancelled when she gave notice that she had issued judicial review proceedings to challenge the decision to remove her. She contended that removal would give rise to a real risk of a violation of her rights under article 3 of the European Convention for the Protection of Human Rights and Fundamental Freedoms (“the ECHR”). She was released on 17 July 2015. She was therefore detained from 8 June 2015 to 17 July 2015.

Mr Fawad Khalili, the second respondent, is a national of Afghanistan. He arrived in the United Kingdom illegally by lorry. On 20 November 2014 he presented himself to the authorities and claimed asylum. He was initially released on temporary admission. On 6 January 2015 he attended a screening interview and made a formal in-country claim for asylum. He was detained on the basis that his removal was imminent, that he had behaved deceptively and, according to the evidence of the Secretary of State, that he posed a risk of absconding. A check revealed that he too had already made an asylum claim in Bulgaria. The United Kingdom proceeded formally to request Bulgaria under the Dublin III procedure to take responsibility for the asylum claim and on 12 February 2015 Bulgaria agreed to do so. Removal directions were set for 23 February 2015.

In the meantime, on 5 February 2015, the second respondent made submissions to the Secretary of State that his removal would breach his rights under article 3 of the ECHR and on 20 February 2015 he issued a claim for judicial review to prevent his scheduled removal. The removal directions were cancelled and on 9 March 2015 he was granted bail by the First-tier Tribunal (“the FTT”). He claims that his detention became unlawful on 5 February 2015, the date he complained that his removal to Bulgaria would be incompatible with article 3 of the ECHR. The relevant period of his detention was therefore 5 February 2015 to 9 March 2015.

Mr Jamal Abdulkadir, the third respondent, is a national of Iraq. On 18 August 2015 he arrived in the United Kingdom illegally by lorry. Upon arrival in Kent, he ran from the lorry but was apprehended and detained. The Secretary of State maintains that the evidence he produced of his identity was inadequate and that there was a risk he would abscond. A check revealed that he had made an asylum claim in Austria in July 2015. At this point he claimed asylum in the United Kingdom. The United Kingdom proceeded formally to request Austria under the Dublin III procedure to take responsibility for the asylum claim and on 15 September 2015 Austria agreed to do so. On 28 September 2015 removal directions were set but later cancelled at Austria’s request. Further removal directions were set for 23 October 2015. On 22 October 2015 the third respondent issued judicial review proceedings challenging the decision to remove him on the basis that in Austria he would be exposed to a real risk of violation of his rights under article 3 of the ECHR. He made an application for bail which the FTT refused on 13 November on the basis that there was a risk he would abscond and that it was likely he would be removed in a short time. On 27 November 2015 he was given permission to apply for judicial review and on 8 December 2015, upon review of his detention, he was released. He was therefore detained from 18 August 2015 to 8 December 2015.

Mr Jwytar Mohammed, the fourth respondent, is also a national of Iraq. On 8 September 2015 he arrived in the United Kingdom illegally by lorry. He ran off when the lorry doors were opened but was apprehended later that day. He claimed asylum and, according to the evidence of the Secretary of State, was detained on the basis he posed a risk of absconding. A check revealed that he had previously claimed asylum in Austria. The United Kingdom proceeded formally to request Austria under the Dublin III procedure to take responsibility for the asylum claim and on 15 September 2015 Austria agreed to do so. Removal directions were set for 12 October 2015 but on 2 October 2015 he began proceedings for judicial review claiming, among other things, that in Austria he would be exposed to a real risk of violation of his rights under article 3 of the ECHR. He was released from detention on 4 November 2015. He was therefore detained from 8 September 2015 to 4 November 2015.

SS, the fifth respondent, is a national of Afghanistan. On 15 September 2015 he arrived in the United Kingdom illegally and by hiding in the back of a train. Upon arrival he claimed asylum and pretended to be a child. A check revealed he had already claimed asylum in Bulgaria, Hungary and Germany. He was detained on that same day on the basis that it was reasonably likely that he would be accepted by another member state under the Dublin III procedure. A month later a notice of detention review stated that it had been decided he should remain in detention because there was reason to believe he would not comply with any conditions of release. The United Kingdom proceeded formally to request Bulgaria, Hungary and Germany under the Dublin III procedure to take responsibility for the asylum claim and on 27 October 2015 Germany agreed to do so. Removal directions were set for 30 November 2015 but were cancelled when he began proceedings for judicial review. He was released from detention on 10 December 2015. He was therefore detained from 15 September 2015 to 10 December 2015.

The judicial review claims brought by the first and second respondents were listed for hearing together with claims brought by three other individuals. The first and second respondents challenged both the lawfulness of their removal and the lawfulness of their detention. The claims were heard by Garnham J who dealt first with their challenges to removal. He gave judgment on 18 April 2016 dismissing all of the claims ([2016] EWHC 857 (Admin)) and an appeal to the Court of Appeal was subsequently dismissed ([2017] EWCA Civ 1871). There has been no further appeal against that decision. Garnham J gave a further judgment on 15 June 2016 dealing with the claims for unlawful detention ([2016] EWHC 1394 (Admin); [2016] 1 WLR 4243). He allowed two of the claims but not those of the first and second respondents.

The judicial review claims brought by the third and fourth respondents were listed for hearing together. Again, they both challenged the lawfulness of their removal and their detention. These claims were heard by Irwin J who dismissed them all for reasons given in his judgment of 28 June 2016 ([2016] EWHC 1504) (Admin)).

The judicial review claim brought by the fifth respondent came on for hearing before Mr John Howell QC, sitting as a deputy High Court judge. The original grounds of claim focused on the assertion by the fifth respondent that he was a child and therefore could not be removed to Germany under the Dublin III scheme. But at the hearing he was permitted to amend his claim to introduce a further claim that he was unlawfully detained because his detention was contrary to articles 28(2) and 2(n) of the Dublin III Regulation. The deputy judge gave judgment on 26 May 2017 ([2017] EWHC 1295 (Admin); [2017] 1 WLR 3641). He found that that the fifth respondent was not a child when he was detained. However, he also found that the fifth respondent had been detained to secure his transfer to the responsible member state under the Dublin III scheme; that it had to be established that he posed a significant risk of absconding; and that his detention was unlawful because, even if he posed a significant risk of absconding, his detention was in conflict with articles 28(2) and 2(n).

The first to fourth respondents in the first four claims and the Secretary of State in the fifth appealed to the Court of Appeal. The principal issues before the court concerned the meaning and effect of articles 28 and 2(n) of the Dublin III Regulation and, in particular, whether the application of the principles explained in R v Governor of Durham Prison, Ex p Hardial Singh [1984] 1 WLR 704 or the Secretary of State’s policy set out in Chapter 55 of the EIG satisfied the requirements of those articles; and, if not, whether damages were payable in respect of the respondents’ detention either under domestic law for false imprisonment or under European Union law pursuant to the Factortame principle. It is important to note that the Court of Appeal was asked to decide these issues on the agreed assumption that the only ground for detaining the respondents was that the Secretary of State wished to remove them using the Dublin III procedure. The position remains the same on this further appeal.

The first and second respondents also raised as a separate issue whether their detention was unlawful because of a failure by the Secretary of State to comply with the Hardial Singh principles.

The Court of Appeal, by a majority (Sir Terence Etherton MR and Peter Jackson LJ), allowed the appeals of the first to fourth respondents and dismissed the appeal of the Secretary of State in the case of the fifth respondent. Critical to the reasoning of the majority was the decision of the second chamber of the Court of Justice of the European Union (the “CJEU”) in Policie ČR, Krajské ředitelství policie Ústeckého kraje, odbor cizinecké policie v Al Chodor (Case C-528/15) [2017] 4 WLR 125. This decision post-dated the decisions of Garnham J and Irwin J but predated that of Mr John Howell QC. The majority held that the touchstone applied by the CJEU in Al Chodor for assessing compliance with articles 28(2) and 2(n) of the Dublin III Regulation 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. The majority also held that it was clear that neither the Hardial Singh principles nor the Secretary of State’s published policy in Chapter 55 of the EIG satisfied these requirements. It followed that the detention of all of the respondents was in breach of article 28(2).

The majority of the Court of Appeal went on to hold that each of the respondents had established all of the necessary ingredients of the common law cause of action for wrongful imprisonment. They had all been detained and that detention was unlawful because it was effected pursuant to the policy in Chapter 55 of the EIG, and that was itself unlawful in so far as it failed to give effect to articles 28(2) and 2(n) of the Regulation. The respondents were therefore entitled to damages for false imprisonment. The Factortame principle had no relevance because the individual right of each person to liberty existed save in so far as it is legitimately cut down by law. The appeals did not concern infringement of rights which were to be found only in European Union law. Further, it was not necessary to consider the additional and discrete claims by the first and second respondents for false imprisonment based upon the alleged breach by the Secretary of State of the Hardial Singh principles.

Sales LJ, dissenting, held that a policy statement such as that contained in Chapter 55 of the EIG was in principle capable of satisfying the requirements of articles 28(2) and 2(n) of the Dublin III Regulation, and that here it did satisfy those requirements. However, conscious that he was in the minority on this issue, he went on to consider whether, on the footing that he was wrong, the respondents were entitled to damages. He concluded that they were not. In his view, the claim turned on the alleged failure by the United Kingdom to adopt a particular form of law when implementing articles 28(2) and 2(n). In these circumstances the proper approach in considering whether the Secretary of State was liable for damages was to ask whether the relevant criteria for an award of damages in respect of a breach of European law had been satisfied and, in particular, whether the breach was sufficiently serious within the meaning of the decision of the CJEU in Factortame , that is to say whether the member state had manifestly and gravely disregarded the limits of its discretion. Here, any breach of articles 28(2) and 2(n) did not satisfy that “sufficiently serious” test.

Sales LJ also addressed the separate claims by the first and second respondents for false imprisonment based upon a breach of the Hardial Singh principles. In his view there was nothing in them, and in this regard he agreed with the decision of Garnham J: the first and second respondents were detained for proper reasons; they were assessed as posing a risk of absconding and that assessment was rational and justified; and throughout the period of their detention, there remained a real prospect that they would be removed eventually.

This further appeal therefore raises the following important issues of principle concerning the limits of the permission conferred by the Dublin III Regulation upon member states to detain an applicant for international protection in order to secure the transfer of that applicant to another member state in accordance with the transfer procedures laid down in the Regulation:

Does the policy published by the Secretary of State in Chapter 55 of the EIG satisfy the requirements imposed by articles 28 and 2(n) of the Regulation for a measure setting out “objective criteria defined by law” for believing that an applicant for international protection who is subject to a transfer procedure may abscond?

If not, are damages payable to an applicant whose detention pursuant to article 28(2) was authorised by the Secretary of State pending such transfer:

under domestic law for the tort of false imprisonment, or

pursuant to European Union law under the Factortame principle?

The European Union has for some time sought to establish an area of freedom, security and justice which is open to those who, forced by circumstances, legitimately seek its protection. To this end the Union has harmonised the procedures and substantive rules of refugee law and, as part of that harmonisation, established a body of law within what is known as the Common European Asylum System (the “CEAS”). A well-functioning Dublin system is seen as essential to the CEAS in ensuring the rapid identification of the member state responsible for examining an application for international protection, and in this way guaranteeing effective access to the procedures for determining refugee status. The Dublin III Regulation replaced Council Regulation (EC) No 343/2003 of 18 February 2003 establishing the criteria and mechanisms for determining the member state responsible for examining an asylum application lodged in one of the member states by a third country national (“the Dublin II Regulation”). Dublin III lays down, in Chapter III, a hierarchy of criteria for identifying the member state responsible for deciding the claim. If the member state where an asylum claim has been lodged considers that another member state is responsible then it may ask the other member state to take charge of the applicant. If the other member state agrees to this request, the first member state will transfer the applicant there in accordance with the procedure laid down in the Regulation.

The Dublin III Regulation permits, subject to strict safeguards, the detention of an applicant for international protection in order to ensure that the Dublin III procedure is implemented effectively. In this regard, recital 20 provides, so far as relevant:

The terms of this recital are reflected in article 28 of the Regulation which authorises member states to detain applicants subject to various conditions. It provides:

Article 28(2) therefore permits, subject to the other provisions of the article, the detention of applicants in order to secure their transfer in accordance with the Regulation, but only where there is a significant risk of absconding; that risk has been identified on the basis of an individual assessment; and the detention is proportional and other less coercive measures cannot be applied effectively. Article 2(n) defines “risk of absconding” as:

I must also refer to another of the group of instruments forming the CEAS body of law: Parliament and Council Directive 2013/33/EU of 26 June 2013 (“the recast Reception Directive”), referred to in recital 20 of the Dublin III Regulation. The United Kingdom has not opted in to this Directive and remains governed by its predecessor, Council Directive 2003/9/EC of 27 January 2003. Nevertheless, its terms do shed some light on the meaning of the relevant terms of the Dublin III Regulation.

Recital 15 of the recast Reception Directive provides:

Article 8 of the recast Reception Directive provides, so far as relevant:

The meaning of these provisions of the Dublin III Regulation was considered by the CJEU in Al Chodor on a reference from the Nejvyšší správní soud, the Supreme Administrative Court of the Czech Republic. This decision is of great importance to the issues arising on this appeal and so I must deal with it in some detail.

The case concerned a family of Iraqi nationals, the Al Chodors, who were stopped by police in the Czech Republic and interviewed. They claimed to be of Kurdish origin and it emerged that they had travelled to the Czech Republic via Turkey, Greece and then Hungary, where they had claimed asylum. The Foreigners Police Section of the Czech police force decided to place the family in detention pending their transfer to Hungary under the Dublin system. They took the view, for perfectly sensible reasons, that there was a serious risk that, unless detained, the Al Chodors would abscond before their transfer. The relevant Czech legislation conferred on the police force the power to detain a foreign national who had entered the Czech Republic illegally for the period of time necessary to secure the transfer of that person in accordance with, among other measures, the Dublin III Regulation. Upon a challenge by the Al Chodors, the Czech Regional Court annulled the decision to detain on the basis that the objective criteria for assessing the risk of absconding were not defined by Czech legislation as required by article 2(n) of the Regulation. The police force then brought an appeal on a point of law before the Supreme Administrative Court, which made the reference to the CJEU.

The referring court was unsure whether the relevant Czech legislation, read together with articles 28(2) and 2(n) of the Dublin III Regulation, provided a sufficient legal basis for detention given that it did not lay down any objective criteria for assessing the risk of absconding. The court was also unsure whether the recognition of such criteria in case law which confirmed a consistent administrative practice of the police could meet the requirement of criteria “defined by law” within the meaning of article 2(n).

It was pointed out by the referring court that the various language versions of article 2(n) diverged and that while the French and German versions required a definition, laid down in legislation, of the objective criteria for the purposes of absconding, other versions required a definition of those criteria by law in the general sense. As a result, the meaning of the term “defined by law” did not follow clearly from the wording of the provision. The referring court therefore asked, in substance, whether articles 2(n) and 28(2) of the Dublin III Regulation require member states to establish, in a national law, objective criteria underlying the reasons for believing that an applicant for international protection who is subject to a transfer procedure may abscond, and whether the absence of those criteria in a national law leads to the inapplicability of article 28(2).

The Czech and United Kingdom Governments argued that, according to the case law of the European Court of Human Rights (“the ECtHR”), the concept of law as referred to in the ECHR was not limited to legislation but also included other sources of law provided they possessed the qualities of precision, foreseeability and accountability. They contended that the Czech case law and relevant administrative practice possessed those qualities in this case. The Greek Government and the Commission disagreed.

Advocate General Saugmandsgaard Øe delivered his Opinion on 10 November 2016 EU:C:2016:865; [2017] 3 CMLR 24. He considered that the concept of “law” as referred to in article 2(n) of Dublin III, read in context and in light of its purpose, had a meaning which was different from that of the concept of “law” in the ECHR; that the provisions of the ECHR established a minimum level of protection and did not exclude the possibility that European Union law might provide more extensive protection; that the European Union legislature had indeed chosen to provide more extensive protection than that arising from article 5(1) of the ECHR; and that the criteria had to be laid down in legislation (Opinion, paras 42-45). He then proceeded to explain his reasons for these views.

The Advocate General began with his reasons for believing that the Dublin III Regulation and the recast Reception Directive were intended to extend the protection afforded to applicants (Opinion, paras 46-58). He pointed out that before these instruments there was only minimal regulation of applicants’ detention. By contrast and as stated in recital 15 of the recast Reception Directive, the European Union legislature intended that detention of such persons under the new regime should be limited to exceptional circumstances. He also explained that the freedom of member states to detain applicants was already subject to the restrictions imposed by article 5(1)(f) of the ECHR and that the compliance of a detention measure with this provision was not conditional on there being a risk of absconding or the absence of other less restrictive measures enabling the removal of the person concerned. In his view, in adopting the Dublin III Regulation and the recast Reception Directive, the legislature intended to provide more extensive protection than that arising from article 5(1)(f). His reasons were twofold: first, article 28(2) of Dublin III permitted detention only where there was a significant risk of absconding; and secondly, article 28(2) of Dublin III and article 8(2) of the recast Reception Directive provided that detention was a measure of last resort and might be taken only in the absence of less coercive alternative measures.

Two objectives of the requirement that the criteria for assessing the risk of absconding must be defined by law were identified by the Advocate General (Opinion, paras 59-70). The first was to ensure that the criteria offered sufficient guarantees of legal certainty, as that concept is understood in European Union law (Opinion, para 62). The second was to ensure that the discretion enjoyed “by the individual authorities responsible for applying the criteria” was exercised within “a framework of certain pre-determined markers” (Opinion, para 63). In the Advocate General’s view, the achievement of each of these objectives was dependent on the objective criteria for the assessment of the risk of an applicant absconding being defined in a legislative text. Legal certainty demanded that individuals should be able to ascertain the scope of their rights and obligations and foresee the consequences of their actions. This requirement had not been satisfied in the instant case because the Czech case law was fragmentary, and the relevant administrative practice could be altered at will and had not been publicised (Opinion, paras 72-80). As for the objective of circumscribing the discretion of the administrative and judicial authorities, the adoption of legislation, in addition to providing advantages in terms of legal certainty, offered 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 detention (Opinion, para 81). Interference with liberty should be limited to exceptional circumstances; the discretion of the authorities should be circumscribed in such a way as to guard applicants against arbitrary deprivations of liberty; and, from this perspective, it was important that the criteria and their application should be determined by institutionally separate authorities (Opinion, para 82). The twofold requirement inherent in article 2(n), for an individual assessment and for the assessment to be based upon pre-defined, objective criteria, required the administrative and judicial authorities to take the circumstances of each case into consideration; and further, it ensured that the discretion of the individual authority was channelled by means of “general, abstract criteria that have been determined in advance by a third authority” (Opinion, para 83).

The CJEU gave judgment on 15 March 2017 [2017] 4 WLR 125. It dealt first with a submission that the Dublin III Regulation was directly applicable in member states and did not require transposition into national law. Thus, it was argued, article 2(n) did not require national legislatures to implement the objective criteria in national law. The court rejected this submission, observing, at para 28, that article 2(n) required that objective criteria defining the existence of a risk of absconding be “defined by law”. Since these criteria had been established neither by Dublin III nor in any other European Union legal act, the elaboration of those criteria was a matter for national law.

The CJEU turned next to the question whether “law” in article 2(n) included settled case law which confirmed a consistent administrative practice. It explained, at paras 30 to 32, that a textual analysis did not provide an answer and that the provision had to be interpreted by reference to the purpose and the general scheme of the rules of which it formed a part. Here, the court continued, at paras 34 and 35, it was relevant that the Regulation was intended to make improvements to the effectiveness of the Dublin system but also to the protection afforded to applicants. Indeed, the high level of protection now afforded to applicants was apparent from the terms of articles 28 and 2(n), the requirement in article 2(n) that a finding of a risk of absconding must be based upon objective criteria defined by law and applied on a case by case basis, and the fact that its predecessor, the Dublin II Regulation, did not contain any provision relating to detention.

The CJEU also thought it relevant that, by authorising the detention of applicants in order to secure their effective transfer under the Dublin III regime, the Regulation was authorising a limitation on the fundamental right to liberty enshrined in article 6 of the Charter of Fundamental Rights of the European Union (“the Charter”), and account also had to be taken of article 5 of the ECHR as providing the minimum threshold of protection. The court reasoned that any law authorising the deprivation of liberty must therefore be sufficiently accessible, precise and foreseeable in its application to avoid all risk of arbitrariness (para 38); that there must be no element of bad faith or deception on the part of the authorities (para 39); and that the detention of applicants, constituting, as it does, a serious interference with their liberty, is subject to strict safeguards, namely the presence of a legal basis, clarity, predictability, accessibility and protection against arbitrariness (para 40).

There followed an analysis by the CJEU of the safeguard of “legal basis” and the type of provision needed to satisfy the other safeguards of clarity, predictability, accessibility and protection against arbitrariness. Here the court’s reasoning is of particular importance and so I set it out in full:

Accordingly, the CJEU ruled as follows:

It is striking that, although the CJEU plainly adopted aspects of the reasoning of the Advocate General, it did not in terms endorse his conclusion that the criteria for assessing the risk of absconding must be embodied in legislation. I must return to this decision in addressing the issues arising on this appeal but first I must say something about our domestic law.

Paragraph 16(2) of Schedule 2 to the Immigration Act 1971 (“the 1971 Act”) confers powers on immigration officers to detain an individual who is held within the immigration system pending a decision whether to give directions for his or her removal, and pending removal pursuant to any such directions. Similar powers are conferred on the Secretary of State by section 62(1) and section 62(2)(c) and (d) of the Nationality, Immigration and Asylum Act 2002. Paragraph 1(3) of Schedule 2 to the 1971 Act provides that, in exercising their functions under the 1971 Act, immigration officers must act in accordance with such instructions as may be given to them by the Secretary of State.

It is rightly accepted by the Secretary of State that there are limits to these powers to detain. First, they are subject to the Hardial Singh principles. These are well known and may be summarised as follows: (i) the Secretary of State must intend to deport the person and can only use the power to detain for that purpose; (ii) the person may only be detained for a period that is reasonable in all the circumstances; (iii) if, before the expiry of the relevant period, it becomes apparent that the Secretary of State will not be able to effect deportation within a reasonable period, he should not seek to exercise the power of detention; and (iv) the Secretary of State should act with reasonable diligence and expedition to effect the removal: see R (I) v Secretary of State for the Home Department [2002] EWCA Civ 888; [2003] INLR 196; affirmed as common ground in R (Lumba) v Secretary of State for the Home Department [2011] UKSC 12; [2012] 1 AC 245, paras 22 and 171. These principles reflect the basic public law duties to act consistently with the purpose of the legislation and reasonably in the Wednesbury sense: Associated Provincial Picture Houses Ltd v Wednesbury Corpn [1948] 1 KB 223. But, as Lord Dyson observed in Lumba at para 30, they are not exhaustive. Were it otherwise, there would be no room for a public law duty of adherence to a published policy, to which I will come in a moment.

We have also been referred by counsel for the Secretary of State to the decision of this court in R (Nouazli) v Secretary of State for the Home Department [2016] UKSC 16; [2016] 1 WLR 1565. There this court accepted (at para 75) that the primary responsibility to comply with the Hardial Singh principles lies with the Secretary of State but that the courts provide supervision of their application and that challenges are brought to secure release and not just for damages after the event.

The Secretary of State’s policy in relation to detention to effect removal was set out in Chapter 55 of the EIG. Paragraph 55.1.1 states:

to effect removal;

initially to establish a person's identity or basis of claim; or

where there is reason to believe that the person will fail to comply with any conditions attached to a grant of immigration bail.

Paragraph 55.1.3 provides:

Paragraph 55.1.4 addresses the implied limitations on the statutory powers to detain and provides:

Paragraph 55.1.4.1 then sets out what are, in substance, the Hardial Singh principles:

Paragraph 55.3 is also concerned with decisions to detain:

Of particular significance is paragraph 55.3.1 which provides:

What is the likelihood of the person being removed and, if so, after what timescale?

Is there any evidence of previous absconding?

Is there any evidence of a previous failure to comply with conditions of temporary release or bail?

Has the subject taken part in a determined attempt to breach the immigration laws? (For example, entry in breach of a deportation order, attempted or actual clandestine entry).

Is there a previous history of complying with the requirements of immigration control? (For example, by applying for a visa or further leave).

What are the person’s ties with the UK? Are there close relatives (including dependants) here? Does anyone rely on the person for support? If the dependant is a child or vulnerable adult, do they depend heavily on public welfare services for their daily care needs in lieu of support from the detainee? Does the person have a settled address/employment?

What are the individual’s expectations about the outcome of the case? Are there factors such as an outstanding appeal, an application for judicial review or representations which might afford more incentive to keep in touch than if such factors were not present? (See also 55.14).

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)?

Is the subject under 18?

Does the subject have a history of torture?

Does the subject have a history of physical or mental ill health?

A policy such as that embodied in Chapter 55 of the EIG is published so that an individual affected by it knows the criteria by which the executive has chosen to exercise the power conferred upon it by statute and so that the individual can make appropriate representations in relation to that exercise in relation to him. In Lumba , Lord Dyson explained its importance in these terms at para 34:

It is also submitted on behalf of the Secretary of State, and I accept, that the executive must follow this stated policy unless there are good grounds for not doing so. Further, and as Lord Dyson explained in Lumba , at para 66, a purported lawful authority to detain may be impugned either because the defendant has acted in excess of jurisdiction or because the jurisdiction has been wrongly exercised. Both species of error render an executive act ultra vires, unlawful and a nullity. There is in this context no difference between a detention which is unlawful because there was no statutory power to detain and a detention which is unlawful because the decision to detain, although authorised by the statute, was made in breach of a rule of public law.

The same point emerges from the decision of this court in R (Kambadzi) v Secretary of State for the Home Department [2011] UKSC 23; [2011] 1 WLR 1299. As Lord Hope of Craighead said at para 42, the published policy in that case narrowed the power of the executive to detain by requiring that any detention be reviewed regularly, and so it was an abuse of the power for any person to be detained if that detention was not reviewed at regular intervals. Lord Hope continued, at paras 51 and 52, that the policy was designed to give practical effect to the Hardial Singh principles and to meet the requirement that, to be lawful, the measures had to be transparent and not arbitrary; that the policy contained a set of instructions with which officials were expected to comply; that the policy and the principles went “hand in hand”; and that the discretion to continue detention had to be exercised in accordance with the principles but also in accordance with the policy.

To summarise the position under European Union law, a member state may not hold in detention a person who is subject to the Dublin III procedure unless there are reasons in that individual case, based on “ objective criteria defined by law ” and “ an individual assessment ”, to believe that person may abscond. In addition, detention must be proportional, is justified only where other less coercive alternative methods cannot be applied effectively and must be for as short a period as possible.

Further and as interpreted by the CJEU in Al Chodor and consistently with the ECHR and the Charter, articles 28(2) and 2(n) of the Dublin III Regulation demand that the detention of applicants, constituting, as such detention does, a serious interference with their right to liberty, is subject to compliance with the strict safeguards of “ legal basis, clarity, predictability, accessibility and protection against arbitrariness ” ( Al Chodor , para 40). The first of these safeguards requires national law to define “ objective criteria indicating the presence of a risk of absconding ” ( Al Chodor , para 41). The others demand that the assessment is carried out within a “ framework of certain predetermined limits ” and that the “ objective criteria ” are defined “ clearly by an act which is binding and foreseeable in its application ” ( Al Chodor , para 42) and are established in a “ binding provision of general application ” ( Al Chodor , para 45). The adoption of such a provision of general application provides the necessary safeguards in so far as its wording sets out the “ limits of the flexibility of [the] authorities in the assessment of the circumstances of each specific case ” in a manner which is “ binding and known in advance ” ( Al Chodor , para 44). The CJEU accordingly ruled that member states are required to establish, in a binding provision of general application, objective criteria underlying the reasons for believing that an applicant for international protection who is subject to a transfer procedure may abscond.

These requirements are of great importance. As we have seen, the Dublin III Regulation was intended to improve not only the effectiveness of the Dublin system but also to confer a high level of protection upon those subject to Dublin III transfers and to impose significant limitations on the powers of member states to detain them. The Regulation was also intended to provide greater guarantees in relation to detention than its predecessor, the Dublin II Regulation.

The Secretary of State accepts that, within the taxonomy of English law, Chapter 55 of the EIG cannot be described as legislation but contends that it nevertheless includes rules and that decision-makers have legal obligations, imposed by settled case law, to comply with them. It is submitted that the policy contained in Chapter 55 constitutes a clear statement by the executive of the circumstances in which the statutory criteria will be exercised; that they are objective and publicly accessible; and that their meaning will, if necessary, be determined by the court. It is also argued that Chapter 55 and domestic case law are integral parts of the law that governs and limits the power to detain and that together they define how the power to detain must be exercised and set out the objective criteria which decision-makers must apply when exercising that power. Consequently, the submission continues, the combination of Chapter 55 and domestic case law ensures that any decision to detain is exercised within a framework of certain predetermined limits and according to criteria established by binding provisions of general application which meet the requirements of clarity, predictability, accessibility and protection against arbitrariness explained by the CJEU in Al Chodor .

These submissions are all directed to the issues set out in para 18(i) above. The majority of the Court of Appeal addressed those issues in two stages. They considered first, whether the contents of Chapter 55 of the EIG met the requirement that the criteria for the assessment of the risk of absconding must be “clear, predictable and accessible”, and whether they constituted a “framework of certain predetermined limits”; and secondly whether Chapter 55, 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, provided the necessary predictability and amounted to a “binding provision of general application” as referred to by the CJEU in Al Chodor and constituted a defining “law” as required by article 2(n). I propose to adopt the same course.

As we have seen, Chapter 55 of the EIG contains the Secretary of State’s policy in relation to detention pending removal. However, it is not a policy which is specifically directed to the detention of persons subject to a Dublin procedure. Indeed, as the majority in the Court of Appeal observed, it contains no reference to Dublin III at all. In these circumstances it comes as no surprise that it makes no reference to the requirement that a person is not to be detained for the sole reason that he or she is subject to a Dublin procedure; nor does it say that the only permissible basis for detaining such a person is that there is a significant risk of absconding. Further, there is no direction that detention must be proportional.

I accept that paragraph 55.1.1 of Chapter 55 states that, in considering the power to detain, there is a presumption in favour of immigration bail and that, where possible, alternatives to detention should be used. It also states that detention is most usually appropriate to effect removal, initially to establish a person’s identity or basis of claim or where there is reason to believe that a person will fail to comply with any conditions attached to the grant of immigration bail. So too, paragraph 55.1.3 makes clear that detention must be used sparingly and for the shortest necessary period; and paragraph 55.1.4 says that detention must be for one of the statutory purposes for which the power is given and must comply with the limitations imposed by domestic and ECHR case law. But all of this amounts to no more than general guidance as to how the power to detain is to be exercised and does not constitute a set of objective criteria against which the risk of absconding is to be assessed. Nor does it set out the limits of the flexibility of the authorities in the assessment of the particular circumstances of each case in a manner which is binding and known in advance.

Paragraph 55.1.4.1 of Chapter 55 sets out the Hardial Singh principles but, as the majority of the Court of Appeal explained at paras 167 to 169, by the end of the hearing before them it appeared not to be in dispute that these principles were insufficient to satisfy the requirements of articles 28(2) and 2(n). Nevertheless, the majority in the Court of Appeal went on to explain why that was in any event their view, and it is a view which I share. In short, the Hardial Singh principles require the power to detain to be exercised reasonably and for the prescribed purpose of facilitating deportation: see, for example, Lumba , para 30, per Lord Dyson. But they do not constitute objective criteria on the basis of which an assessment may be made as to the likelihood that a person who is subject to a transfer procedure may abscond.

The Secretary of State places particular reliance on paragraphs 55.3 and 55.3.1 of the EIG. Paragraph 55.3 does no more than set out further general guidance to the effect that, for detention to be justified, there must be strong grounds for believing a person will not comply with conditions of temporary admission or temporary release; that all reasonable alternatives to detention have been assessed; and that each case has been considered on its merits, with due regard to the need to safeguard and promote the welfare of any children involved. The Secretary of State therefore focuses on the criteria set out in paragraph 55.3.1 and contends these are objective and publicly accessible and that their meaning will, if necessary, be determined by the court. It is also argued that they have formed the basis of detention decisions for many years and are well known and well understood by asylum seekers and the courts. So too, the argument continues, binding guidance as to the application of these criteria is set out in case law, and the courts have ensured that the decision-making process is transparent. The Secretary of State accepts that this list relates not just to Dublin III cases but also to all other cases in which it may be appropriate to consider detention for immigration purposes but contends this is not incompatible with article 2(n). It is submitted that this article does not say that the objective criteria have to be set out in a separate stand-alone document focused exclusively on article 28, nor that there has to be an express reference to that provision. Further, an individual who is subject to the Dublin III procedure can readily identify which of the criteria in the list are relevant to any assessment of whether he poses a risk of absconding.

I do not find these submissions persuasive. Paragraph 55.3.1 sets out 11 factors which may be relevant in considering the need for initial or continued detention. It does not purport to be a list of criteria for assessing whether a person in a Dublin III procedure may abscond. Further, it is not an exhaustive list for, as its opening words make clear, all relevant factors are to be taken into account. As for the list itself, only one factor, the second, refers in terms to absconding. Some of the others might be relevant to a risk of absconding but might also be relevant to the need for detention for different purposes underpinned by different policy considerations. In the case of the last three (whether the subject is under 18, whether the subject has a history of torture and whether the subject has a history of physical or mental ill health), it is hard to see their relevance to a risk of absconding at all. In the result, persons subject to a Dublin III procedure cannot know which criteria will be used for the basis of an assessment whether they are likely to abscond and cannot identify the limits of the flexibility of the relevant authorities in carrying out their evaluation. In my judgement a provision such as this does not constitute a framework of certain predetermined limits for assessing whether a person in a Dublin III procedure is likely to abscond, does not identify the criteria which are to form the basis of the assessment and does not set out the limits of the flexibility of the authorities in a manner which is binding and known in advance. Nor is there any evidence or finding that asylum seekers were aware of the provisions of the Dublin III Regulation or the existence or significance of the Al Chodor decision and so could in some way factor these matters into their understanding of the assessment processes to which they were subjected.

The Secretary of State also points to the subsequent legislative history and submits this shows how little difference formal compliance with any requirement for secondary legislation would make. On the 15 March 2017, the day the CJEU gave judgment in Al Chodor , the Secretary of State made the Transfer for Determination of an Application for International Protection (Detention) (Significant Risk of Absconding Criteria) Regulations 2017 (SI 2017/405) (“the 2017 Regulations”). These set out, at regulation 4, the criteria to be considered when deciding whether a person (“P”) poses a significant risk of absconding for the purposes of article 28(2):

The Secretary of State contends that these criteria reflect the relevant bullet points referred to in paragraph 55.3.1 and observes that paragraph 55.3.1 could have been drafted in precisely these terms. This, so it is said, shows how technical and formal the following argument of the respondents is: promulgation of the criteria by means of a policy such as the EIG does not comply with articles 28(2) and 2(n), yet the making of secondary legislation in the same terms by the Secretary of State alone and without any Parliamentary scrutiny does comply with them.

It will be appreciated that this argument is primarily directed at the second stage of the analysis as explained at para 56 above, and I address this below. But in my view the 2017 Regulations are also relevant to the first stage. The contrast between regulation 4 of the 2017 Regulations and paragraph 55.3.1 of the EIG is striking. Paragraph 55.3.1 contains a non-exhaustive list of criteria, only some of which may be relevant to an assessment of the risk of an applicant absconding for the purposes of article 28(2). Regulation 4, on the other hand, says that the Secretary of State must consider the factors which are identified and the potential relevance of them all to such an assessment is plain.

In summary and for the reasons I have given, I am satisfied that Chapter 55 of the EIG does not establish objective criteria for the assessment of whether an applicant for international protection who is subject to a Dublin III transfer procedure may abscond; its contents do not constitute a framework with certain predetermined limits; and it does not set out the limits of the flexibility of the relevant authorities in assessing the circumstances of each case in a manner which is binding and known in advance. It follows that Chapter 55 of the EIG cannot satisfy the requirements of articles 28(2) and 2(n) of the Dublin III Regulation and the majority of the Court of Appeal were right so to hold.

In the light of the foregoing it is not strictly necessary to deal with this further stage of the analysis. Nevertheless, since we heard argument about it, I will address it.

The Secretary of State contends that the circumstances of these appeals are very different from those of Al Chodor . It is argued that Chapter 55 of the EIG is not mere administrative practice and that in reality it is prescriptive and imposes restrictions on the executive power to detain that go beyond the Hardial Singh principles. Compliance with the provisions of Chapter 55 is enforceable by individuals before the courts and an unlawful failure to comply will result in the detention being held to be unlawful and lead to an order for the release of the person concerned or an award of compensation, or both. It is also contended that, in English law, settled case law defines the legal powers which limit the statutory power to detain and permits enforcement of the criteria which restrict the power of the executive to detain and which in other legal systems might exist only in legislation. As such, case law and Chapter 55 EIG are integral components of the law that governs and limits the power to detain and together define how the power to detain must be exercised. Therefore, Chapter 55 constitutes the kind of non-legislative instrument that the CJEU contemplated would satisfy article 28(2) and 2(n). Indeed, the submission continues, if a non-legislative instrument like Chapter 55 does not constitute the kind of non-legislative instrument that the CJEU contemplated would satisfy articles 28(2) and 2(n) then it is difficult to imagine what the CJEU was contemplating when it departed from the Advocate General’s view.

The circumstances of the Al Chodor case were plainly very different from those the subject of these appeals. It will be recalled that Czech law conferred on the police force a wide power to detain. Nevertheless, the national court had laid down in a series of judgments some criteria for assessing the risk of absconding. However, as the Advocate General explained, the presentation of these criteria was fragmentary. In addition, there was doubt as to whether that administrative practice had been publicised and it was alterable at will.

There can also be no doubt that in this jurisdiction a policy statement such as Chapter 55 of the EIG has significant legal effects. I have referred to these at paras 50-51 above. In broad terms and as Laws LJ explained in R (Nadarajah) v Secretary of State for the Home Department [2005] EWCA Civ 1363, para 68, where a public authority has issued a promise or adopted a practice which represents how it proposes to act in a given area, the law requires that promise or practice to be honoured unless there is good reason not to do so. Moreover, the court is the final arbiter of what a policy means : Kambadzi , at para 36, per Lord Hope; Mandalia v Secretary of State for the Home Department [2015] UKSC 59; [2015] 1 WLR 4546, para 31, per Lord Wilson of Culworth. It is also well established that compliance with such a policy is enforceable by individuals before the courts.

Moreover, the word “law” is used in articles 5(1)(f), 8(2), 9(2), 10(2) and 11(2) of the ECHR which require that any interference with the rights affirmed by these provisions be in accordance with “a procedure prescribed by law”, “in accordance with the law” or “prescribed by law”. The meaning of each of these expressions is the same and the word “law” within them encompasses not only legislation but also case law. As the ECtHR explained in Sunday Times v United Kingdom (1979) 2 EHRR 245, para 47, were it otherwise it would deprive a common law state which was party to the ECHR of the protection of article 10(2) and strike at the roots of that state’s legal system. Indeed, the applicants in that case did not argue that the expression prescribed by law required legislation in every case; they contended that legislation was required only where the common law rules were so uncertain that they did not satisfy the requirement of legal certainty. The court went on to explain, at para 49, that two of the requirements that flow from the expression “prescribed by law” are that a law must be accessible and sufficiently precise to enable a person adequately to foresee the consequences of his actions and so regulate his conduct. But, the court continued, the consequences need not be foreseeable with absolute certainty; that was unattainable and might carry with it excessive rigidity, preventing the law from keeping pace with changing circumstances.

Similarly, R (Munjaz) v Mersey Care NHS Trust [2005] UKHL 58; [2006] 2 AC 148 concerned a policy adopted by an authority for the seclusion of detained psychiatric patients. The House of Lords held any interference with the article 8 rights of patients was justified under article 8(2). Seclusion under the policy was necessary for, among other things, the prevention of disorder, the protection of health and the protection of the rights and freedoms of others, and if properly used it would not be disproportionate. The procedure adopted by the authority did not permit arbitrary or random decision making and the rules were accessible, foreseeable and predictable. In these circumstances, it could not be said that they were not in accordance with or prescribed by law. Lord Bingham of Cornhill, with whom Lord Hope and Lord Scott of Foscote agreed, rejected a submission that the interference was not in accordance with the law because it was not prescribed by a binding general law:

Nevertheless, in my view it does not follow that Chapter 55 of the EIG and domestic case law constitute “law” within the meaning of article 2(n) of Dublin III. That is so because a provision can only amount to a “law” within the meaning of article 2(n) if it has the necessary quality of certainty and that is something that Chapter 55 does not have. To ignore the need for certainty would be impermissibly to remove the word “law” from its context.

As I have explained, the Advocate General identified the two objectives of the requirement that the criteria for assessing the risk of absconding be defined by law as being first, to ensure that those criteria offer sufficient guarantees in terms of legal certainty, that is to say that the measures adopted by member states enable the individuals concerned to ascertain the scope of their rights and obligations and to foresee the consequences of their actions; and secondly, to ensure that the discretion enjoyed by the individual authorities responsible for applying those criteria is exercised within a framework of pre-determined limits. All of this reasoning is echoed in the judgment of the CJEU, in particular at paras 41 to 43. The CJEU also explained, at para 44, that the adoption of rules of general application would provide the necessary guarantees in so far as they set out the limits of the flexibility of the authorities in a manner which is binding and known in advance, and further, that the criteria are best placed for the external direction of the discretion of those authorities for the purpose of protecting applicants against arbitrary decision making. In this latter connection it referred, with apparent approval, to the observations of the Advocate General at paras 81 and 82 of his Opinion (see para 33 above).

In my judgement and for the reasons I have given at paras 57-65 above, it is clear that Chapter 55 does not satisfy these requirements. It does not set out the limits of the flexibility of the relevant authorities in assessing the circumstances of each case in a manner which is binding and known in advance and so lacks the necessary qualities of certainty and predictability. Therefore, it does not constitute a “law” for the purposes of articles 28(2) and 2(n).

A broader question is whether a statement of policy and public law adherence to it can ever amount to a binding provision of general application and so a “law” within the meaning of article 2(n). The Secretary of State maintains that it can for the reasons summarised at paras 67-71 above. Reliance is also placed on the 2017 Regulations. It is said that the respondents’ argument that promulgation of the criteria set out in the 2017 Regulations by means of a policy would not comply with articles 2(n) and 28(2) whereas the making of secondary legislation in the same terms would so comply, shows how technical and formal the objection is.

The respondents have advanced powerful arguments to the contrary, however. First, it is a feature of the policy-adherence principle that the decision maker is entitled for good reason to depart from the policy: see, for example, Lumba at para 54, per Lord Dyson (with whom Lord Hope, Lord Walker of Gestingthorpe and Lord Collins of Mapesbury agreed); at para 202, per Baroness Hale of Richmond; at para 245, per Lord Kerr of Tonaghmore; at para 351, per Lord Brown of Eaton-Under-Heywood (with whom Lord Rodger of Earlsferry agreed); and at para 312, per Lord Phillips of Worth Matravers. Hence, the respondents continue, a statement of policy cannot be a principle of general application.

The respondents’ second argument focuses on the view expressed by the Advocate General in Al Chodor [2017] 3 CMLR 24, paras 81 and 82 of his Opinion that the discretion of the authorities should be circumscribed in such a way as would best guard applicants against arbitrary deprivations of liberty, and so the content of the criteria and their application in a particular case should be decided by institutionally separate authorities (see para 33 above). Further, in explaining why a provision of general application is required, the CJEU referred to this aspect of the Advocate General’s Opinion with apparent approval ( Al Chodor [2017] 4 WLR 125, para 44). The respondents also point out that, under United Kingdom constitutional arrangements, a statement of immigration policy and the common law principle of policy-adherence do not involve external direction and that, as a matter of constitutional theory, under the Carltona doctrine ( Carltona Ltd v Comrs of Works [1943] 2 All ER 560), the decisions of the Secretary of State’s officials count as his. It follows, say the respondents, that a statement of policy cannot be a binding general law.

The third argument advanced by the respondents is directed at the reliance by the Secretary of State upon the position under the ECHR that interferences with human rights must be “prescribed by law”, and how this requirement has been interpreted. The respondents contend that this reliance is misplaced because articles 28(2) and 2(n) require more than the conventional ECHR standards of prescription. They also say, fairly in my view, that what is needed is a “ high level of protection ” ( Al Chodor [2017] 4 WLR 125, paras 34 and 43), that there are to be “ strict safeguards ” ( Al Chodor , para 40), and that there are to be “ greater guarantees ” than applied under Dublin II ( Al Chodor , para 35). They contend that, in giving this guidance and contrary to the position taken by the Secretary of State, the CJEU plainly had in mind the settled case law of common law systems and that such is apparent from its reference ( Al Chodor , para 21) to the decision of the ECtHR in Kruslin v France (1990) 12 EHRR 547 in which that court discussed the role of case law in both civil law and common law systems.

It is not necessary to resolve these rival contentions in this appeal and in my judgement the question whether a statement of policy and public law adherence to it can ever amount to a binding provision of general application and so a “law” within the meaning of article 2(n) should be decided in a case in which it is necessary to do so.

For the reasons I have given, the policy published by the Secretary of State in Chapter 55 of the EIG does not satisfy the requirements of articles 28(2) and 2(n) of the Dublin III Regulation.

The right to personal freedom is of fundamental importance and is reflected in the guarantees contained in articles 5(4) and 5(5) of the ECHR. A person who is unlawfully detained in this jurisdiction has (a) a right to be released; and (b) a right to damages for the tort of false imprisonment. This tort has just two ingredients: the fact of imprisonment and the absence of lawful authority to justify it: R v Deputy Governor of Parkhurst Prison, Ex p Hague [1992] 1 AC 58, 162C-D per Lord Bridge of Harwich. All that a claimant has to prove in order to establish false imprisonment is that he was directly and intentionally detained by the defendant, whereupon the burden shifts to the defendant to show that there was lawful justification for doing so: Lumba , para 65, per Lord Dyson.

Here the Secretary of State relies upon the discretionary power to detain which is conferred by paragraph 16(2) of Schedule 2 to the 1971 Act. But that reliance can only assist the Secretary of State so far as the power to detain was exercised in accordance with the 1971 Act. As we have seen, the power is limited in various ways. It is limited by the Hardial Singh principles such that a failure by the Secretary of State to comply with those principles will render the detention unlawful. So too, a failure by the Secretary of State to adhere to a published policy under the 1971 Act without good reason can amount to an abuse of power which will render the detention unlawful: Kambadzi , paras 41-42, per Lord Hope.

The respondents’ primary submission, which the majority of the Court of Appeal accepted, is that the exercise by the Secretary of State of the power to detain under the 1971 Act is also constrained by any applicable obligations under European Union law by operation of section 2(1) of the European Communities Act 1972. The respondents were detained pursuant to the policy in the EIG which was unlawful in so far as it failed to give effect to articles 28(2) and 2(n) of the Dublin III Regulation. Put another way, nothing in the 1971 Act authorised an exercise of the power to detain in breach of European Union law. The detention of the respondents under the 1971 Act and pursuant to the policy in the EIG was in breach of Union law. It follows that they were detained without lawful authority and their detention amounted to false imprisonment, and they are entitled to damages.

This argument is clear and, in my opinion, compelling. However, the Secretary of State argues that it is simplistic and that, were it to be accepted, it would ride roughshod over the careful balance of interests that is inherent in the principles of European Union law which govern and restrict the availability of damages as a remedy for breach by a member state of Union law, particularly where the precise effect of that law is unclear until it has been established by a decision of the CJEU. There are three limbs to the Secretary of State’s submissions.

It is contended first, that there are only two ways in which a claimant in domestic proceedings is entitled to damages payable by a member state for a breach of European Union law: either where the law specifies the penalties to be imposed or the compensation to be provided in the event of breach, or under the principles established by the CJEU in Francovich v Italian Republic (Joined Cases C-6/90 and C-9/90) [1995] ICR 722. Here, the Secretary of State continues, the Dublin III Regulation does not provide for compensation for its breach and so the respondents’ only possible claim under Union law is for Francovich damages.

Francovich was a case concerning the obligation upon a member state to take, within a given period, the measures necessary to implement a Directive. The court explained that it is a principle of European Union law that member states are obliged to pay compensation for harm caused to individuals by breaches of the law for which they can be held responsible, but that the conditions under which that liability gives rise to a right to compensation depend upon the nature of the breach giving rise to the claim. This issue and in particular the approach to be adopted where the legislature of the member state has a wide discretion when acting in a field governed by Union law, was explored further by the court in Factortame . The court explained that Union law confers a right to reparation where: the rule of law infringed is intended to confer rights on individuals; the breach of the obligation resting on the state is sufficiently serious; and there is a direct causal link between the breach and the damage sustained by the injured parties: Factortame , para 51. The decisive test for determining whether the breach is sufficiently serious is whether the member state manifestly and gravely disregarded the limits on its discretion: Factortame , para 55. The factors which the national court may take into consideration include the clarity and precision of the rule breached; the measure of discretion left by the rule to the national authority; whether the infringement and the damage caused was intentional or involuntary; whether any error of law was excusable or inexcusable; the fact that the position taken by a European Union institution may have contributed towards the omission; and the adoption or retention of national measures contrary to European Union law: Factortame , para 56. On any view, a breach of European Union law will be sufficiently serious if it is persisted in once it is clear that the impugned conduct constitutes an infringement: Factortame , para 57.

The Secretary of State has invoked all of these principles on this appeal. It is contended that European Union law was impenetrable before the decision of the CJEU in Al Chodor and that this is highly relevant to the question whether any breach by the United Kingdom of Union law was intentional or voluntary, or excusable or inexcusable, and therefore sufficiently serious to trigger a liability for Francovich damages. The Secretary of State also points to the wide margin of discretion given to member states in giving effect to articles 28(2) and 2(n), and contends that in circumstances such as those of this appeal a manifest and grave disregard of the limits of the discretion must be established before damages are available; and that the same criteria should inform, if not govern, a claim for damages for false imprisonment under common law.

There can be no dispute about the correctness of the principles of European Union law which I have summarised in para 86. I accept too that it is by reference to these principles that any claim by the respondents for damages under Union law must be judged. But it does not follow that the same principles constrain the claim by the respondents for damages for wrongful imprisonment and in my judgement and for the reasons which follow, they do not.

First, the consequence of a failure by a member state to establish, in a binding provision of general application, objective criteria underlying the reasons for believing that an applicant for international protection who is subject to a Dublin III procedure may abscond, is that article 28(2) of the Regulation does not apply. This in turn means that the detention of such an applicant in such a state is unlawful and he or she must be released: see Al Chodor , paras 17 and 46. In this appeal, it has the consequence that the decision to detain the respondents lay outside the boundaries of any permissible exercise of the power to detain conferred by paragraph 16(2) of Schedule 2 to the 1971 Act.

Secondly and as I have explained, the right to liberty is a fundamental human right enshrined in article 5 of the ECHR. Immigration detention is only in accordance with article 5(1)(f) in so far as it is in accordance with a procedure prescribed by law. Moreover, a person who is detained unlawfully must be released: article 5(4); and is entitled to compensation: article 5(5). These principles are also general principles of European Union law: article 6(3) of the Treaty on European Union.

Thirdly, the right to compensation is provided in domestic law by the right at common law to claim damages for false imprisonment. Generally, damages for false imprisonment are awarded as compensation and so the level of damages will depend on the circumstances and degree of harm the claimant has suffered by reason of his or her wrongful detention. There is no reason to believe that the impact of loss of liberty is likely to be affected by whether lack of legal authority for the detention is the consequence of a failure to comply with European Union or domestic legislation, and in my judgement the source of the lack of legal authority does not justify treating those who have been wrongfully detained differently from one another.

The second limb to the Secretary of State’s submissions is founded on the decision of this court in Lumba . It is argued that in Lumba this court rejected the submission that any public law error in a decision to detain would result in the subsequent detention being unlawful, regardless of any of the circumstances of that public law error, and instead adopted an approach which involved the weighing of a number of countervailing considerations, such as the nature and extent of the public law error, the absence of procedural safeguards which are normally available in cases of judicial review but are not available in a private law action for damages for false imprisonment, and the discretionary nature of judicial review remedies. We are urged to adopt the same approach in this appeal in considering the elements of the tort of false imprisonment and correct approach to the assessment of damages, if liability is established. It is submitted that such an approach echoes that of the CJEU in Factortame when formulating the necessary elements of a claim for damages for a breach of European Union law.

Lumba was a case in which the Secretary of State applied an unpublished policy of detention for all foreign national prisoners on completion of their sentences of imprisonment and pending the making of deportation orders against them. This court held that the unpublished policy was unlawful because it was a blanket policy which admitted of no exceptions and was inconsistent with the Secretary of State’s published policy. It also held that a public law error can found a claim for damages for false imprisonment but recognised that not all public law errors will have this effect. The majority expressed themselves in slightly different ways. Lord Dyson explained, at para 68, that the error must be material to the decision to detain:

Lord Hope considered, at para 175:

Baroness Hale put it this way, at para 207:

Lord Kerr stated the test in these terms, at para 251:

All of these formulations have at their heart a recognition that a public law error will not render detention unlawful unless that error bears upon and is relevant to the decision to detain, and so is capable of affecting the result. Similar expressions were adopted subsequently in Kambadzi at paras 41-42 (Lord Hope), para 69 (Baroness Hale) and para 88 (Lord Kerr).

In my judgement, there can be no doubt that the test laid down in Lumba and Kambadzi , however expressed, is met in the circumstances of the cases before us in this appeal. There was a requirement for a binding provision of general application containing objective criteria underlying the reasons for believing that an applicant might abscond, and that requirement was not satisfied. This was fundamental to the decision to detain and it can make no difference whether the source of that requirement lay in European Union or domestic legislation.

The third limb to the Secretary of State’s submissions adopts the reasoning in the dissenting judgment of Sales LJ in the Court of Appeal at paras 132 to 147. The essential elements of that reasoning are these:

The Dublin III Regulation does not stipulate that damages are to be awarded for detention in breach of its terms.

The test laid down in Factortame for attaching liability to a state to pay damages to an individual for a breach of European Union law reflects the fact that Union legislation is frequently not clear.

The domestic tort of false imprisonment was framed without reference to the particular problems to which the Dublin III Regulation gives rise, and ignores the fair balance of interests which the Regulation aims to achieve.

In effect, a choice of law question arises when assessing whether a person within article 28 of Dublin III who has been wrongfully detained is entitled to substantial damages, and the appropriate law to govern that question is Union law.

A person who is subject to a Dublin III procedure and who has been wrongfully detained must be released but such a person is only entitled to damages if there has been a sufficiently serious breach of article 2(n), and any other approach would be disproportionate. Put another way, article 28 excludes any application of the power to detain conferred by Schedule 2 of the 1971 Act. The detention of such a person is therefore completely covered by Union law and that person is only entitled to damages under Union law.

The disconnection between Union law as set out in articles 28(2) and 2(n) of the Regulation and the domestic law of false imprisonment is illustrated by the case of the third respondent who was refused bail and so was detained by order of the court, and could not claim damages for false imprisonment in respect of his detention thereafter. However, he could still claim damages for breach of Union law if the breach was sufficiently serious.

The Secretary of State’s position is also supported by the fact that damages for a breach of article 6 of the ECHR fall to be assessed by reference to ECtHR authority.

These arguments overlap to a considerable extent with those I have already addressed. I of course accept that the Dublin III Regulation does not require member states to confer a right to damages on persons who have been detained in breach of its terms. So too I recognise that one of the matters informing the formulation by the CJEU of the conditions under which a member state may incur liability for damage caused to individuals by a breach of Union law is whether the state concerned manifestly and gravely disregarded the limits on its discretion, and that one of the factors which may be relevant to this issue is the clarity and precision of the rule breached: Factortame , paras 55-56. I also accept that the domestic tort of false imprisonment was framed without reference to the particular problems to which the Dublin III Regulation gives rise.

These points aside, however, I cannot agree with Sales LJ’s analysis. The power to detain applicants for international protection who are subject to a Dublin III procedure is conferred, not by the Dublin III Regulation, but by Schedule 2 to the 1971 Act. That power to detain is constrained in various ways, three of which I have discussed: the Hardial Singh principles, the policy-adherence principle and the provisions of the Dublin III Regulation. Here the Secretary of State’s published policy in Chapter 55 of the EIG did not comply with articles 28(2) and 2(n) of the Dublin III Regulation with the consequence that, in the case of each of the respondents, the decision to detain lay outside the scope of any legitimate exercise of the discretion conferred by Schedule 2 to the 1971 Act. In these circumstances, the two ingredients of the tort of wrongful imprisonment were undoubtedly present. As the respondents submit and I accept, the right under domestic law to claim damages for wrongful imprisonment is not dependent on the law being clear. Nor is it dependent upon the nature of the illegality, that is to say whether it is the consequence of a failure to comply with European Union legislation, as in this case, or has some other cause, as it did in Lumba .

Further, there is no disconnection between a failure to comply with articles 28(2) and 2(n) of Dublin III Regulation and the tort of false imprisonment in circumstances such as those of the cases before us. Nor can the Secretary of State derive any assistance from the position of the third respondent. He was not detained pursuant to an order of the court. He was simply denied bail. A decision on a bail application is not a determination of whether or not the detention is lawful, whether at common law or for the purposes of article 5(4) of the ECHR: see, for example, Lumba at para 118.

The approach adopted in this jurisdiction to claims for damages for violations of article 6 of the ECHR does not assist the Secretary of State either. Sales LJ referred to the decision of the House of Lords in R (Greenfield) v Secretary of State for the Home Department [2005] UKHL 14; [2005] 1 WLR 673. In that case Mr Greenfield, a prisoner, failed a mandatory drug test and was charged with a drugs offence under the relevant prison rules. The charge was heard by the deputy controller, a Crown servant, for whom the Secretary of State was responsible. The deputy controller refused a request by Mr Greenfield that he be legally represented. The charge was proved and Mr Greenfield was ordered to serve an extra 21 days of imprisonment. He applied for judicial review of the decision, alleging that his rights under article 6 of the ECHR had been violated. In due course the Secretary of State conceded there had been a breach of article 6 on the basis that the proceedings involved a criminal charge, that the deputy controller was not an independent tribunal and that Mr Greenfield had been denied legal representation of his own choosing. Mr Greenfield nevertheless pursued his claim for damages for the violations of article 6 which had taken place. The House of Lords held that, in deciding whether an award of damages was necessary, it was appropriate to look to the jurisprudence of the ECtHR for guidance, and in the great majority of cases in which that court had found a breach of article 6 it had treated the finding of violation as, in itself, just satisfaction under article 41, and that it would only award monetary compensation where it was satisfied that the loss or damage was caused by the violation.

The important point of difference between Greenfield and the cases before us on this appeal is that, in Greenfield , the claim for damages was based entirely on the breach of article 6. There was no claim for damages for wrongful imprisonment or for any other tort and Mr Greenfield had not suffered any loss. Indeed, as Lord Bingham of Cornhill explained at paras 27 to 29, the hearing had been conducted in an exemplary manner and, while it could be accepted that Mr Greenfield thought that the authorities were biased against prisoners and that he would not receive a fair hearing, the manner of his adjudication had been the norm, he had been treated no differently from any other prisoner and there was no feature of the case which justified an award of damages.

In my judgement the majority in the Court of Appeal were therefore right to hold that the respondents were wrongfully detained. The respondents are also entitled to compensation for any loss their wrongful detention has caused them.

The Secretary of State contends that the respondents should be awarded no more than nominal damages. It is said that it is inevitable that the respondents would have been detained lawfully, had the Secretary of State appreciated the unlawfulness of Chapter 55 of the EIG. That is demonstrated by the 2017 Regulations, which came into force on 15 March 2017, the day the CJEU gave judgment in Al Chodor . Further, the argument continues, these regulations meet all of the criteria set out in articles 28(2) and 2(n) of the Dublin III Regulation as interpreted by the CJEU in Al Chodor , and their application would have resulted in the same outcome: the respondents would have been detained.

The Secretary of State relies in support of this contention upon the decision of this court in Lumba . One of the issues in that appeal was whether the appellants had suffered any loss as a result of their wrongful imprisonment. The majority of the court held they had not because, had the Secretary of State acted lawfully and applied her published policy on detention as opposed to her unpublished policy of blanket detention, it was inevitable that the appellants would have been detained. In other words, the detention of the appellants was at all times justifiable.

The same point emerges from the decision of this court in Kambadzi . There the claimant’s detention pending deportation was unlawful because it had not been reviewed in accordance with the Secretary of State’s published policy and rule 9(1) of the Detention Centre Rules 2001 (SI 2001/238). Although it could be no defence to a claim for false imprisonment to demonstrate that, if reviews had been carried out, the claimant would still have been detained, this would be relevant to the claim for damages. Lord Hope said this at para 55:

Baroness Hale of Richmond summed up the position this way at para 74:

Similarly, Lord Kerr said this at para 89:

These principles were subsequently applied by the Court of Appeal in Parker v Chief Constable of Essex Police [2018] EWCA Civ 2788; [2019] 1 WLR 2238, another decision on which the Secretary of State relies. Here the claimant was arrested on suspicion of murder and rape. The investigating officer was delayed by traffic so the arrest was carried out by a surveillance officer who was present at the scene but did not personally have reasonable grounds for suspecting the claimant was guilty of an offence, as required by section 24(2) of the Police and Criminal Evidence Act 1984. It was perfectly clear that, had this requirement been appreciated, it could and would have been met, either by waiting for the investigating officer to arrive or by properly briefing the officer who carried out the arrest. Accordingly, the claimant could only recover nominal damages.

In my view the Secretary of State is seeking to apply these principles well beyond their proper limits. In Lumba , this court considered what would have happened had the Secretary of State applied his published policy. In Kambadzi , the question was whether the claimant would have been detained had regular reviews been carried out. In Parker , it was established that, had things been done as they should have been, the claimant could and would have been arrested lawfully. In other words, a claimant will be awarded nominal damages if it is established that the detention could have been effected lawfully under the existing legal and policy framework. Article 5(1) of the ECHR requires any deprivation of liberty to have a legal basis in domestic law, and that law must be sufficiently precise and accessible in order to avoid all risk of arbitrariness: see Dougoz v Greece (2002) 34 EHRR 61, para 55. Similarly, a person is entitled to know what the law and any policy made under it is, so he or she can make relevant representations in relation to it: see Lumba , at paras 34-36, per Lord Dyson. It can be no answer to a claim for damages for unlawful imprisonment that the detention would have been lawful had the law been different.

The respondents say that they are also entitled to damages pursuant to European Union law in the light of the principles explained by the CJEU in Francovich and Factortame . It is not contended that any award of damages for such a breach would exceed those payable for false imprisonment. It is therefore not necessary to consider this alternative claim in this appeal.

The respondents’ claims do not require remittal for any further consideration of the lawfulness of their detention. They were all detained unlawfully and are entitled to damages under domestic law for false imprisonment. I would transfer these proceedings to the County Court for the assessment of the quantum of those damages, if that quantum cannot be agreed.

I would dismiss this appeal.