(with whom Lord Wilson and Lord Hughes agree)
This is another round in the series of important points of law which arise as preliminary issues in actions brought by people who claim to have been wrongfully detained or mistreated by British or American troops in the course of the conflicts in Iraq and Afghanistan. The litigation is being expertly managed by Leggatt J, but so far there have been no trials and so the points of law are being decided mainly on the basis of assumed facts. To summarise the issues which have so far been heard in this court:
Mr Rahmatullah is a Pakistani national who was captured by the British forces in Iraq on 28 February 2004, transported to a United States detention facility that same day, and transferred by the US to a detention facility in Afghanistan on 29 March 2004, where he remained until his release on 15 May 2014. He is suing the Ministry of Defence and the Foreign and Commonwealth Office, first in respect of the United Kingdom's own treatment of him, and second in respect of the UK's alleged complicity in his detention and treatment by the United States. In relation to the second aspect of his claim, the UK government has raised the defences of state immunity and foreign act of state. The arguments relating to these defences were heard together with the claims of Mr Belhaj and his wife, Mrs Boudchar, against Mr Jack Straw and a number of UK officials and agencies, for alleged complicity in their rendition by Malaysian, Thai and US officials to Libya and their detention and torture there, where the same defences were raised. Judgment is given today: see Belhaj (Respondents) v Straw (Appellants) and Rahmatullah (No 1) (Respondent) v Ministry of Defence (Appellants)  UKSC 3. In relation to the first aspect of his claim, which is based on both the Iraqi law of tort and the UK Human Rights Act 1998, the UK Government has raised the doctrine of Crown act of state in relation to the tort claim, and this judgment is concerned with that doctrine.
A large number of Iraqi citizens have made claims similar to that of Mr Rahmatullah in respect of their detention and treatment by UK troops and transfer to the US authorities at various times during the UK's military presence in Iraq. In relation to many of these claims, the UK Government raised the defence that they were statute-barred by the Iraqi law of limitation. Judgment on that issue was given on 12 May 2016: see Iraqi Civilians v Ministry of Defence  UKSC 25; 1 WLR 2001. The UK Government has also raised the doctrine of Crown act of state. Three of the claimants, known as XYZ, ZMS and HTF, have been chosen as representative for the purpose of deciding this issue.
Mr Serdar Mohammed is an Afghan national who was captured in a planned International Security Assistance Force (ISAF) operation targeting a senior Taliban commander on 7 April 2010. He was detained by British troops until 25 July 2010 when he was transferred into Afghan custody. He was subsequently tried, convicted and sentenced to ten years' imprisonment for offences relating to the insurgency in Afghanistan. He too claims that his detention was unlawful both under the Afghan law of tort and the Human Rights Act 1998. In relation to his Human Rights Act claim, the UK Government argues that his detention was not in breach of article 5 of the European Convention on Human Rights, because article 5 has to be modified to take account of detention during armed conflict which is permitted, either under resolutions of the United Nations Security Council or under International Humanitarian Law. The argument about article 5 was heard together with a similar argument raised against the Ministry of Defence by Mr Al Waheed, an Iraqi national detained in the course of the conflict in Iraq. Judgment is given today: see Abd Ali Hameed Al-Waheed (Appellant) v Ministry of Defence (Respondent) and Serdar Mohammed (Respondent) v Ministry of Defence (Appellant)  UKSC 2. In relation to Mr Mohammed's tort claim, the UK Government has raised the same doctrine of Crown act of state as is raised in Mr Rahmatullah's and the Iraqi civilians' cases, and with which this judgment is concerned.
For completeness, there should also be mentioned the claims brought by the "PIL three" under both the Human Rights Act 1998 and UK public law in respect of their detention in Afghanistan. They bring no claim under the Afghan law of tort and so the question of Crown act of state does not arise in their cases.
The issues relating to Crown act of state
So what is this doctrine of Crown act of state? An act of state has been very widely defined, for example, by ECS Wade (in "Act of State in English Law: Its Relations with International Law" (1934) 15 British Yearbook of International Law 98, at p 103):
Act of state means an act of the Executive as a matter of policy performed in the course of its relations with another state, including its relations with the subjects of that state, unless they are temporarily within the allegiance of the Crown.
That definition is cited, not entirely approvingly, in the leading case of Nissan v Attorney General  AC 179, at 212 (Lord Reid), 218 (Lord Morris) and 231 (Lord Wilberforce). It is also cited in the footnotes to the current issue of Halsbury's Laws of England, with the comment that act of state is "not a term of art". Halsbury refines the definition slightly:
An act of state is a prerogative act of policy in the field of international affairs performed by the Crown in the course of its relationship with another state or its subjects.
No doubt it is a necessary component of the doctrine that the act in question falls within some such definition. But, as Lord Wilberforce pointed out in Nissan, that does not tell us what the doctrine is, or to what rule or rules of law it gives rise.
The doctrine is very rarely pleaded and so recent authority is scant. In this century, it has been raised in the context of the conflicts in Iraq and Afghanistan, first in Al-Jedda v Secretary of State for Defence (No 2)  EWCA Civ 758;  QB 773, which was decided on other grounds, and now in the current cases. In the 20th century, there are only two reported House of Lords cases in which it was raised, Johnstone v Pedlar  2 AC 262 and Nissan v Attorney General, above, and in neither of them was it successful, although it did succeed in a number of Indian appeals before the Judicial Committee of the Privy Council. We have therefore to go back to the 19th century and beyond to discover its origins and rationale.
The starting point is that English law "does not recognise that there is an indefinite class of acts concerning matters of high policy or public security which may be left to the uncontrolled discretion of the Government and which are outside the jurisdiction of the courts" (H Street, Governmental Liability, A Comparative Study, Oxford University Press, 1953, p 50). That there is no general defence of state necessity to a claim of wrongdoing by state officials was firmly established in the landmark case of Entick v Carrington (1765) 19 St Tr 1029, following on from Leach v Money (1765) 19 St Tr 1001and Wilkes v Wood (1763) 19 St Tr 1029,98 ER 489. This principle was reiterated by Viscount Finlay in Johnstone v Pedlar, at 271:
It is the settled law of this country, applicable as much to Ireland as to England, that if a wrongful act has been committed against the person or the property of any person the wrongdoer cannot set up as a defence that the act was done by the command of the Crown. The Crown can do no wrong, and the Sovereign cannot be sued in tort, but the person who did the act is liable in damages, as any private person would be.
It was thus no defence to a claim for the return of money and a cheque, taken by the police from a person arrested in Ireland for illegal drilling in 1918, that the Chief Secretary for the Treasury had signed a certificate formally to "ratify, adopt and confirm the said seizure and detention of the said cash and cheque" as an act of state for the defence of the realm and for the prevention of crime. It made no difference that the person arrested was a US citizen: the United Kingdom was not at war with the United States. As a friendly alien resident here he was "a subject by local allegiance with a subject's rights and obligations", per Viscount Cave at 276.
However, there was an exception, which Viscount Finlay stated in very wide terms, at 271:
This rule of law has, however, been held subject to qualification in the case of acts committed abroad against a foreigner. If an action be brought in the British Courts in such a case it is open to the defendant to plead that the act was done by the orders of the British Government, or that after it had been committed it was adopted by the British Government. In any such case the act is regarded as an act of state of which a municipal court cannot take cognizance. The foreigner who has sustained injury must seek redress against the British Government through his own Government by diplomatic or other means.
The question for this court is whether there is indeed a qualification such as that expressed by Viscount Finlay and, if so, how far that qualification goes. It is not contended on behalf of the Government that it is so broad as to cover any act committed against a foreigner abroad which is authorised or ratified by the Crown. The contention of the Government is that the doctrine of Crown act of state covers two distinct principles. The first is a principle of non-justiciability: this is that certain acts committed by a sovereign state are, by their very nature, not susceptible to adjudication in the courts. The obvious examples (given by Lord Pearson in Nissan v Attorney General, at 237)are "making war and peace, making treaties with foreign sovereigns, annexations and cessions of territory". The second is a defence to an action in tort: that a foreigner cannot sue the Government, or its servants or agents, in the courts of this country in respect of certain acts committed abroad pursuant to deliberate UK policy in the conduct of its foreign affairs.
The respondent claimants, on the other hand, argue that there is only the first rule, a narrow rule of non-justiciability whereby certain acts of government in the conduct of foreign affairs are by their very nature not justiciable in the courts. The decision to go to war in Iraq, and to remain there after the cessation of hostilities between the allied invaders and the state of Iraq in order to bring about internal peace and stability, and the decision to contribute to the International Security Assistance Force in Afghanistan, were of that nature. But the decision to detain these particular individuals in the course of those operations was of a completely different character. The question of whether the detention of an individual is lawful, under whichever system of law is applicable, is "quintessentially a matter for a court" (per Leggatt J, in Mohammed v Ministry of Defence  EWHC 1369 (QB), para 381). None of the reasons that might make it non-justiciable (helpfully summarised at para 377, referring to Buttes Gas and Oil Co v Hammer (No 3)  AC 888 and R (Al-Haq) v Secretary of State for Foreign and Commonwealth Affairs  EWHC 1910 (Admin); see also Shergill v Khaira  UKSC 33; AC 359) apply: there is no "absence of judicial or manageable standards" by which to judge it; the courts have the relevant expertise; and this is not a matter of high policy, which is constitutionally in the hands of Government Ministers who are accountable to Parliament and not in the hands of the courts.
For those reasons, in Mr Mohammed's case, Leggatt J held that the non-justiciability rule did not apply. However, he went on to hold that the tort defence did apply, at para 395:
It is not the business of the English courts to enforce against the UK state rights of foreign nationals arising under Afghan law for acts done on the authority of the UK government abroad, where to do so would undercut the policy of the executive arm of the UK state in conducting foreign military operations.
He went on to emphasise how narrow this second rule was: it applied only to executive acts done abroad pursuant to deliberate UK foreign policy and might well be confined to acts involving the use of military force (para 397). It was "analogous to the conflict of laws rule that English courts will not enforce a right arising under the law of a foreign country if to do so would be contrary to English public policy, and to the rule that English courts will not enforce the penal, revenue and public law of a foreign state" (para 396).
He returned to this question in the case of Mr Rahmatullahand the Iraqi civilians and rejected the argument that there was no good authority for such a rule; he also rejected the argumentsthat, if there were, it had been abolished by the Crown Proceedings Act 1947 or was incompatible with the right to a fair hearing under article 6 of the European Convention on Human Rights: Rahmatullah v Ministry of Defence and Foreign and Commonwealth Office; R (Rahmatullah and Ali) v Secretary of State for Defence and Secretary of State for Foreign and Commonwealth Affairs  EWHC 3846 (QB), paras 179-223.
The cases were taken together in the Court of Appeal: Mohammed (Serdar) v Ministry of Defence, Qasim v Secretary of State for Defence, Rahmatullah v Ministry of Defence, Iraqi Civilians v Ministry of Defence  EWCA Civ 843;  2 WLR 247 . That court also accepted that there was a tort defence as well as a non-justiciability rule. But it was "an exception to the general principle that proceedings may be brought in this country founded on a tort which is actionable under the law of a foreign country where the law of that country is the applicable law". They agreed with the judge that the rationale for the exception was to be found in domestic public policy (para 349). Accordingly, it was necessary to identify, in each case, the public policy interests which justified denying access to the courts in this way (para 352). In Mr Mohammed's case there were no compelling considerations of public policy which should prevent reliance on Afghan law as the basis of his tort claims (para 364). In the other cases, the relevant facts and evidence had not yet been pleaded. The court held that claims would be barred by the doctrine of act of state "only if the defendant is able to establish that there are compelling grounds of public policy to refuse to give effect to Iraqi law" (para 377).
On the Government's appeal against those decisions, therefore, the parties have defined the issues thus:
Is the doctrine of Crown act of state limited to a non-justiciability rule or does it also encompass a tort defence?
If it does encompass a tort defence, what is its scope?
In particular, is the test to be applied, by analogy with section 14(3)(a) of the Private International Law (Miscellaneous Provisions) Act 1995, whether there are compelling grounds of public policy to refuse to give effect to the local law of tort?
Was the tort defence extinguished by the Crown Proceedings Act 1947?
Is the tort defence incompatible with article 6 of the European Convention on Human Rights?
It may well be, however, that issues (i), (ii) and (iii) can be reduced to a single issue, that is, the circumstances in which a claim is not cognisable in the courts of England and Wales because it relates to a "Crown act of state". The following issues are not now in dispute:
if the doctrine is limited to a narrow non-justiciability rule, of the sort identified before Leggatt J and the Court of Appeal, it is not applicable in these cases, for the reasons given by the judge;
the doctrine is not a defence to claims made under the Human Rights Act 1998 and so those will continue in any event.
For the purpose of the Crown act of state issue, it is to be assumed that the claimants' detentions were in fact contrary to the Afghan or Iraqi laws of tort, although that too is in issue in the case of Serdar Mohammed.
In order to discover the nature and content of the doctrine, it will be necessary to look at some old authorities which, although culminating in Nissan v Attorney General, were decided against a legal landscape which was very different from the legal landscape of today. The conduct of foreign affairs, making treaties, making peace and war, conquering or annexing territories, are all aspects of the Royal prerogative. Until the decision of the House of Lords in Council of the Civil Service Unions v Minister for the Civil Service  AC 374 (the GCHQ case), the general position was that the courts would review whether what had been done fell within the scope of the Royal prerogative but would not review how that prerogative had been exercised. After that case, the exercise of executive power might be excluded from the scope of judicial review, not because of its source, whether statute or the prerogative, but because of its subject matter: hence the need to distinguish between certain acts of high policy, which by their very nature are not subject to judicial review, and other actions taken in pursuance of that policy, which are.
Second, the old cases were decided against the backdrop of the principle that "the King can do no wrong". The King could not be sued in his own courts. The officials who carried out his policies could be sued for their unlawful actions, and the practice developed of nominating an official as a defendant to claims in tort. But the courts had to grapple with the circumstances in which the King's prior authority or subsequent ratification might import the doctrine that the King could do no wrong and thus afford a defence to such a claim. We have already seen how the courts distinguished between acts done to foreigners abroad and acts done to citizens or residents here. The Crown Proceedings Act 1947 abolished the general immunity of the Crown from liability in tort and enabled litigants to sue Government departments such as the defendants in these cases.
Thirdly, it may be worth bearing in mind that until the Private International Law (Miscellaneous Provisions) Act 1995, a tort committed abroad could only be the subject of a claim in the English courts if the conduct complained of was tortious, both by the law of the place where it took place and by the law of this country. This rule was abolished by the 1995 Act, which established the general rule that the applicable law in an action in tort is the law of the country in which the events took place. Hence it is now accepted that the tort claims have to be determined according to the law of Afghanistan or Iraq respectively, subject to the doctrine of Crown act of state if applicable, while the human rights claims have to be determined according to the Human Rights Act 1998.
Finally, as already noted, the term "act of state" is also used in a completely different context, that of whether the courts of this country will adjudicate upon the acts of a foreign legislature or executive. Sometimes, of course, both doctrines may arise in the same case (they are both, for example, discussed by Lord Wilberforce in the leading case of Buttes Gas and Oil Co v Hammer (No 3)  AC 888). And insofar as the doctrine of foreign act of state covers governmental acts outside the territory of the state concerned, there may be some similarities, as Lord Mance has shown. But Crown act of state was not raised as a defence in Belhaj v Straw and foreign act of state is not the subject-matter of this judgment.
Does the Crown act of state doctrine encompass two rules?
The clearest judicial statement that the doctrine does encompass two rules is that of Lord Wilberforce in Nissan's case, at 231:
The first rule is one which provides a defendant, normally a servant of the Crown, with a defence to an act otherwise tortious or criminal, committed abroad, provided that the act was authorised or subsequently ratified by the Crown. It is established that this defence may be pleaded against an alien, if done abroad, but not against a friendly alien if the act was done in Her Majesty's Dominions. It is supported in its positive aspect by the well-known case of Buron v Denman (1848) 2 Exch 167 and in its negative aspect by Johnstone v Pedlar  2 AC 262.
The second rule is one of justiciability: it prevents British municipal courts from taking cognisance of certain acts. The class of acts so protected has not been accurately defined: one formulation is 'those acts of the Crown which are done under the prerogative in the sphere of foreign affairs' (Wade and Phillips's Constitutional Law, 7th ed (1956), p 263). As regards such acts it is certainly the law that the injured person, if an alien, cannot sue in a British court and can only have resort to diplomatic protest. How far this rule goes and how far it prevents resort to the courts by British subjects is not a matter on which clear authority exists. From the terms of the pleading it appears that it is this aspect of the rule upon which the Crown seeks to rely.
It would appear, however, that the case was only concerned with the second, the non-justiciability rule. The United Kingdom had made a treaty with the government of Cyprus (then an independent country) to provide troops in order to restore peace between the Greek and Turkish Cypriot communities. The claimant, a British subject, ran a successful luxury hotel near Nicosia. The British troops took over the hotel as their headquarters. No claim in tort was made in respect of the occupation of the hotel, to which the claimant had apparently consented. There was a tort claim in respect of damage to furniture and other chattels, but it was accepted that this should go to trial. The main claim was that there was a contractual right to compensation, which was disputed. But if there was such a contract, act of state could not be a defence. The disputed claim was one in restitution, for compensation for the use and occupation of the hotel. To this the government pleaded that the actions of the British forces in Cyprus were acts of state of Her Majesty on the territory of an independent sovereign performed in pursuance of an agreement with the Cyprus government and as such not cognisable by the court. The House of Lords unanimously rejected this defence, Lord Reid on the ground that it could not be pleaded against a British subject, but the other members of the House on the ground that the occupation of the hotel did not have the character of an act of state. A distinction should be drawn between the making of the treaty with Cyprus, which was an act of state, and the actions of the troops complained about, which were not so closely connected with the treaty as to amount to an act of state. Lord Morris regarded Professor Wade's definition of an act of state (para 2 above) as "helpful" but went on to explain why it did not apply (at 218):
I do not think that such actions as securing food or shelter in peace time for troops situate abroad are to be regarded as acts of the executive performed inthe course of relations with another state within the conception of the above definition. But, even if they were, I would be surprised if the contention were advanced that it was 'a matter of policy' on the part of the executive to take food or shelter and not to make payment.
It is worth noting that the House rejected the contention that the UK was acting as agent for the government of Cyprus. But, as Lord Wilberforce pointed out at p 230, if it had been, then the doctrine of foreign act of state might well have applied: they would have been acts attributable to a foreign government in its own territory. As Lord Mance has shown in his judgment in Belhaj, that doctrine may extend to a state's appropriation of property within its own territory even if this is illegal by the law of that state.
The question for us is not whether the type of rule, with which Nissan was concerned, exists: there is no reason to doubt that it does. The question is whether a different type of rule, affording a tort defence even though the subject matter is entirely suitable for adjudication by a court, also exists. Although its existence was acknowledged in Nissan, and indeed Lord Morris stated that it was "so recognised that it cannot now be overthrown" (at 220), the foundations upon which it is built are very shaky.
The source of such a rule is the direction to the jury by Parke B in the well-known case of Buron v Denman (1848) 2 Exch 167, 154 ER 450 (the background is explained by C Mitchell and L Turano in C Mitchell and P Mitchell (eds), Landmark Cases in the Law of Tort (Hart, 2010) and also by A Perreau-Saussine, "British Acts of State in English Courts"  British Yearbook of International Law 176). Briefly, in 1835, as part of its campaign to suppress the slave trade, Britain made a treaty with Spain, which allowed British ships to stop and search Spanish vessels on the high seas if they were suspected of trading in slaves. Commander Denman's patrol was looking for slaving ships at the mouth of the Gallinas river in West Africa. He was asked by the Governor of Sierra Leone to liberate two British subjects who were being held as slaves in one of the "barracoons" (slave pens) on islands at the mouth of the river. Denman and his crew landed on the islands, chased away the Spanish slavers, and liberated the slaves they were trying to take with them. Denman then made a treaty with the local chiefs, outlawing slavery in their lands, handed over the slavers' trade goods in return, took possession of the barracoons, liberated the slaves and burned the barracoons down. He carried several hundred former slaves back with him to Sierra Leone. He also rescued some of the slavers, who were fleeing local retribution, one of whom was SeñorBuron.
All of this was greeted with great jubilation when the news reached England, the Colonial Secretary and the Foreign Secretary exchanged letters praising Denman's actions, Parliament voted a bounty to him and his crew, and he was promoted to Captain. SeñorBuron, however, brought an action in trespass against him, claiming damages for the loss of his chattels, including the slaves. Parke B directed the jury that, slave-owning not being shown to be against the law in the Gallinas, Denman's actions could amount to a trespass; but their subsequent ratification by the Government turned them into an act of state, for which he could not be sued (although the judge left open whether SeñorBuron might able to proceed against the Crown by petition of right or whether he could only pursue a remedy by diplomatic means, because this was irrelevant to the action he was trying).
It appears that the only act of state case cited in argument was Elphinstone v Bedreechund (1830) 1 Kn 316, 12 ER 340. But that was essentially a non-justiciability case: the issue was whether the Supreme Court of Bombay had jurisdiction to hear a claim for damages for the seizure of property of the governor of a fortress conquered in the course of military hostilities. There is a whole series of cases, not all of them easy to reconcile (helpfully discussed by Perreau-Saussine, loc cit ), concerning the appropriation of property in the course of annexing territory in India (and on occasions in Africa) supporting the proposition that "the transactions of independent states between each other are governed by other laws than those which municipal courts administer: such courts have neither the means of deciding what is right, nor the power of enforcing any decision which they may make" (Secretary of State in Council of India v Kamachee Boye Sahaba (1859) 7 Moore Ind App 476, at 529, 19 ER 388, at 407; Cook v Sprigg  AC 572, at 578). The leading case is Kamachee, where the East India Company, as agent for the British Crown, had seized the Raj of Tanjore, and the whole of the property of the Rajah, who had died without male issue, under Treaties authorising the annexation of the Raj. The Rajah's widow sued for the return of his private property. She succeeded in the Supreme Court of Madras but failed before the Judicial Committee of the Privy Council. No distinction could be drawn between private and public property for the purpose of such an act of state. Most of the discussion relates to the character of the act, but there is a brief reference to Buron v Denman.
The subject matter of Buron v Denman, however,was something different. It was not a transaction between states. Denman's actions were not carried out on the high seas in accordance with the treaty with Spain. Britain was not at war with another state or conquering territory (although it could be said to be conducting a "war on the slave trade"). His actions were by ordinary standards, both of the local law and of English law, unlawful. The case has therefore been treated as establishing a defence to an action in tort over which the ordinary courts would otherwise have jurisdiction. It was so regarded by the House of Lords in Johnstone v Pedlar ,in which the scope of such a defence was directly in point, and also in Nissan v Attorney General, where it was not.
No doubt it was so regarded, in part at least, because this was how it was regarded by some eminent academic authorities. For example, Lord McNair, in International Law Opinions (1956) in a chapter dealing with "The Position of the United Kingdom Government, its Servants and Agents, as Defendants instituted in Actions in British Courts", distinguished between
"the defence called act of state available to certain defendants in British courts" and
"the rule which entirely excludes from British courts certain areas of British governmental action in the realm of foreign affairs".
As to (a):
Act of state has been defined in a standard text book as follows: 'The plea, act of state, can be raised as a defence to an act, otherwise tortious or criminal, committed abroad by a servant of the Crown against a subject of a foreign State or his property, provided that the act was authorised or subsequently ratified by the Crown'. (citing Wade and Phillips, Constitutional Law (4th ed (1950), 193-196))
To this very wide definition McNair added the slight qualification:
Its scope of operation is the whole field of governmental or official activity in relation to the Crown's dealings with foreign states.
The only authorities cited are Buron v Denman, Johnstone v Pedlar, where the argument that it could be relied upon in claims brought by friendly aliens was rejected, and Commercial and Estates Co of Egypt v Board of Trade  1 KB 271, at 290 and 297, where the argument that it could be relied upon in respect of actions within the realm was rejected; however, Scrutton LJ noted that the owners of the cargo of timber on a British ship which had been requisitioned abroad during the first world war and brought to this country "did not rely on any dealings with the cargo outside the realm for the probable reason that on the authority of such cases as Buron v Denman a claim by a foreigner for such acts would be successfully met by the defence that the interference was an act of state".
The doctrine was also relied upon in Al-Jedda v Secretary of State for Defence (No 2)  EWCA Civ 758;  QB 773. This was another tort claim arising from detention by British forces in Iraq, this time of a person with dual British and Iraqi nationality. As here, the applicable law was the law of Iraq and Underhill J held that the detention was lawful under the law of Iraq. But he also held that the defence of act of state would have been available. In the Court of Appeal, Lord Dyson JSC and Elias LJ agreed with the judge that the detention was not unlawful under Iraqi law. Lord Dyson declined to deal with the act of state issue, on the ground that it did not arise and raised points of very considerable difficulty, on which they had not heard full argument (para 127). Elias LJ did discuss it (paras 192-226) and his "tentative" view was that it would not be an answer to the claim (para 193). The act did fall into the category of act of state, in that it would have removed the jurisdiction of the courts to question the detention of a foreign subject (para 195); but "the courts would be failing in their constitutional duty if they were to leave the executive with unfettered powers to intern British citizens merely because the act of internment occurred abroad" (para 216); however, he did float an intermediate possibility, that even if this were to be an act of state, and thus not to give rise to liability for damages in tort, it would be amenable to judicial review on conventional principles. Arden LJ, on the other hand, agreed with Underhill J that act of state was a defence; but this was on the basis of the House of Lords' decision in R (Al-Jedda) v Secretary of State for Defence  UKHL 58;  AC 332, that the UK was entitled and bound under its obligations under article 103 of the United Nations Charter and the applicable UN Security Council Resolution to intern people where this was necessary for the internal security of Iraq(para 108); thus, not only was the decision to join the Multi-National Force an act of state, but acts required to be done pursuant to that decision were also acts of state; the fact that the claimant was a British national made no difference.
Clearly, therefore, Al-Jedda (No 2) is of very little help in resolving the issue between the parties here. Both Underhill J and Arden LJ appeared to be treating the case as non-justiciable, because internment was required under international law; Elias LJ appears to have agreed with them, but considered that this afforded no defence to the internment of a British citizen. In the cases before us, both Leggatt J, at  EWHC 3846, para 197, and the Court of Appeal, at  2 WLR 247;  EWCA Civ 843, para 330, thought that the issue in Al Jedda (No 2) was better analysed under the tort defence than under the non-justiciability rule.
In the light of these shaky foundations, it is scarcely surprising that the respondent claimants argue that the tort rule does not, in fact, exist. The only rule is that certain decisions of high policy in the conduct of foreign relations are non-justiciable. The arguments to the contrary are of two kinds: first, that the existence of a wider rule is long-established both in the case law and in academic commentaries and texts, as already discussed; and second, that there are good reasons for it, certainly in the context of military operations abroad. As I understand it, we are not asked to consider whether it exists outside that context.
Sir James Fitzjames Stephen, in his History of the Criminal Law (1883), Vol II, argued that where an injurious act done to a foreigner "is an act of open war, duly proclaimed, there can be no doubt at all that it does not amount to a crime .... the very essence of war is that it is a state of things in which each party does the other all the harm they possibly can". The same should apply to "acts which are in their nature warlike done in time of peace": "I think that if such acts are done by public authority, or, having been done, are ratified by public authority, they fall outside the sphere of the criminal law" (pp 62-63). He could cite no criminal law authority but relied upon Buron v Denman and Secretary of State for India v Kamachee Boye Sahaba. Of course, in those days, the criminal law was even more territorially limited than it is today and so only killings by British subjects abroad would be within the jurisdiction of the English courts. But the point is, if act of state is a defence to the use of lethal force in the conduct of military operations abroad, it must also be a defence to the capture and detention of persons on imperative grounds of security in the conduct of such operations. It makes no sense to permit killing but not capture and detention, the military then being left with the invidious choice between killing the enemy or letting him go.
There are conceptual advantages in confining the doctrine to a non-justiciability rule; but in doing so, a rather broader concept of non-justiciability would be required than that which was espoused in the courts below. It would have to encompass aspects of the conduct of military operations abroad as well as the high policy decision to engage in them, and perhaps also some other aspects of the conduct of foreign relations, even though their subject matter was entirely suitable for determination by the court. It is necessary that the courts continue to recognise that there are some acts of a governmental nature, committed abroad, upon which the courts of England and Wales will not pass judgment. They may, of course, have to hear evidence and find facts in order to determine whether the acts in question fall into that category. It is also necessary to confine that category within very narrow bounds. Contrary to the impression given in some accounts (for example, J Collier, "Act of state as a Defence against a British Subject" (1968) 26 CLJ 102) it cannot give carte blanche to the authorities to authorise or ratify any class of tortious acts committed abroad in the conduct of the foreign relations of the state.
Is this aspect of the doctrine one of public policy?
The approach of the Court of Appeal has very real attractions. It is consistent with the policy of the Private International Law (Miscellaneous Provisions) Act 1995, that where the English courts have jurisdiction over a tort committed abroad, the applicable law is the law of the state where the conduct took place, unless to apply that law "would conflict with principles of public policy" (section 14(3)(a)). It would enable a case-by-case approach depending upon a range of policy factors, such as those identified by the Court of Appeal. There is, however, no hint of such an approach in the cases concerned with act of state. In essence, public policy may be the reason why the courts of this country will apply the domestic doctrine of act of state rather than the tort law of the state where the events took place, but it cannot tell us what the content of that defence is to be. It is not enough to say that it is not for the courts of this country to enforce the tort laws of another state, because that is exactly what the 1995 Act expects us to do. The question is in what circumstances we should decline to do so.
The Court of Appeal, when considering Mr Mohammed's case, concluded that there was no authority to detain him, either under the regime established by the United Nations, or under local Afghan law, and the UK government had deliberately decided to apply a policy outside both of these without promoting UK legislation to permit this. The public interest that "not doing harm to the nation by precluding HM Armed Forces from detaining a commander in the Taliban for more than 96 hours because it appeared that questioning him would provide significant new intelligence vital for force protection purposes" was not sufficiently compelling to outweigh the public interest in the protection of liberty. In effect, therefore, military necessity, however compelling, provided no defence. If this aspect of the doctrine is to have any content at all, this cannot be right.
What is its scope?
It would be unwise for this court to attempt a definitive statement of the circumstances in which this aspect of the doctrine might apply. The question is whether it applies in the circumstances of Mr Mohammed's case, some of which have been explored in pleadings and evidence, and how it might apply in the circumstances of Mr Rahmatullah and the Iraqi civilians' cases, which have not yet been explored in pleadings and evidence. For the reasons already given, it cannot apply to all torts committed against foreigners abroad just because they have been authorised or ratified by the British Government. It can only apply to acts which are by their nature sovereign acts, acts which are inherently governmental, committed in the conduct of the foreign relations of the Crown. The Government accepts that it cannot apply to acts of torture, even supposing that the Government of the United Kingdom would ever authorise or ratify such acts. The Government also accepts that it cannot apply to the maltreatment of prisoners or detainees, such as happened in Baha Moussa's case. Bearing in mind that this is a doctrine of the law of the United Kingdom, I would prefer to regard this as an acknowledgement that such acts are not inherently governmental, rather than creating exceptions to a general rule. The Government of the United Kingdom can achieve its foreign policy aims by other means. Nor would it generally apply to the expropriation of property, for which compensation can always be paid, but there could be circumstances in which the expropriation, or more probably the destruction, of property, for example in the course of battle, was indeed a governmental act.
We are left with a very narrow class of acts: in their nature sovereign acts – the sorts of thing that governments properly do; committed abroad; in the conduct of the foreign policy of the state; so closely connected to that policy to be necessary in pursuing it; and at least extending to the conduct of military operations which are themselves lawful in international law (which is not the same as saying that the acts themselves are necessarily authorised in international law). For the purpose of these cases, we do not need to go further and inquire whether there are other circumstances, not limited to the conduct of military operations which are themselves lawful in international law, in which the defence might arise. Buron v Denman was at the borderline. The slaves were freed in the conduct of military operations pursuant to British foreign policy. Commander Denman was clever enough to negotiate a treaty with the local chiefs outlawing slavery before he freed most of them and burned down the barracoons but that did not necessarily render his acts lawful by local law. Nor do we need to decide whether the doctrine can ever be pleaded against British citizens: it was freely acknowledged in the courts below that there are arguments either way. They do not arise in these cases.
The Crown Proceedings Act 1947
The respondents argue that the defence was abolished by section 2(1) of the Crown Proceedings Act 1947. This provides:
Both Leggatt J and the Court of Appeal held that the 1947 Act had not had the effect of abrogating the defence of act of state. Section 2(1)(a) made the Crown vicariously liable for the torts of its servants or agents. The proviso makes it clear that this does not apply where the act or omission would not have given rise to a cause of action in tort against the servant or agent. The servant or agent could claim the defence of Crown act of state before the Act and the effect is, now, that the Crown can do so too. That is certainly how it was understood at the time, for example, by Professor Glanville Williams, in Crown Proceedings (Stevens, 1948, p 44), although he also suggested that the proviso was an unnecessary "bludgeon" as the so-called defence meant that there was no liability in any event.
The claimants argue that the purpose of the 1947 Act was to put the Crown in the same position as any other litigant in civil proceedings. Other litigants did not enjoy the benefit of the defence of Crown act of state. It would be illogical if a provision intended to put the Crown on the same footing as anyone else had a proviso having precisely the opposite effect. It is also odd to do so by a proviso aimed at the agent rather than the Crown. Where the Act intended to create immunities or restrictions, or preserve existing common law rules, or make savings, it did so clearly and expressly; an example is section 2(5), which gives the Crown immunity from liability for anything done by any person while discharging responsibilities of a judicial nature.
There is, however, nothing odd about preserving the previous law by means of a proviso aimed at the agent. The Act imposed vicarious liability for the acts of its servants or agents upon the Crown. It is natural, therefore, to make it clear that the Act is not making any difference to the previous law relating to the liability of that servant or agent, even if the previous law in question is one which applies only to Crown servants or agents. It may be that the proviso was unnecessary but there is no reason to doubt that the previous law of Crown act of state, whatever it was, was left intact.
The respondents accept that if, as held by the Court of Appeal, Crown act of state is only a defence where there are overriding reasons of public policy not to apply the tort law of a foreign state, this would be a proportionate interference with the right of access to a court and thus compatible with article 6 of the European Convention on Human Rights. However, they argue that any wider defence would be an unjustifiable impediment to that right of access. This depends upon two questions:
first, whether the defence is an aspect of the substantive law or whether it is a procedural restriction on the right to go to court to vindicate a right;
secondly, if it is merely a procedural bar, whether it is justified as a proportionate means of pursuing a legitimate aim.
The respondents argue that the defence is not an aspect of a substantive right, akin to the rule of Italian law considered by the Grand Chamber of the European Court of Human Rights in Markovic v Italy 44 EHRR 52. Claims had been brought in an Italian court against the Prime Minister, Ministry of Defence and Commander of NATO forces in southern Europe in respect of deaths caused by military action in the former Yugoslavia in 1999. The Cour de Cassation had held that the Italian court had no jurisdiction, under a rule very like the non-justiciability aspect of our own Crown act of state doctrine, that certain "acts of government", including the conduct of hostilities, did not give rise to civil liability. The Grand Chamber held that the claimants had not been deprived of access to a court: their claims had been fairly examined in the Italian courts in the light of the applicable domestic legal principles. Those principles "marked out the bounds of the law of tort" so that the inability to sue was not the result of an immunity but of the principles governing the substantive right of action. By contrast, argue the respondents, the Crown act of state defence is a procedural bar, which prevents the United Kingdom courts from enforcing rights and liabilities in tort which would otherwise be justiciable. This means, they argue, that the government must justify it and this they cannot do. It cannot be justified on the basis that it pursues the legitimate aim of ensuring that the government and the courts speak with "one voice" on matters of foreign policy. As the Court of Appeal pointed out (para 372), they are not required to do so in public law claims or in claims under the Human Rights Act. Leggatt J held that "it serves the legitimate aim of protecting the interests of the nation abroad, in particular where military action is considered necessary by the executive in the national interest" (Rahmatullah, para 217) and was proportionate to that aim (para 218). The respondents accept that it might be justified on the basis that it pursued the legitimate aim of enabling the legality of the Government's conduct of foreign affairs, and in particular military operations abroad, to be determined by international law rather than the law of the place where those operations took place. Thus it enables the Government to comply with its obligations in international law without having to concern itself with local domestic law. But the rule as contended for by the Government goes wider than is necessary to meet that aim and is thus disproportionate to it.
The Government, on the other hand, contends that article 6 is "not engaged". This is a Markovic case. Article 6 does not guarantee any particular content to the civil rights and liabilities protected by domestic law. It merely guarantees a right of access to the courts to have those rights and liabilities determined: Z v United Kingdom 34 EHRR 3, paras 87, 92. The claimants have the right to a fair hearing of whether the doctrine of Crown act of state, in either of its aspects, applies. But that doctrine defines the content of their rights. It means that there are certain actions which do not give rise to civil liability. It is not simply a procedural bar. Even if it were, they argue, it pursues the legitimate aim identified by Leggatt J and is proportionate to that aim for the reasons he gave: it applies only to acts done pursuant to deliberate United Kingdom foreign policy and only to claims arising under foreign law. It does not therefore apply to claims under the Human Rights Act.
In my view, this is clearly a rule of the substantive law rather than a procedural bar. It does not confer an immunity on a particular class of actors. It defines the circumstances in which there may be liability for a particular type of activity. The rules are the same whether that activity is governed by English law or by foreign law. As these cases show, the claimants do have access to a court to determine the scope of their rights. This court is concerned with the question of law as to how far their rights extend. In Mr Mohammed's case, the facts necessary to determine the extent of his rights have already been examined. In the other cases, the facts have yet to be fully pleaded and evidence filed. When they have, the court will have to decide what the facts are and whether they fall within the defence of Crown act of state as defined by this court. There has been no procedural bar to the claimants bringing these claims and fighting them vigorously through the courts. It is the substantive law which will determine whether, on the facts found, they succeed.
I would therefore allow the Government's appeal. I would substitute a declaration to the effect that, in proceedings in tort governed by foreign law, the Government may rely on the doctrine of Crown act of state to preclude the court passing judgment on the claim if the circumstances are such as stated in paras 36 and 37 above. It may well be that the declarations made by Leggatt J should be restored. In the case of Serdar Mohammed, he declared that, on the assumption that the facts relating to his arrest and detention pleaded by the Government were true, and without prejudice to his right to challenge the factual basis of his arrest and detention at any further trial, the defendants could rely on the doctrine of Crown act of state to preclude the enforcement of a claim under Afghan law. In the case of Yunus Rahmatullah, he declared that the claims in tort in relation to his arrest and detention by UK armed forces were barred by the doctrine of Crown act of state, "if the defendants established that his arrest and detention was authorised pursuant to lawful UK policy". He made a declaration in similar terms in the Iraqi Civilian Litigation. I would, however, invite further submissions as to the precise form of declaration which would be appropriate in each of these cases.
LORD MANCE: (with whom Lord Hughes agrees)
Crown act of state certainly presents terminological and conceptual difficulties. But I think it clear that the underlying principle is one of non-justiciability or (as I would prefer to say: see para 54 below) abstention or restraint. It creates unnecessary confusion to suggest that the principle has two branches, one non-justiciability, the other a defence based on Buron v Denman (1848) 2 Exch 167.
Lord Sumption suggests a dichotomy between two rules (para 79). But he ends with a proposition that, in the present context, the two rules merge into one (para 81). This is achieved by defining non-justiciability in the present context as going "to the existence or scope of legal rights" (para 80) and so as a defence (para 81). To my mind, this involves confusion. Lord Sumption seeks to support it in para 80 by suggesting that the case of R (Campaign for Nuclear Disarmament) v Prime Minister  EWHC 2777 (Admin) proceeded on the footing that the Divisional Court "had both jurisdiction and competence to determine whether a resolution of the United Nations Security Council authorised military operations against Iraq" but that it declined to do so because there were no relevant domestic law rights. This, however, in my opinion, misreads the Divisional Court's judgment. The essential ground of the decision in Campaign for Nuclear Disarmament was that the subject matter was non-justiciable: see per Simon Brown LJ at para 47(ii), Maurice Kay LJ at para 50, and Stephen Richards LJ at paras 59-60. Absence of domestic foothold was a separate and (as Maurice Kay and Stephen Richards LJJ make clear in these passages) lesser objection to the claim (see also Simon Brown LJ at paras 35-36).
The case of Campaign for Nuclear Disarmament therefore lends no support to a proposition that Crown act of state involves a defence. On the contrary, it places Crown act of state, involving foreign policy action including the deployment of armed forces, firmly within the domestic principle of non-justiciability or abstention, and this is so whether the Crown or its agent is being impleaded. Crown act of state is at the same time (as Lady Hale notes: paras 43-45) a principle of substantive law outside the scope of article 6 of the European Convention on Human Rights.
Crown act of state is in short based on the same underlying principle of abstention that can in some circumstances also apply to preclude adjudication of the third type of foreign act of state identified in my separate judgment on that topic handed down concurrently with this judgment in the present case and in Belhaj v Straw  UKSC 3. That is not to suggest that the principle of abstention applies with the same force or by reference to the same considerations in relation to the latter context. Because of the way in which the issues in these cases were identified and divided for determination, this was not an aspect on which the submissions before us focused. I will say some provisional words about it.
Both Crown act of state and the third type of foreign act of state are based on an underlying perception of the role of domestic courts. The constitutional relationship of a domestic court with its own State differs from its relationship with that of any foreign sovereign state. Crown act of state is reserved for situations of sovereign authority exercised overseas as a matter of state policy. In these circumstances a straight-forward principle of consistency directly underpins Crown act of state (as identified by Lord Sumption in para 87). In contrast, if and when the third type of foreign act of state applies, its underpinning is a more general conception of the role of a domestic court, and, more particularly, the incongruity of a domestic court adjudicating upon the conduct of a foreign sovereign state, even though the foreign state is neither directly or indirectly impleaded or affected in its rights. However, concern for the international relations of the domestic with the foreign state, and in that sense a concern that the domestic court's stance should not be out of line with that of its own state's, may probably in some cases play some part: see the discussion in paras 103 to 105 of my judgment in Belhaj v Straw .
This analysis is supported by what Lord Wilberforce said in Buttes Gas, 938A-C, quoted in para 42, as well as with my own observations in para 91, of my judgment in Belhaj v Straw. But to immunise the home state of the domestic court from action anywhere (or, in a Buttes Gas type case to refuse to adjudicate upon civil litigation between third parties), by reference to the conduct of a foreign state, is self-evidently an extreme step. This is no doubt why the principle of abstention recognised in Buttes Gas has rarely found application. These differences in underpinning and analysis between Crown act of state and foreign act of state mean, in my opinion, that it must be easier to establish that a domestic court should abstain from adjudicating on the basis of Crown act of state than on the basis of the third type of foreign act of state. The relationship is closer and the threshold of sensitivity lower in the case of the former than the latter.
It is necessary to grasp the considerations which may make a case non-justiciable. A consideration which in the past may have encouraged an overly narrow view is Lord Wilberforce's pithy references in Buttes Gas & Oil Co v Hammer  AC 888, 938B-C to an absence of "judicial or manageable standards" by which to judge the issues in that case, placing the court in a "judicial no-man's land". But it is clear that these references do not represent the definitional limit of non-justiciability in the present context. They represent – as Lord Sumption pointed out in Shergill v Khaira  UKSC 33;  AC 359, para 40 – only one of two reasons why the issue in Buttes Gas was "political" or "non-justiciable". The other was that the issue "trespassed on the proper province of the executive, as the organ of the state charged with the conduct of foreign relations", a consideration which takes one back to the discussion in the previous paragraphs.
Bearing in mind the potential for misunderstanding the concept of non-justiciability in the light of Lord Wilberforce's aphorisms in Buttes Gas, it may be preferable to refer instead, as I have done above, to a principle of abstention or restraint, whereby Crown decisions and/or activities of a certain nature in the conduct of foreign affairs are not open to question (or are "not cognisable") in domestic civil proceedings, at the instance of anyone injured thereby (except, perhaps, someone owing allegiance to the Crown – a point which can be left open on this appeal).
I have already indicated that I do not accept what was the primary case of the appellants, the Ministry of Defence and the Foreign and Commonwealth Office, namely that there is a dichotomy between two rules (non-justiciability and a tort defence). However, the appellants never committed themselves to bringing their case within either rule. On the contrary, they submitted in paras 101, 107 and 173(4) and (5) of their written case: that "each of the claims should be dismissed on the grounds that Crown act of state operates as a separate defence and/or a jurisdictional bar": that "the Court of Appeal mischaracterised the scope of the justiciability limb of the Crown act of state doctrine, and consequently concluded that it did not apply in the present case"; that "the Court of Appeal erred in holding that the non-justiciability rule only applies in circumstances where the issue is one which the court is constitutionally incompetent to determine"; and that "in any event, the tort defence and/or non-justiciability rule applies to each of the cases". In his oral submissions, Mr Eadie QC maintained this position, in submissions to the effect that "how you characterise the doctrine, whether as a defence or non-justiciability, does not matter; its criteria and effect are what matters". There is good sense in this. It is, nonetheless, helpful to identify the conceptual basis of Crown act of state. In my view, there is, as I have stated, a single doctrine based on non-justiciability or, in the terms which I prefer, judicial abstention or restraint.
I turn to examine further the core rationale behind cases where courts will abstain from adjudicating upon civil claims against the Crown or its servants. This is also discussed in some detail in the passages from Campaign for Nuclear Disarmament cited above. Since Council of Civil Service Unions v Minister for the Civil Service ("GCHQ")  AC 374, reinforced by R (Bancoult) v Secretary of State for Foreign and Commonwealth Affairs (No 2)  UKHL 61;  1 AC 453, the exercise of prerogative powers, including prerogative legislation in the form of an order in council, has not enjoyed any general immunity from judicial scrutiny. But the nature and subject-matter of the particular prerogative power being exercised may make it inappropriate for adjudication before a domestic court. Thus in GCHQ Lord Roskill said, at p 418, that
Prerogative powers such as those relating to the making of treaties, the defence of the realm, the prerogative of mercy, the grant of honours, the dissolution of Parliament and the appointment of ministers as well as others are not, I think, susceptible to judicial review because their nature and subject matter are such as not to be amenable to the judicial process.
In Shergill v Khaira, para 42, Lord Sumption referred to this category of case as "beyond the constitutional competence assigned to the courts under our conception of the separation of powers", and included within it "the non-justiciability of certain transactions of foreign states and of proceedings in Parliament". The thinking behind all these concepts is linked in Blackstone's Commentaries on the Laws of England (vol 1) pp 251 and 257-258. There, the non-justiciability of the royal prerogative of making war and peace or treaties is explained on the basis that the appropriate forum for its control is Parliament (including, in the last resort, as Blackstone notes, by impeachment). In the case of certain foreign activities of the British state, there is in my view an additional parallel aspect at the international level to their non-justiciability in domestic courts. That is that representations and redress in respect of activities involving foreign states and their citizens may be more appropriately pursued at a traditional state-to-state level, rather than by domestic litigation brought by individuals.
In any event, it is, as already shown, wrong to regard either Lord Roskill in GCHQ or the Supreme Court in Shergill v Khaira as basing this category of case on an absence of judicial or manageable standards or the presence of a judicial no-man's land. When there is an appropriate domestic foothold and the matter is otherwise justiciable, domestic courts are well able to adjudicate upon and give effect to international law (see the citations from Campaign for Nuclear Disarmament, above). Indeed, customary international law was long said to be automatically incorporated into domestic law: see the discussion in R (Keyu) v Secretary of State for Foreign and Commonwealth Affairs  3 WLR 1665, paras 117-122 and 144-151. As Lord Sumption also recognised in Shergill v Khaira, para 43, a private claim may require adjudication upon issues of international law, where there is a domestic foothold in the sense of a prima facie domestic law right under whatever may be the relevant law: see eg Republic of Ecuador v Occidental Exploration Production Co  EWCA Civ 1116;  QB 432 .
The Report of William Murray (later Lord Mansfield) and other Law Officers on the Rules of Admiralty Jurisdiction, etc in time of war dated 18 January 1753 and the decision in "The Rolla" (1807) 6 Robinson 364, to which Lord Sumption refers (para 83), do no more than exemplify the same point. Both concerned the civil rights of neutrals whose property had been seized during the blockade of an enemy port. They turn on the customary international law of war and prize, treated as incorporated into domestic law. The Law Officers' Report was thus sought concerning the consistency of certain Prussian prize proceedings with "the Law of Nations, and any Treaties ..., the established Rules of Admiralty Jurisdiction, and the Laws of this Kingdom" (p 889). The Report was given on the basis that "By the Maritime Law of Nations, universally and immemorially received, there is an established method of determination, whether the Capture be or be not lawful Prize" (p 890), and that "In this method, by Courts of Admiralty acting according to the Law of Nations and particular Treaties, all captures at sea have immemorially been judged of, in every Country of Europe" (p 892). The Admiralty Court had "immemorially" held trading with enemy subjects to be illegal: see Oppenheim's International Law (7th ed) paras 101, footnote 2, and 192; and see also McNair and Watts, The Legal Effects of War (1966) pp 336-337.
Neither the Law Officers' Report nor "The Rolla" bears on the issue currently under discussion. They concern the application of customary international law as and where incorporated into domestic law, giving rise to a foothold for domestic adjudication, as there was for example in the case of "The Rolla", where the issue concerned a claim to seize and condemn a ship as prize for contravention of a legal blockade.
The Secretary of State in Council of India v Kamachee Boye Sahaba (1859) 7 Moo Ind App 476, 19 ER 388is, in contrast, an example of a case falling within the conduct of foreign affairs which the Privy Council held was unsuitable for adjudication in a domestic court. It concerned the annexation by the East India Company (exercising sovereign power on behalf of the British Government) of a foreign territory and the taking of its late ruler's public and private property. The late ruler's eldest widow brought an action claiming to be entitled to the private property which (the Privy Council was ready to accept) would under Hindoo law pass and belong to the late ruler's eldest widow. In relation to the appropriation of the private property, there was clearly a domestic foothold for the claim. But the claim failed in its entirety. This was not because it was non-justiciable in some narrow sense, involving absence of judicial or manageable standards. A domestic court could, if necessary, identify standards and rules of international law by reference to which to adjudicate upon such a dispute. It was because the whole case fell within the category of non-justiciability identified by Lord Roskill in GCHQ and the Supreme Court in Shergill v Khaira, para 42. It was a case upon which domestic courts should not adjudicate because of its nature and subject-matter. It was a classic case of intervention by forces acting for the British Crown intervening in and taking over a foreign territory and property of its subjects.
Attempts at a bifurcation in this area between
cases of non-justiciability (or cases not open to question) in a domestic court and
cases falling within the so-called rule in Buron v Denman are in my view incorrect and confusing.
The correct analysis is that the rule applied in Buron v Denman is no more than a corollary of the principle of non-justiciability, abstention or restraint. In short, the rule applied protects the Crown's servants or agents in circumstances where that principle precludes a claim against the Crown itself. A claim which is non-justiciable against the Crown itself cannot be justiciable against the servants or agents who, with the Crown's authority or subsequent ratification, undertook the relevant acts. Otherwise, the principle of Crown act of state could and would be subverted.
The point is illustrated by the case of Secretary of State in Council of India v Kamachee Boye Sahaba itself. The action there was not against the Crown, but for an act done, as Lord Kingsdown said, by the East India Company as "its delegate", over which act domestic courts had "no jurisdiction" and "of the propriety or justice of that act, neither the court below nor the Judicial Committee have the means of forming, or the right of expressing, if they had formed, any opinion" (p 540). See also, in the quotations from Lord Kingsdown's judgment that Lord Sumption sets out in para 86: "such [viz municipal] courts have neither the means of deciding what is right" and "an act not affecting to justify itself on grounds of municipal law"; "over which the Supreme Court of Madras has no jurisdiction" and "It is sufficient to say that, even if a wrong has been done, it is a wrong for which no Municipal Court of Justice can afford a remedy".
This is the language of non-justiciability, abstention or restraint. The authorities relied on by counsel for the East India Company also demonstrate the same point. They included Tandy v Earl of Westmoreland (1792) 27 State 1246, in which "the official acts of the Lord Lieutenant of Ireland were considered acts of state, and not within the cognizance of the Municipal jurisdiction", and Mostyn v Fabrigas (1775) Cowp 161, where "Lord Mansfield laid it down that no Governor of a Colony could be sued while he is exercising the functions of a Governor". In each case, the principle identified is one of non-justiciability.
Further, in Lord McNair's magisterial and influential work International Law Opinions vol 1 (1956), from which Lord Wilberforce quoted in Nissan v Attorney General  AC 179, at p 234C-F, the general principle of non-justiciability is described in terms making clear that it protects the Crown's servants and agents as much as the Crown. The reality is that the rule in Buron v Denman is a necessary aspect of the principle of abstention or non-justiciability. Likewise, in the analogical context of state immunity, not only the Crown, but also its servants and agents are protected, or otherwise the rule would be subverted: see Jones v Saudi Arabia  1 AC 270 . Equally, it cannot be open to a Crown servant or agent by waiving a plea of Crown act of state to enable a domestic court to adjudicate upon an area falling within the scope of the concept. Yet that would seem the consequence if Crown act of state were a mere defence, rather than a bar to adjudication based upon a principle of abstention or restraint.
There were only two issues in the Court of Exchequer case of Buron v Denman. One, swiftly dispatched by Parke B in summing up (p 187), was whether the plaintiff could show sufficient property or possession in slaves. The other, the main issue, was whether a Crown servant or agent could by ratification of his acts by the Crown be put in the same position as if those acts had been authorised by the Crown from the outset. Had there been prior authorisation, it is clear from the way that Buron v Denman was argued, that there would have been no doubt that the action was not maintainable. The Attorney General cited briefly in this connection at the end of his submissions (p 185) Elphinstone v Bedreechund, which concerned British forces' seizure of a military fortress during military hostilities and is, correctly, analysed by Lady Hale in para 25 as a case of non-justiciability.
Conversely, without prior authorisation or ratification, it was well established before Buron v Denman that a Crown servant could be held liable for unauthorised naval action taken against foreign slavers conducting a trade which remained lawful under the (Spanish) law of the flag of the ship they were using. In Madrazo v Willes (1820) 3 B & Ald 353, relied on by counsel for the plaintiff in Buron v Denman, Captain Willes, commanding a Royal Navy vessel, had without authority taken possession of a Spanish brig engaged in the slave-trade between Africa and Cuba, detaining her, her stores and other goods as well as 300 slaves, and preventing her from further trading. The Spanish owner sued Captain Willes, maintaining (consistently with the principle in Entinck v Carrington (1765) 19 State Tr 1030) that he was personally liable. The only question which arose was as to the measure of damages. Reluctantly, the Court of King's Bench found itself obliged to award SeñorMadrazo damages which included not merely the deterioration of the ship's stores and goods, but also the alleged profit which would have been made from the ship's cargo of slaves.
When Buron v Denman was argued nearly 30 years later, the only substantial question was therefore whether a different result could and should follow if the Crown had, after the event, purported to ratify what its naval captain had done. There was a faint suggestion, which led to nothing, that Lord John Russell's and Viscount Palmerston's commendations of Captain Denman's "very spirited and able" and "highly meritorious" conduct and their expressions of desire that such conduct should be repeated, whenever occasion arose, were insufficient to amount to ratification in fact. The plaintiff's real argument was that ratification was only permissible if the act would have been justified, if done by the principal, whereas here it was "not for the purpose of showing that the act was justifiable, but for the purpose of protecting the party committing it against examination as to whether it was right or wrong". That ratification was permissible, if the act done would under domestic law have been justified if done by the principal, was in fact demonstrated by "The Rolla". There a British blockade of Monte Video (a Spanish enemy port) was in principle legal under international and domestic law, had however actually been imposed by the local British fleet commander, Sir Home Popham, without governmental authority, but was ratified by the British government after the event. Parke B was in Buron v Denman concerned by the plaintiff's submission that a plea of Crown act of state was of a different character, since it did not turn on any conclusion that the act would, if authorised by the Crown be lawful (but simply withdrew it from domestic adjudication). Ultimately, however, he joined with the other members of the Court in holding that ratification was in this context also equivalent to prior authorisation – even if it left the plaintiff without remedy against the Crown because the injury would count as "an act of state without remedy" (p 189).
there is only one principle of Crown act of state;
Buron v Denman is simply authority for the proposition that conduct capable by its nature of being an act of state may be so not only when authorised in advance but also when subsequently ratified by the Crown; and
Crown servants or agents committing an act of state with prior authorisation or subsequent ratification by the Crown enjoy the same immunity from liability that the Crown does – otherwise, indeed, the doctrine of Crown act of state would have very little bite at all.
In support of this analysis, I note the following further points:
The suggestion that there are two separate rules of Crown act of state, operating somehow in parallel but at different levels, stems essentially from dicta of Lord Wilberforce in Nissan v Attorney General  AC 179, 231C-E. Nothing said by other members of the House supports such a bifurcation: see eg per Lord Reid at pp 207G, 208C-G and 212C-D and Lord Morris at pp 219B-221B. Both analysed the issue in Nissan as turning on the scope of the rule in Buron v Denman – ie as treating the rule in Buron v Denman and the principle of non-justiciability as interdependent.
It is far from clear that Lord Wilberforce intended the conceptual distinction now proposed between two separate rules. He himself spoke of Crown act of state as "a principle .... that .... includes within itself two conceptions or rules". His first conception or rule can be seen to have been focused on the liability or immunity of Crown servants whose acts have been authorised or subsequently ratified by the Crown. On that basis, his second conception or rule represents the sole principle focusing on the case of a claim against the Crown itself.
A precursor to Lord Wilberforce's dicta consists in Lord McNair's International Law Opinions vol 1 (1956), from which, as I have already noted, Lord Wilberforce quoted at p 234C-F in Nissan. Lord McNair confined discussion of the rule in Buron v Denman to circumstances where a claim is made against a Crown servant, and dealt with non-justiciability as a "wider and more fundamental" principle precluding claims against the Crown, its servants or agents (pp 111-112). But, in circumstances where Crown servants are protected under the rule in Buron v Denman, the Crown itself must also be protected. The inference again is that the rule in Buron v Denman is simply an aspect of the protection afforded by the wider and more fundamental principle of non-justiciability. The two rules cover different facets of the same situation.
Further, in so far as Buron v Denman addressed the Crown's immunity from suit in relation to foreign military activity at all, it was based on authority addressing circumstances of non-justiciability and has subsequently been analysed in the same terms. Thus:
the Attorney General in Buron v Denman successfully advanced Crown act of state as a defence by referring to circumstances which were and are clearly non-justiciable, referring (at p 184) to acts under a treaty and, as I have already noted, by referring (at p 185) to Elphinstone v Bedreechund, which concerned seizure of a military fortress during military hostilities and is a case of non-justiciability;
the rationale of Buron v Denman is clearly identified in later authority at the highest level as being that it concerned non-justiciable activity, that is (in the light of the ratification) state activity undertaken abroad as a matter of policy at an inter-state level or in the course of something like military operations against a foreign state or its subjects: see Johnstone v Pedlar  AC 262, per Viscount Finlay, p 271 foot; per Viscount Cave, p 275 foot; per Lord Atkinson, p 279; and per Lord Sumner, p 290 and pp 291-292. In these passages, both Viscount Cave and Lord Atkinson assimilated Buron v Denman and Kamachee Boye Sahaba, which is another case correctly analysed by Lady Hale at para 25 as an instance of non-justiciability.
In addition, no rationale for or explanation of the contours of any distinction between circumstances of non-justiciability and circumstances falling within a supposedly separate rule to be derived from Buron v Denman is available. Indeed, Lord Sumption argues that the two rules are in the present context one (para 80), but only (as I have pointed out in para 48 above) by assigning to non-justiciability the unnatural meaning of a defence. He suggests that, contrary to contemporary and later views, Buron v Denman was a case of a tort law defence.
Taking the criteria for activities which are non-justiciable or inappropriate to be questioned in domestic civil proceedings, they clearly include all those identified by Lord Sumption in para 82, that is
they must involve an exercise of sovereign power, inherently governmental in nature;
done outside the United Kingdom;
with the prior authority or subsequent ratification of the Crown; and
in the conduct of the Crown's relations with other states or their subjects (possibly excluding persons owning allegiance to the Crown).
I add two points.
First, Crown act of state must be potentially applicable as much to acts in the execution of policy-makers' decisions as it is to the decisions themselves. It would not otherwise be a coherent doctrine. In this, I am at one with Lady Hale (para 33) and Lord Sumption (para 90).
Second, in relation to the availability of Crown act of state as a plea in relation to conduct towards the subjects of foreign states: see eg the citations which Lady Hale gives in her para 2, the first of which was also quoted and endorsed by Lord Wilberforce in Nissan at p 231B; see also per Lord Reid at p 212C-D.
The upshot is that the criteria suggested for the rule in Buron v Denman are the same criteria as lead to a conclusion that circumstances are non-justiciable or inappropriate for adjudication in domestic civil proceedings. The reason is clear. There is only one principle, though it has different aspects protecting Crown servants or agents (Buron v Denman) and the Crown more generally. What matters in any case is therefore its scope and application in relation to the particular circumstances the subject of the relevant civil proceedings.
As to this, I agree with Lady Hale and Lord Sumption that the present claimants' detention by Her Majesty's forces and their transfer from British to United States and Afghan custody were, as such, Crown acts of state which are not justiciable or open to question in domestic proceedings for common law damages such as the present. They were, on the actual or presently assumed facts, steps taken pursuant to or in implementation of deliberately formed policy against persons (none owing any allegiance to the Crown) reasonably suspected to be insurgents or terrorists in the context and furtherance of foreign military operations during a time of armed conflict.
I also agree with Lady Hale's conclusions regarding the Crown Proceedings Act 1947 and article 6 of the European Convention on Human Rights. As she observes, the rule of Italian law considered by the European Court of Human Rights in Markovic v Italy 44 EHRR 52was effectively a rule of non-justiciability. The Italian Court of Cassation had before it claims by relatives of persons killed in the NATO bombing of Belgrade, in which Italian forces had participated. The Court of Cassation categorised the impugned act as an act of war, and said that "since such acts were a manifestation of political decisions, no court possessed the power to review the manner in which that political function was carried out" (para 106). The European Court of Human Rights said, at para 114, that
the Court of Cassation's ruling .... does not amount to recognition of an immunity but is merely indicative of the extent of the courts' powers of review of acts of foreign policy such as acts of war. It comes to the conclusion that the applicants' inability to sue the state was the result not of an immunity but of the principles governing the substantive right of action in domestic law.
This statement fits precisely the circumstances of the present case on my approach to Crown act of state.
Further, in the light of the above, I agree that there should be a declaration in each appeal as Lady Hale proposes in her para 46, and that we should invite further submissions on its precise form. In the cases of Yunus Rahmatullah and the Iraqi Civilian Litigation, I would specifically invite further assistance as to the effect and appropriateness of the qualifying adjective "lawful" quoted by Lady Hale in her para 46.
In Nissan v Attorney General  AC 179, 231, Lord Wilberforce, whose speech comes closest to supplying a coherent judicial statement of the doctrine of Crown act of state, reviewed the main relevant authorities on the doctrine and concluded that it comprised two rules. One was a rule of non-justiciability, by which he meant a rule which "prevents British municipal courts from taking cognisance of certain acts". The other was a rule which
provides a defendant, normally a servant of the Crown, with a defence to an act otherwise tortious or criminal, committed abroad, provided that the act was authorised or subsequently ratified by the Crown.
The dichotomy between these two rules had previously been suggested by Lord McNair in International Law Opinions (1956), pp 111-116.
"Non-justiciability" is a treacherous word, partly because of its lack of definition, and partly because it is commonly used as a portmanteau term encompassing a number of different legal principles with different incidents. Strictly speaking, as this court observed in Shergill v Khaira  AC 359at para 41, it should be reserved for cases where an issue is said to be inherently unsuitable for judicial determination by reason only of its subject matter. This may result in a court declining to determine an issue notwithstanding its relevance to the dispute between the parties, for example because there are no juridical standards by which to determine it, as in Buttes Gas and Oil Co v Hammer (No 3)  AC 888; or because its determination is not within the constitutional competence of the courts, for example because it would trespass on Parliamentary privilege, as in Prebble v Television New Zealand Ltd  1 AC 321. These are mandatory rules of public policy, originating in the law's recognition of the separation of powers between different organs of the state. They define the limits of the court's jurisdiction or juridical competence. But there are other principles, also originating in the separation of powers and described as principles of non-justiciability, which do not go to the court's jurisdiction or competence but to the existence or scope of legal rights. Thus in R (Campaign for Nuclear Disarmament) v Prime Minister  EWHC 2777 (Admin), the court proceeded on the footing that it had both jurisdiction and competence to determine whether a resolution of the United Nations Security Council authorised military operations against Iraq, but declined to do so because, among other reasons, there were no relevant " rights, interests or duties under domestic law": paras 14-15, 36. I venture to suggest that if domestic law rights, interests or duties had been engaged, the court would not have regarded the issues as non-justiciable.
Crown act of state is a rule of substantive law which belongs in this latter category. The court is not disabled from adjudicating on a Crown act of state by virtue of its subject-matter. The acts of the Crown and its agents are always in principle subject to the adjudicative power of the courts. They unquestionably have both jurisdiction and competence to determine the legal effects of a Crown act of state on the rights of those adversely affected by it. The real question is what are those rights. The rule of law relating to Crown acts of state defines the limits which as a matter of policy, the law sets upon certain categories of rights and liabilities, on the ground that they would otherwise be inconsistent with the exercise by the executive of the proper functions of the state. In principle an agent of the Crown is liable as a matter of English law for injury or detention of persons or goods without lawful authority. But that liability does not extend to a limited class of acts constituting Crown acts of state. It follows that the agent has a defence if his acts fall within that class. Like other members of the court, I doubt whether it helps to treat the doctrine as comprising two rules. But in this context, it can fairly be said that Lord Wilberforce's two rules merge into one.
I agree with Lady Hale that a Crown act of state gives rise to no liability on the part of the Crown or its agents. I also agree with her upon the essential elements of a Crown act of state in this context. They are (i) that the act should be an exercise of sovereign power, inherently governmental in nature; (ii) done outside the United Kingdom; (iii) with the prior authority or subsequent ratification of the Crown; and (iv) in the conduct of the Crown's relations with other states or their subjects. There may be a fifth requirement, that the alleged tort should have been committed against a person not owing allegiance to the Crown. But that raises a distinct and controversial question which does not need to be decided on these appeals. The claimants in these proceedings did not owe allegiance to the Crown.
Although the label "act of state" is modern, the concept is very ancient. The earliest illustrations relate to the right to seize ships or cargoes at sea. The right, without incurring liability under English law, to seize property under letters of marque and reprisal issued on the authority of the Crown, even in peacetime, dates back to the 13th century. It was not, however, until the 18th century that the underlying rationale of the doctrine began to emerge. The growth of British seapower made it necessary to consider the interrelation between international and municipal law concerning captures at sea. In a celebrated opinion of 1753, written by Sir William Murray, later Lord Mansfield, the law officers of the Crown advised that a belligerent power was entitled in international law to seize not only enemy property but the property of neutrals destined for an enemy: British and Foreign State Papers, 20 (1836), 889ff. The result was that the seizure of the property on behalf of the Crown gave rise to no right to damages or possession at the suit of the former owner. In "The Rolla" (1807) 6 Robinson 364, 365-367, Sir William Scott, perhaps the greatest British international lawyer of his day, identified the basis of the rule as being the authority or ratification of the Crown in the exercise of its sovereign power. The result was that the American owner of a cargo had no rights under English municipal law in respect of the seizure of his property by Admiral Sir Home Popham in the course of his highly irregular (but ratified) blockade of the River Plate in 1806.
Greater definition was brought to this area of law in two seminal cases decided in the middle of the 19th century: Buron v Denman (1848) 2 Exch 167, and Secretary of State in Council of India v Kamachee Boye Sahaba (1859) 7 Moo Ind App 476, 19 ER 388.
Buron v Denman is one of those cases which is more significant for what it has always been understood to have decided, than for anything Parke B actually said in the course of his summing up to the jury. Its significance is that Captain Denman's act in seizing the plaintiff's slaves and destroying his property in the Gallinas in West Africa was not a valid act of war, since Britain was at peace with Spain. Nor was it justifiable in international law, since the slave trade had been held to be lawful by the law of nations: see Le Louis (1817) 2 Dods 210. Although the indigenous ruler of the Gallinas had undertaken by treaty with Captain Denman to destroy the barracoons and surrender the slaves, he had not authorised Captain Denman to do these things, which was presumably why the treaty was not relied upon by Captain Denman and ignored by Parke B. There was therefore no legal basis whether in international or municipal law for the invasion of SeñorBuron's proprietary rights. In those circumstances, the only plea available to Captain Denman was that by virtue of the Crown's adoption of his acts, they were acts of state. The judge took it to be axiomatic that the prior authority of the Crown would have constituted a defence. The defendant would in that case be "irresponsible" (p 190), ie not liable. The only contentious issue was whether subsequent ratification was equivalent to prior authority. He held that it was.
Parke B did not explain why it went without saying that the authority of the Crown was a defence, but the basis of the rule became clearer a decade later in the advice of the Privy Council in Secretary of State in Council of India v Kamachee Boye Sahaba. Lord Kingsdown, delivering the advice of the Board, applied the principle in Buron v Denman (see pp 539-540) to the annexation of the Indian state of Tanjore and the seizure of the late Rajah's property there by the East India Company in the exercise of the sovereign power of the Crown. He declared, at p 529:
The general principle of law was not, as indeed it could not, with any colour of reason be disputed. The transactions of independent states between each other are governed by other laws than those which Municipal Courts administer: such courts have neither the means of deciding what is right, nor the power of enforcing any decision which they may make.
Lord Kingsdown went on, at p 531, to inquire what was the nature of the act of the East India Company's officers:
Was it a seizure by arbitrary power on behalf of the Crown of Great Britain, of the dominions and property of a neighbouring state, an act not affecting to justify itself on grounds of municipal law? Or was it, in whole or in part, a possession taken by the Crown under colour of legal title of the property of the late Rajah of Tanjore, in trust for those who, by law, might be entitled to it on the death of the last possessor?
He concluded (p 540) that
.. the property now claimed, by the respondent has been seized by the British Government, acting as a Sovereign power, through its delegate the East India Company; and that the act so done, with its consequences, is an act of state over which the Supreme Court of Madras has no jurisdiction. Of the propriety or justice of that act, neither the court below nor the Judicial Committee have the means of forming, or the right of expressing, if they had formed, any opinion. It may have been just or unjust, politic or impolitic, beneficial or injurious, taken as a whole, to those whose interests are affected. These are considerations into which their Lordships cannot enter. It is sufficient to say that, even if a wrong has been done, it is a wrong for which no Municipal Court of Justice can afford a remedy.
Leggatt J regarded this as a "perverse doctrine under which the executive can be held to account if it purports to act legally, but not if it openly flouts the law." But I think that the judge has misunderstood Lord Kingsdown's reasoning. Lord Kingsdown was not saying that the East India Company could not be held to account because it had openly flouted the law. He was doing two things. In the first place he was pointing out that even if the Crown's annexation of Tanjore was unlawful in international law that could not of itself give rise to any legal rights in municipal law. Secondly, he was distinguishing between the sovereign and non-sovereign acts of the Crown. If the seizure of the late Rajah's property had been carried out under colour of municipal law, for example as a taking of possession by a trustee, it would not have been a sovereign act but an act such as any non-sovereign could have done. As it was, it was an extraterritorial exercise of sovereign power, and as such an act of state. It therefore gave rise to no actionable duty owed to the late Rajah's heirs.
The judgment of Lord Kingsdown has been treated by the House of Lords and the Privy Council on many occasions since it was decided as an authoritative statement of the law: see, among other cases,
Sirdar Baghwan Singh v Secretary of State for India  LR 2 Ind App 38, 47,  UKPC 63;
Cook v Sprigg  AC 572;
Johnstone v Pedlar  2 AC 262, 275 (Viscount Cave), 278-279 (Lord Atkinson), 290-291 (Lord Sumner);
Vajesingji Joravarsingji v Secretary of State for India  LR 51 Ind App 357,  UKPC 51;
Secretary of State for India v Sardar Rustam Khan  AC 536;
Nissan v Attorney General  AC 179, 218 (Lord Morris of Borth-y-Gest), 225 (Lord Pearce), 231-232 (Lord Wilberforce), 238 (Lord Pearson).
The reason why the liabilities of the Crown in municipal law do not extend to sovereign acts done in the course of military operations outside the United Kingdom is essentially a principle of consistency. The deployment of armed force in the conduct of international relations, or the threat of its deployment (express or implicit) is one of the paradigm functions of the state. The law vests in the Crown the power to conduct the United Kingdom's international relations, including the deployment of armed force in support of its objectives. Constitutionally, as Blackstone observed, the result is that "what is done by the royal authority with regard to foreign powers is the act of the whole nation:" Commentaries, para 252. Or, as Willes J put it a century later in Esposito v Bowden 7 EL & BL 763, 781 (1857), speaking of a declaration of war, "as an act of state, done by virtue of the prerogative exclusively belonging to the Crown, such a declaration carries with it all the force of law". In the nature of things, the use of armed force abroad involves acts which would normally be civil wrongs not only under English law but under any system of municipal law. People will be detained or killed. Their property will be damaged or destroyed. It would be incoherent and irrational for the courts to acknowledge the power of the Crown to conduct the United Kingdom's foreign relations and deploy armed force, and at the same time to treat as civil wrongs acts inherent in its exercise of that power.
In this respect, Crown act of state differs from foreign act of state. When the courts consider an exercise of sovereign authority by a foreign state, no question of consistency arises because the sovereign authority of the foreign state is not derived from English law. Foreign act of state operates purely as a rule of non-justiciability. Its effect in the very limited class of cases to which it applies is not to afford a defence but to preclude the courts from taking cognisance of an alleged civil wrong if it necessarily depends on determining the lawfulness of a foreign act of state.
None of this means that whatever an agent of the Crown does pursuant to its decisions in the conduct of the United Kingdom's foreign relations gives rise to the defence of act of state. The boundaries are admittedly difficult to draw. The only extended discussion appears in the speeches in the House of Lords in Nissan v Attorney General  AC 179 . But the inconsistencies between them, the unsatisfactory terms of the pleading on which the argument was based and the obscurity of the facts combine to make it hard to extract any very clear ratio from this decision. It is unquestionably right to say, as Lord Pearson did at p 237F, that an act of state "must be something exceptional". He cited the making of war and peace, the making of treaties and annexations or cessions of territory as "obvious examples" (p 237F-G), and the dispatch of a peacekeeping force to the territory of an independent sovereign as having "to some extent the character of acts of state" even if it did not follow that everything that it did there was an act of state (pp 239F-240B). But this brings one no closer to a workable criterion on which to decide cases like the present ones. In my opinion, the main relevant limitations on the act of state doctrine are implicit in the doctrine itself. In particular, they are implicit in the requirement that the act must be inherently governmental in nature, and either specifically authorised or ratified by the Crown or inherent in what the Crown has authorised or ratified. Without seeking to formulate a comprehensive definition of a rule whose application is inevitably fact-sensitive, I consider that the following points can fairly be made.
The first is that an act does not need to raise questions of "high policy" in order to give rise to a plea of Crown act of state. This is because the rule extends not just to the decisions of policy-makers, but to actions taken by the Crown's agents in the execution of those decisions, often at a relatively low level, far below the level of policy-making. Moreover, as Lord Reid pointed out in Nissan, at p 212, acts which are unauthorised but ratified after the event are unlikely to have been done in accordance with any high policy of the Crown.
Secondly, it is sometimes said that the act must be the necessary consequence of a decision made by the Crown through its ministers. I think that this is right, provided that we are careful about what we mean by necessary. In rejecting the Crown's reliance on act of state in Nissan, two members of the Appellate Committee observed that while the agreement with the government of Cyprus to station troops on the island was itself an act of state, the occupation of the Cornaro Hotel was not necessary to its implementation: pp 216-217 (Lord Morris of Borth-y-Gest), 227B-C (Lord Pearce). But it is important to guard against the suggestion that the availability of the act of state defence depends on a judicial assessment of the political or tactical alternatives, an exercise which must be left to the judgment of the executive or its officers on the spot. As Sir William Scott observed in "The Rolla", at p 366, "a commander going out to a distant station may reasonably be supposed to carry with him such a portion of sovereign authority delegated to him, as may be necessary to provide for the exigencies of the service on which he is employed." In my opinion, the question depends on the character of the act. It is whether an act of that character is inherent in what the Crown has authorised or ratified. It is in this sense that the concept of necessity is used in this context.
Thirdly, however, the fact that the act is of a kind which is inherent in what the Crown has authorised or ratified, although undoubtedly a necessary condition, cannot be a sufficient one. It must also be by its nature a sovereign, ie an inherently governmental act, for the Crown to be capable of authorising or ratifying it as an act of state. In his speech in Nissan (p 218F), Lord Morris, after quoting the definition of act of state in the then current edition of Halsbury's Laws of England ("an act of the executive as a matter of policy performed in the course of its relations with another state including its relations with the subjects of that state"), observed, at p 218F:
I do not think that such actions as securing food or shelter in peace time for troops situate abroad are to be regarded as acts of the executive performed in the course of relations with another state within the conception of the above definition.
I think that this was the true ratio of the decision. The appropriation of the hotel was not an inherently governmental act in the circumstances pleaded. It was an ordinary case of the army acquiring accommodation in peacetime, in respect of which they were in no different position from any other organisation acquiring accommodation. They therefore had to pay like any one else. As Lord Pearson suggested at p 240B, the position might be different if there had been an "urgent military necessity" to occupy the hotel.
I would prefer to reserve my opinion on the question whether the appropriation of property, with or without compensation, can be an act of state. I think that the answer would be likely to depend on the circumstances. In Buron v Denman and in Kamachee Boye Sahaba and other colonial annexation cases, the seizure of property without compensation was held to be an act of state. The same would, I suspect, be true of most appropriations of property in the course of active military operations. In other circumstances, like those in Nissan, the position would be different. We have heard no argument on this question.
In the present cases, Crown act of state is raised by the Secretary of State only so far as the allegations are based on the mere fact of the claimants' detention by Her Majesty's forces or the mere fact of their transfer from British custody to that of the United States. In my opinion these were acts of state so far as they were authorised by the United Kingdom's detention policy or required by the United Kingdom's agreements with the United States, these being the only particulars of authority relied upon. If these criteria were satisfied, they were both inherently governmental in character and authorised by the Crown in the conduct of the United Kingdom's international relations. The Crown and its servants could not therefore be liable for them in tort. I would make a declaration to that effect.
The Secretary of State denies that the claimants were maltreated, but does not contend that any maltreatment which may have occurred was an act of state. That is as one would expect. Any maltreatment of detainees was not authorised by the United Kingdom's detention policy. It is not alleged to have been authorised in any other way, or to have been ratified. It is therefore unnecessary to address the question whether the maltreatment of detainees ever could be an act of state, in the highly improbable event of its being done with the authority of the Crown. I would merely record my reservations about Lady Hale's suggestion that the torture or maltreatment of prisoners is not an inherently governmental act, although I agree that in the light of the government's statements on the subject this is a moot point. As an international crime and a statutory offence in the United Kingdom, torture is by definition a governmental act: see Jones v Saudi Arabia  1 AC 270, paras 19 (Lord Bingham of Cornhill) and 81-85 (Lord Hoffmann). There are, unfortunately, well documented modern instances across the world of the use of torture and other forms of maltreatment as an instrument of state policy authorised at the highest levels. There is a more satisfactory answer to the hypothetical problem of governmental torture and deliberate governmental maltreatment. Given the strength of the English public policy on the subject, a decision by the United Kingdom government to authorise or ratify torture or maltreatment would not as a matter of domestic English law be a lawful exercise of the royal prerogative. It could not therefore be an act of state. Nor would there be any inconsistency with the proper functions of the executive in treating it as giving rise to civil liability.
I have nothing to add to Lady Hale's analysis of the Crown Proceedings Act 1947, or her conclusions about article 6 of the European Convention on Human Rights, with which I entirely agree.
(with whom Lord Hughes agrees)
This aspect of these proceedings concerns the principle or doctrine of Crown act of state, which has been raised by the defendants in circumstances which have been explained by Lady Hale in paras 1-14 above.
Crown act of state, like foreign act of state, is a doctrine which has been developed by judges over the years, as explained in the judgments of Lady Hale, Lord Mance and Lord Sumption. It would be a fruitless exercise to try and reconcile all the judicial dicta, even from the House of Lords, on this doctrine. Indeed, it is very difficult to identify a comprehensive definition of the doctrine, as is clear from the somewhat different approaches in the speeches of Lord Reid, Lord Morris and Lord Wilberforce in the most recent decision of the House of Lords on the topic, Nissan v Attorney General  AC 179 .
A remarkable aspect of this doctrine is the weight that has been given to Buron v Denman (1848) 2 Exch 167, given that it wasa direction to a jury where the nature and extent of the doctrine was not really in issue. The fact that the judge concerned was Baron Parke no doubt helps to explain why the ruling has been accorded particular respect. However, in agreement with Lord Mance's analysis of the report, it seems to me that, to the extent that the case is strictly an authority, it is simply for the proposition that an action which would have been an act of state if it had been authorised in advance, will (or at least may) be treated by the court as an act of state if it is subsequently ratified by the Crown. Nonetheless, Buron was cited with approval by Lord Reid, Lord Morris and Lord Wilberforce in Nissan, and by Viscount Finlay, Viscount Cave, and Lord Sumner in Johnstone v Pedlar  2 AC 262, in relation to what Baron Parke said about the doctrine of Crown act of state.
The fact that any attempt to define the precise nature and extent of the principle of Crown act of state is doomed to failure is unsurprising. The doctrine is ultimately based on judicial decisions and dicta as to when the judiciary should decline to rule on the lawfulness of an act on the ground that any challenge to the act should be left to the executive, at least normally where the act is based on the Royal prerogative. However, it is only in relation to some acts based on the Royal prerogative that when a court will decline to adjudicate, namely acts which, because of their nature or circumstances, call for judicial self-restraint. There have been very few cases in the past 100 years when the doctrine has been considered and hardly any in which it has been held to apply. And decisions given even 50 years ago may reflect a somewhat different approach to that which appears appropriate today, following the growth of judicial review and the introduction of human rights into our domestic law. The difficulty in identifying or delimiting the doctrine is reinforced by the flexibility and imprecision of the United Kingdom's constitutional settlement. And it appears to me that any observation on the doctrine prior to the decision of the House of Lords in Council of the Civil Service Unions v Minister for the Civil Service  AC 374 must be considered with particular caution, essentially for the reason given by Lady Hale in para 15.
However, despite these points, there is no doubt that the doctrine of Crown act of state remains a constitutionally important, if rarely invoked, feature of the common law. For the reasons already given, I agree with Lady Hale (in para 36) that it would be unwise for us to propound a definitive statement as to when the Crown act of state doctrine can be invoked. However, I entirely endorse the attempts in the preceding judgments to give as much guidance as we can on the extent of the doctrine. In that connection, I consider that the formulations of Lady Hale (in paras 32 and 36-37), of Lord Mance (in paras 56-58 and 64) and of Lord Sumption (in paras 88-93) provide helpful guidance as to what may constitute (or may not constitute) a Crown act of state, as do the definitions cited in para 2 of Lady Hale's judgment.
That leaves the question of the proper characterisation of the doctrine.
In those rare cases where an issue involving a Crown act of state arises, it does not mean that a judge lacks the information or expertise to resolve the issue (although in some exceptional cases that may be a different reason for the court not determining an issue). As Lady Hale says in para 45, this is not because the doctrine bestows any sort of immunity, or indeed because of any judicial discretion: it is because there are certain acts of the UK government (sc the executive) which, owing to their nature or circumstances, are not susceptible to judicial assessment. As Lord Sumption says in para 88, the doctrine is ultimately based on the need for consistency or coherence in the distribution of functions between the executive and the judiciary in the United Kingdom's constitutional arrangements. Accordingly, if a claim depends on establishing the unlawfulness of a Crown act of state, then, as a matter of United Kingdom law, the claim must fail, as a Crown act of state cannot give rise to a legal liability.
When Crown act of state applies to a particular act, that act is often described as being non-justiciable. However, as Lord Sumption explains in para 79, the expression non-justiciable can have a number of different meanings, and, for that very reason it seems to me that it is one which is best avoided if one is seeking to explain precisely why an issue cannot be resolved because the doctrine of Crown act of state applies. Thus, the expression "non-justiciable" could well be understood as suggesting that the court is incapable of determining, or choosing not to determine, the lawfulness of the act in question, or that the court is declining to address any legal liability flowing from that act. But, as I have just explained (and is explained more fully in the preceding judgments), none of those analyses represents the basis of Crown act of state.
I agree with Lady Hale, for the reasons which she gives, that neither the Crown Proceedings Act 1947 nor article 6 of the European Convention on Human Rights assists the respondents. However, I also agree with her, Lord Mance and Lord Sumption, for the reasons which they all give, that the doctrine of Crown act of state can be relied on by the defendants in this case, and accordingly I would join them in allowing this appeal.
I have read all these judgments with interest and admiration. There has been much debate as to whether Crown act of state involves one or two principles. However nobody has so far suggested a case in which it would make a difference as to which of the two principles applied. Like Lord Sumption (the principal proponent of the two principles approach), at the end of para 80, I doubt whether it helps to treat the doctrine as comprising two rules or one. I also agree with him that in this context it can fairly be said that Lord Wilberforce's two rules merge into one.
The only point that I would stress is this. It does seem to me that whether there is one principle or two, the question whether a defendant can successfully rely upon Crown act of state does not involve the court exercising a discretion or anything approaching a discretion. The defendant either has a legal right to rely upon Crown act of state or it does not. Only in the former case will it succeed, whether it is held that it does so by way of defence or by the application of the principle of non-justiciability.
I agree that the disposition of this appeal should be as proposed by the other members of the court for the reasons they give. In so far as there may be differences between them, I do not detect any difference which is critical to the resolution of this appeal.
Derek Sweeting QC & James Purnell (instructed by The Government Legal Department) for Appellants (MOD in XYZ and ors).
Phillippa Kaufmann QC, Ben Jaffey & Edward Craven (instructed by Leigh Day) for Respondents (Rahmatullah and XYZ and ors).
all rights reserved