C3 & Anor v The Secretary of State for Foreign, Commonwealth & Development Affairs

I should as a preliminary set out the standard terms of the writ of habeas corpus ad subjiciendum, as it appears in the Civil Procedure Forms in the current edition of the Supreme Court Practice:

The writ of habeas corpus has a long and important history in English law as a remedy against unlawful detention. We are, however, on this appeal concerned only with the specific question of the circumstances in which it lies against a person who does not have actual custody of the applicant but who is said nevertheless to have the power to procure their release. We can ignore the case of detention by someone who is the respondent’s agent or employee: in such a case there is obviously no difficulty in treating the principal or employer as the responsible person. But the situation is less straightforward where the actual detainee is an independent actor but the respondent is said nevertheless to have de facto control over the detention. That question is addressed in a small number of authorities to which we were referred in some detail. In my view only three directly address the relevant principles – Barnardo v Ford [1892] AC 326; R v Secretary of State for Home Affairs, ex p O’Brien [1923] 2 KB 361, [1923] AC 603; and Rahmatullah v Secretary of State for Defence [2012] UKSC 48, [2013] 1 AC 614. I will consider them first, but it is necessary also to refer to two other authorities – Ex p Mwenya [1960] 1 QB 241 and Re Sankoh (unreported 1.9.00 (HC) and 27.9.00 (CA)) – which illustrate how those principles have been applied in particular circumstances.

In Barnardo v Ford a boy had been placed in a Barnardo’s home, but his mother later asked for him to be moved elsewhere. She was told by Dr Barnardo, as he confirmed on affidavit, that the child had been transferred to the care of a Mr Norton, who had taken him to Canada, and that he did not know where Mr Norton was and had no means of communicating with him. It appears to have been accepted that the boy was no longer in the Barnardo’s home, but the mother did not accept that Dr Barnardo had no control over his current circumstances. The House of Lords upheld the order of the High Court for the issue of a writ of habeas corpus for the purpose of establishing, as Lord Macnaghten put it (see p. 340), “the circumstances under which Dr Barnardo parted with the child, and ascertaining beyond all doubt whether the child is or is not still under Dr Barnardo’s control or within his reach”.

The primary point established by Barnardo v Ford is that a writ of habeas corpus can properly be issued where there is a doubt about whether the respondent does in fact have custody of the putative detainee. In such a case the respondent will in the return to the writ required by CPR 87.10 state, if that remains their case, that they do not have custody and adduce appropriate supporting evidence, on which they will be liable to be cross-examined on the return date: as to this, see the speech of Lord Herschell at p. 340.

Equally important for our purposes is that the question that would be examined by the Court on the return date was not defined by the House simply as whether the child was in Dr Barnardo’s actual custody (which on the face of it he was not) but extended to whether he was in his “power or control” – a phrase used by Lord Herschell at p. 340, which is plainly to the same effect as Lord Macnaghten’s language of “control … or reach”.

A related point which emerges from Barnardo v Ford is that the sole purpose of the writ is to procure the release of a person currently being detained, unless their detention were shown to be lawful, and not to investigate the lawfulness of any past detention with a view to punishment or compensation: see in particular the speeches of Lord Halsbury LC at pp. 332-333 and Lord Watson at pp. 333-334. That point is not directly significant for our purposes because the Applicants’ object is indeed to seek release from their current detention; but it is an illustration of the specific and limited nature of the habeas corpus jurisdiction.

In O’Brien the applicant, who was living in London, had originally been detained by the Home Secretary under emergency powers contained in the Restoration of Order in Ireland Regulations 1920. Under the terms of an arrangement with the newly independent Irish Free State, he was transferred to Dublin and interned by the Free State authorities in Mountjoy prison. His application for habeas corpus was dismissed by the High Court, but the Court of Appeal held both that his detention had not been authorised by the Regulations and that the writ should issue. As regards the latter decision, it held that it was at least arguable that the Home Secretary retained control over the applicant’s custody notwithstanding his transfer to Ireland, and accordingly, applying Barnardo v Ford, that the writ should issue in order that that question should be fully explored. A note in the King’s Bench reports records that the applicant was on the return date duly produced to the Court and discharged.

As regards the nature of the requisite “control”, the fullest account is in the judgment of Atkin LJ. He said, at p. 398 (I italicise three propositions of particular importance):

Bankes and Scrutton LJJ adopted essentially the same analysis. Bankes LJ described the question requiring further exploration as being “how far, if at all, by arrangement with the Free State Government the body of the applicant is under the control of the Home Secretary” (p. 381). Scrutton LJ referred in particular to the Home Secretary’s statement in the House of Commons which I have quoted above (see p. 392).

In the interval between the decision of the Court of Appeal and the return date of the writ the Home Secretary appealed to the House of Lords, and there was an expedited hearing. The majority held that it had no jurisdiction to entertain the appeal, but in his dissenting speech Lord Atkinson observed, as part of his reasons for concluding that the House did have jurisdiction, that (p. 624):

So far as relevant for our purposes, O’Brien is authority for three propositions, all established by the passage in the judgment of Atkin LJ which I have quoted (reinforced as regards the third by the dictum of Lord Atkinson):

In Rahmatullah the applicant was a Pakistani national who had been captured by British forces in Iraq. Under the terms of a Memorandum of Understanding (“the MoU”) between the British and the US armed forces he was transferred to the custody of the US; but cl. 4 of the MoU, reflecting article 45 of the Fourth Geneva Convention (“GC4”), required the US to return him on request. (There was a later MoU but it is immaterial for our purposes.) He was subsequently taken to Bagram airbase in Afghanistan, where he was held in custody by the US. He brought proceedings for habeas corpus against the Foreign and Defence Secretaries on the basis that his continuing detention was unlawful and that, although he was no longer in the custody of British forces, he was under the control of the UK Government because it had the right under the MoU to require his return. As to the allegation of unlawfulness, it was not part of his case that his initial detention by UK forces was unlawful; rather, he relied on the fact that article 49 of GC4 prohibits the removal of a detained civilian to a different country, so that his transfer to, and detention in, Afghanistan was unlawful.

That application was dismissed by the Divisional Court, but the Court of Appeal held that the writ should issue on the basis, following O’Brien, that it was necessary in order to resolve the question of what control over the applicant’s detention the UK Government in fact had ([2011] EWCA Civ 1540, [2012] 1 WLR 1462). I will refer to that as “the first CA decision”.

On the (adjourned) return date before the Court of Appeal the Secretaries of State informed the Court that a formal request had been made to the US Government requesting the return of the applicant to UK custody but referred it to a letter from the US Deputy Assistant Secretary of State for Defense, Mr Lietzau, which it submitted amounted to a definitive refusal of the request. The Court accepted that submission and held that in those circumstances it could make no further order ([2012] EWCA Civ 182, [2012] 1 WLR 1462, at pp. 1491-4): I refer to this as “the second CA decision”.

The Secretaries of State appealed to the Supreme Court against the first CA decision, and the applicant cross-appealed against the second decision. The case was heard by a panel of seven justices. Although both the appeal and the cross-appeal were dismissed, those conclusions were not reached by the same route by all the members of the Court. The position can be summarised as follows:

Rahmatullah is important because it gives the authority of the Supreme Court to the decision and reasoning in O’Brien. But the Applicants also relied on paras. 41-44 of Lord Kerr’s judgment, where he makes some introductory observations about the remedy of habeas corpus before going on to consider the issue of “control”. These read:

In Mwenya a politician detained by the authorities in Northern Rhodesia applied for habeas corpus both against the Governor and against the Secretary of State for the Colonies. There was an issue about whether the English courts had power to issue habeas corpus against the Governor: that turned on the status of Northern Rhodesia as a protectorate and is not material for our purposes. As regards the claim against the Secretary of State, the applicant’s case, relying on Barnardo v Ford and O’Brien, was that he had de facto power to procure his release.

The Divisional Court rejected that argument. Lord Parker CJ, after referring to the facts of O’Brien, said, at p. 279:

In Re Sankoh an opposition leader detained by the Government of Sierra Leone sought habeas corpus against the Foreign and Defence Secretaries on the basis that there was evidence that British forces operating in Sierra Leone were assisting the government, and might indeed have assisted in his detention, and that the UK Government might therefore have de facto power to obtain his release, so that the writ should issue in order to allow that question to be determined. Elias J dismissed the application on the basis that there was no evidence that the UK Government had any control over the applicant’s custody: the evidence was that he had not been originally arrested by British forces, nor was he at any time in their custody or control. It might have significant political influence but that was not enough: in that regard he cited Mwenya.

Elias J’s decision and reasoning were upheld by this Court. Laws LJ said, at para. 9:

The principles governing the obligations of the UK Government to take steps to seek to procure the release of a person detained by a foreign government are conveniently summarised in the decision of this Court in R (Abbasi) v Secretary of State for Foreign & Commonwealth Affairs [2002] EWCA Civ 159. The claimant was a British citizen detained by the US in its detention facility at Guantanamo Bay. He sought judicial review of the refusal of the UK Government to make representations to the US Government, or take other appropriate steps, to attempt to secure his release. The Court of Appeal upheld Richards J’s dismissal of his claim. At para. 106 of the judgment of the Court, delivered by Lord Phillips MR, the effect of the relevant case-law was summarised as follows:

Assistance of the kind being sought in Abbasi no doubt goes beyond the more usual examples of “consular assistance”, but the label does not seem to me inapt.

Ms Kaufmann submitted that the habeas corpus authorities summarised above clearly establish that the writ should be issued in the circumstances of the Applicants’ case. The AANES’s offer means that the Foreign Secretary has control over their custody in the sense established by O’Brien and endorsed in Rahmatullah, namely that he has the de facto power to effect their release. There is every reason to suppose that the AANES would indeed release the Applicants on receipt of an official request, consistently with the many other cases in which they have released detained women and children to their governments. Even if there were any doubt about the matter, the authorities establish that habeas corpus will lie in order for that doubt to be resolved.

The starting-point in considering that submission is that the passages about control – and more particularly de facto control – on which the Applicants rely in O’Brien and Rahmatullah must be read in the context of the issues in those cases. As Jay J says at para. 59 of his judgment below, “what is required is an examination of how [the statements of principle] were applied in individual cases and an identification of the particular features of those cases that either led to the issue of the writ, or not”. In both cases the UK Government was responsible for the original detention of the applicant but had transferred him to the custody of a foreign government on the basis (it appeared) of agreements which entitled it to require his re-transfer to its custody on request – that is, in O’Brien, the undertakings which the Home Secretary told Parliament that he had had from the Free State and, in Rahmatullah, cl. 4 of the MoU. The observations in both Courts about de facto control were made in that context, and more particularly in response to an argument that the agreements in question did not create enforceable legal rights. They cannot fairly be read as authority for a general proposition that in any case where A is being prima facie unlawfully detained by B habeas corpus will lie against C if there is reason to believe that C is for any reason able to procure A’s release. (I would add that such a reading would be inconsistent with both Mwenya and Sankoh.) The factual context here is different. The UK Government was not responsible for the Applicants’ detention and accordingly no question of an agreement to re-transfer on request arises.

I therefore do not believe that the outcome of this case is determined by the outcome in O’Brien and Rahmatullah. I have focused on those cases because the discussion of “control” was more developed there; but the same goes for Barnardo v Ford, where Dr Barnardo was the child’s original custodian.

It is accordingly necessary to consider whether as a matter of principle the effect of the AANES’s offer is to give the UK Government control over the Applicants’ detention in the sense necessary for habeas corpus to lie. I do not believe that it is. My reasons fall under two heads, though they overlap to some extent.

First, it is in my view of fundamental importance that the UK Government not only does not have actual custody of the Applicants but had no involvement in their original detention. For a writ of habeas corpus to issue in those circumstances would be unprecedented. In the course of his discussion of the “unexplored issue” in Rahmatullah Lord Phillips said (at paras. 104-105):

Lord Phillips is in that passage contemplating precisely the circumstances of the present case – that is, that the UK Government had nothing to do with the original detention but was, or might be, nevertheless in a position to procure their release. His observation that it is “at least questionable” whether habeas corpus would lie in such circumstances falls short of a definitive view on the question, but it carries real weight. (It may also be supported to a limited extent by Lord Reed’s statement that it was important that Mr Rahmatullah was initially detained by British forces; but it is fair to say that that statement is on its face directed to a different question.) The emphasis placed in Sankoh on the fact that the UK Government had no involvement in the applicant’s detention points in the same direction, though it may not be part of the ratio.

In my view the absence of any case in this jurisdiction (Footnote: ¹) in which the writ has issued in circumstances of this kind reflects a general understanding that it lies only against those responsible for the original detention. I believe that that understanding accords with the nature of habeas corpus as a remedy for unlawful detention. Where B has first detained A but has passed actual custody to C, it is fair to regard B as responsible for A’s continuing unlawful detention if in practice they have the power to procure his or her release by C: it was B who created the “detention situation” (if I may be forgiven the phrase) in the first place. But the case is different where B had nothing to do with the original detention. It is hard to see why in that case the fact that they may have, for an unconnected reason, de facto power to procure A’s release justifies subjecting them to a peremptory remedy for a situation which they have done nothing to create. That is all the more so given that if the writ is available at all it issues as of right, leaving no room for any exercise of discretion by the Court. B may of course in particular circumstances, as discussed in Abbasi, be under a public law duty to take steps to try to procure A’s release, but that is not the same thing as treating them as a constructive custodian for the purpose of habeas corpus.

Ms Kaufmann submitted that it should make no difference in principle whether B was responsible for the original detention: all that matters is whether they in fact have the power to procure A’s release. For the reasons given above I do not accept that. She also emphasised that, as Lord Kerr said in Rahmatullah, habeas corpus is a flexible remedy. That is no doubt the case, but it does not justify extending the scope of the remedy outside its principled boundaries.

Second, acceptance by the Foreign Secretary of the AANES’s offer would not by itself be effective to procure the Applicants’ release or their transfer to his custody. In the first place, the offer was not simply to release them: as Mr Hargreaves makes clear, they would be released only on the basis that the UK undertook to repatriate them. Secondly, that condition could only be satisfied by the Foreign Secretary taking the steps set out at para. 17 above – requiring, among other things, the transporting of the Applicants from Syria into Iraq or Turkey, with the permission of the government in question, in order to get them to an airport from which they could be flown to the UK. I would thus accept the Foreign Secretary’s submission that such control over the Applicants’ detention as the AANES’s offer gave him was qualified and conditional.

I do not believe that a power to procure the release of a detainee which is qualified in that way is sufficient to justify the issue of a writ of habeas corpus. I believe that it is a necessary part of the rationale underlying O’Brien and Rahmatullah that the control enjoyed by the constructive custodian is unqualified. As a matter of principle, it can only be because a respondent has an unconditional power (de facto if not de jure) to obtain the applicant’s (re-) transfer to their custody that it would be right to treat them as custodians. That was the basis of the claim in both O’Brien and Rahmatullah. In O’Brien the Home Secretary had told Parliament that the undertakings that he had been given by the Free State Government gave him “complete control”; and in Rahmatullah the right to require transfer in cl. 4 of the MoU was apparently unqualified. In neither case was there a need for the Government to take any steps of the kind necessary in this case in order to resume custody of the applicant if its request was acceded to: Mr O’Brien and Mr Rahmatullah would simply have been transferred from the custody of the one government to the other at the relevant port or airport (Footnote: ²).

Ms Kaufmann’s initial position before us was that it did not matter that the AANES’s offer was conditional: if it was in the power of the UK Government to comply with a condition of release required by a detainer it had de facto control of the detention and should be treated as a constructive custodian against whom habeas corpus would lie. Holroyde LJ put to her a case where a foreign organisation unlawfully detained a British citizen and offered to release them on payment of £10m – i.e. a hostage situation; other examples can readily be devised, such as release of prisoners or publication of a terrorist manifesto. She at first accepted that it followed from her position that in such a case habeas corpus would lie against the Government, who would accordingly be required to pay the ransom in order to bring the detainee before the Court. But on reconsideration she modified her response. She accepted that a conditional offer of release would normally not be sufficient to make the offeree a constructive custodian, but she said that the present case was different because the only condition imposed by the AANES was that the UK Government should arrange the repatriation of the Applicants, which was no more than the writ required in any event.

Ms Kaufmann denied that that argument was circular; but, whether it is or not, I would not accept it. The nature of the condition goes beyond a simple requirement to admit the Applicants to the UK. It requires the UK Government to take the various steps identified above, which, as the Foreign Secretary contends, involve the deployment of significant resources in terms of personnel, cost and diplomatic capital. That may not be as stark as making a payment of £10m, but the principle is the same. A decision on the part of the UK Government whether to take steps of the kind required in this case to procure the release of a British citizen held abroad involves an exercise of judgment which may be challengeable on the basis of the principles identified in Abbasi but which cannot simply be bypassed by the issue of a writ of habeas corpus. (Footnote: ³)

I would add that the artificiality of fitting the present situation into the framework of habeas corpus is illustrated by considering the Applicants’ position if the Foreign Secretary accepted the AANES’s offer. They would be released from the custody of the AANES at the moment of hand-over somewhere in its territory. But it is not clear to me in what sense they could be regarded as being thereafter in the custody of the UK Government. No doubt in practice they would in their own interests consent to being accompanied by its representatives on the journey back to the neighbouring country and until they had received their travel documents, and probably also as far as the airport; but that would be a matter of choice rather than compulsion. On arrival in the UK they would on the face of it be free to go wherever they liked (unless arrested on some lawful basis). That being so, I do not see what detention on his part the Foreign Secretary could be expected to justify on their production (actual or notional) to the Court on the return date. It may be that some analysis could be found that would cover the situation; but the difficulty of doing so reflects the fact that the relief that the Applicants really seek in their case is not simply their release from unlawful detention by the AANES but their repatriation.

In response to the Foreign Secretary’s skeleton argument in this Court the Applicants submitted a Supplemental Note in which they advanced a fallback argument which was not part of their case in the Divisional Court. This is that even if it was not enough to show that the Foreign Secretary had de facto power to procure their release, he should nevertheless be treated as having constructive custody on the basis that he was “responsible for” their continuing detention. The Court sought to explore the basis of this submission with Ms Kaufmann in her oral submissions. In the end she summarised her case as being that habeas corpus should lie because of what she described as a unique combination of four circumstances – (1) that the Foreign Secretary was in reality the only person who could bring about the Applicants’ release, because, given the conditions in North East Syria, the AANES could not responsibly release them except to the UK Government, which could arrange their and their children’s repatriation; (2) the shocking conditions of their detention; (3) the fact that the real reason for his unwillingness to make the request was his concerns about national security; and (4) that his conduct deprived them of their fundamental rights as citizens to enter the UK – she submitted that if the Applicants were able to present themselves at a consulate in Iraq or Turkey the Foreign Secretary would not be entitled to refuse to issue them and their children with passports. She acknowledged that “responsibility” might not be an apt label for that combination of factors and that she was in truth basing her case on the requirements of justice and humanity.

It is impossible not to feel the force of that appeal. The Applicants’ current situation is indeed dire, however much it may be the result of their own choices; and it is made worse that it is shared by their innocent children. There may be – though we are not in a position to form a view – a powerful case that neither the difficulties about the repatriation arrangements described by Mr Hargreaves nor (which appears to be the essential point) the Foreign Secretary’s concerns about the risk that the Applicants pose to national security can justify his refusal to take steps to secure their release and repatriation. But it is essential to remember the limited scope of the issue before us: see para. 6 above. I do not see how the factors relied on by Ms Kaufmann could justify granting the remedy of habeas corpus in circumstances where the Foreign Secretary does not have the necessary control over the Applicant’s detention. As already indicated, I believe that the only proper vehicle for such a case is a claim for judicial review, in which the Court would consider the lawfulness of the Foreign Secretary’s refusal of assistance in accordance with the principles in Abbasi.

Ms Kaufmann invited us, if that was the view that we took, not to dismiss the present proceedings but to remit them to the Divisional Court to be pursued as a claim for judicial review. I see no advantage in that course. The case would in practice have to be re-pleaded from the start, and supported by up-to-date evidence. It is in my view more straightforward for the Applicants, if so advised, to start fresh proceedings.

For those reasons, which I believe essentially accord with those of the Divisional Court, I would dismiss this appeal.

I too would dismiss this appeal, for the reasons given by Underhill LJ.

I agree.