Benkharbouche v Secretary of State for Foreign and Commonwealth Affairs

The question at issue on this appeal is whether two provisions of the State Immunity Act 1978 are consistent with the European Convention on Human Rights and the European Union Charter of Fundamental Rights. The two provisions are section 4(2)(b) and section 16(1)(a). I shall set out both below, but in summary the effect of section 4(2)(b) is that a state is immune as respects proceedings relating to a contract of employment between a state and a person who at the time of the contract is neither a national of the United Kingdom nor resident there; and the effect of section 16(1)(a) is that a state is immune as respects proceedings concerning the employment of members of a diplomatic mission, including its administrative, technical and domestic staff. It is common ground that the answer depends in both cases on whether these provisions have any basis in customary international law although, as I shall explain, there is an issue about what kind of basis it must have.

Ms Minah Janah, the Respondent to this appeal, is a Moroccan national. In 2005, when she was resident in Libya, she was recruited to work for the Libyan government as a domestic worker at its embassy in London. She entered the United Kingdom on a visa which recorded her status as “Domestic Worker (Diplomatic)”, and continued to work for the embassy until she was dismissed in 2012. During that time, she worked successively in a number of Libyan diplomatic households, and latterly in the residence of the ambassador. Her duties were cooking, cleaning, laundry, shopping and serving at meals. In April 2012, she began proceedings against Libya in the Employment Tribunal in support of a claim for failure to pay her the National Minimum Wage, breaches of the Working Time Regulations, failure to provide her with payslips or a contract of employment, unfair dismissal, discrimination and harassment. At all material times since her arrival in the United Kingdom, Ms Janah has been resident, but not permanently resident here.

Ms Fatimah Benkharbouche is also a Moroccan national. In 2000, when she was working for the Sudanese government in Iraq, she agreed to move to the United Kingdom to work for its embassy in London as a housekeeper and cook to the ambassador. Her employment by the London embassy began on 16 May 2000 and continued until the autumn of 2001. She then returned for some years to Iraq, before being re-engaged to work for the London embassy in the same role as before. Her second term of employment began on 28 January 2005 and continued until she was dismissed on 27 November 2010. She subsequently began proceedings in the Employment Tribunal in support of claims for unfair dismissal, failure to pay her the National Minimum Wage, unpaid wages and holiday pay, and breaches of the Working Time Regulations. By the time of her dismissal, she was permanently resident in the United Kingdom, having been granted indefinite leave to remain with effect from 25 January 2010.

It is common ground that under the terms of the State Immunity Act 1978, Libya is entitled to state immunity in respect of Ms Janah’s claim and Sudan in respect of Ms Benkharbouche’s. In Ms Janah’s case, this is because she has never been a Libyan national and was not a national or permanent resident of the United Kingdom at the time when her contract was made. Both section 4(2)(b) and section 16(1)(a) therefore apply to her. In Ms Benkharbouche’s case, it is because section 16(1)(a) applies to her. There are as yet no findings about whether the facts of her case bring her within section 4(2)(b). Both claims were dismissed by different judges in the Employment Tribunal on the ground that the employer was immune.

In the Employment Appeal Tribunal the two cases were heard together. The EAT declared that sections 4(2)(b) and 16(1)(a) of the Act should be disapplied so far as they prevented Ms Janah from bringing claims based on EU law, on the ground that they were contrary to the right of access to a court guaranteed by article 47 of the EU Charter. The claims based on discrimination and harassment and breaches of the Working Time Regulations were accordingly allowed to proceed. Leave to appeal to the Court of Appeal was granted, inter alia, in order to enable it to consider whether to make a declaration of incompatibility under section 4 of the Human Rights Act 1998. This led to the joinder of the Secretary of State under section 5 of the Human Rights Act 1998 so as to participate in the appeal. The Court of Appeal affirmed the judgment of the EAT, disapplying the relevant provisions so far as they applied to the EU law claims. It also made a declaration of incompatibility affecting all the claims, whether founded on domestic or EU law.

Sudan elected not to participate in the proceedings before the Court of Appeal and has not appealed to this court. Ms Benkharbouche was represented in the Court of Appeal but has not appeared before us. Libya participated in the proceedings in the Court of Appeal, but although it was granted permission to appeal to this court, it has not been permitted to pursue the appeal because it has failed to comply with an order of this court for security for costs. In those circumstances, the effective participants in the appeal to this court have been the Secretary of State, who appeals in both cases, and Ms Janah. We have also received written and oral submissions on behalf of the AIRE centre, and written submissions on behalf of 4A Law. It is agreed that Ms Janah’s appeal raises all of the issues in either case, but I shall refer from time to time to Ms Benkharbouche’s position also.

I propose first to examine the provisions of the State Immunity Act and then the requirements of the Human Rights Convention, before turning to the relationship between the Act and the international law of state immunity.

Before 1978, state immunity was governed in the United Kingdom by the common law. Properly speaking, it comprised two immunities whose boundaries were not necessarily the same: an immunity from the adjudicative jurisdiction of the courts of the forum, and a distinct immunity from process against its property in the forum state. During the second half of the nineteenth century, the common law had adopted the doctrine of absolute immunity in relation to both. The classic statement was that of Lord Atkin in Compania Naviera Vascongada v S S Cristina (The Cristina) [1938] AC 485, 490:

Before the adoption of the restrictive doctrine at common law, the United Kingdom had signed a number of treaties limiting the scope of state immunity in particular respects. It was a signatory to the International Convention for the Unification of Certain Rules concerning the Immunity of State-owned Ships (Brussels, 1926), which restricted the immunity of state-owned trading vessels. It had also signed the European Convention on State Immunity (Basle, 1972), a regional treaty drawn up under the auspices of the Council of Europe which identified specified categories of acts done by foreign states in the territory of the forum state which would not attract immunity. These treaties were concerned mainly with acts of a kind which would generally not attract immunity under the restrictive doctrine. But neither of them sought to codify the law of state immunity or to apply the restrictive doctrine generally. In addition, they have attracted limited international support. The Brussels Convention of 1926 has attracted 31 ratifications to date. The Basle Convention of 1972 has to date been ratified by only eight of the 47 countries of the Council of Europe.

One purpose of the State Immunity Act 1978 was to give effect to the Brussels and Basle Conventions, and thereby enable the United Kingdom to ratify them. It did this in both cases in 1979. But by this time, the conventions had been largely superseded by the adoption of the restrictive doctrine of state immunity at common law. The Act therefore dealt more broadly with state immunity, by providing in section 1 for a state to be immune from the jurisdiction of the courts of the United Kingdom except as provided in the following sections of Part I. The exceptions relate to a broad range of acts conceived to be of a private law character, including widely defined categories of “commercial transactions” and commercial activities, as well as contracts of employment and enforcement against state-owned property used or intended for use for commercial purposes. In Alcom Ltd v Republic of Colombia[1984] AC 580, 597-598, Lord Diplock, with whom the rest of the Appellate Committee agreed, observed that given the background against which it was enacted, the provisions of the Act

For present purposes, the relevant provisions of the State Immunity Act are sections 1, 3, 4 and 16. So far as they bear on the points at issue, they provide as follows:

Since the passing of the State Immunity Act, the United Kingdom has signed, but not ratified, a further treaty, the United Nations Convention on Jurisdictional Immunities of States and their Property (2004). The Convention is the result of the long drawn out labours of the United Nations International Law Commission between 1979 and 2004. For the most part, it is consistent with the United Kingdom Act, which indeed was one of the models used by the draftsmen. But there are differences, in particular relating to contracts of employment, which would require the Act to be amended before the United Kingdom could ratify it. To date, however, the United Nations Convention has attracted limited support. Twenty-eight states have signed it, including the United Kingdom. Of these, 21 have ratified it, not including the United Kingdom. Libya and Sudan have neither signed nor ratified it. It will not come into force until it has been ratified by 30 states.

The respondents’ case is that sections 4(2)(b) and 16(1)(a) of the State Immunity Act 1978 are incompatible with article 6 of the Convention, because they unjustifiably bar access to a court to determine their claims. Article 4(2)(b) is also said to be incompatible with article 14 read in conjunction with article 6, because it unjustifiably discriminates on grounds of nationality. For the moment I shall put the case on discrimination to one side, to return to it later. The main point argued before us was based on article 6.

Article 6 of the Human Rights Convention provides that “in the determination of his civil rights and obligations, or of any criminal charge against him, everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal established by law.” Two points are well-established, and uncontroversial in this appeal. The first is that article 6 implicitly confers a right of access to a court to determine a dispute and not just a right to have it tried fairly: Golder v United Kingdom(1975) 1 EHRR 524. The “right to a court” corresponds to a right which the common law has recognised for more than two centuries. As early as the 1760s, Blackstone wrote in his Commentaries, 4th ed (1876), 111:

One of the perennial problems posed by the right to a court is that article 6 is concerned with the judicial processes of Convention states, and not with the content of their substantive law. When the Duke of Westminster complained in James v United Kingdom(1986) 8 EHRR 123 that the Leasehold Reform Act 1967 allowed qualifying leaseholders to enfranchise their properties without providing any grounds on which the freeholder could object, he was met with the answer (para 81) that article 6 “does not in itself guarantee any particular content for (civil) rights and obligations in the substantive law of the Contracting States.” In Fayed v United Kingdom(1994) 18 EHRR 393, the Court explained (para 65) that it was not at liberty to “create through the interpretation of article 6(1) a substantive civil right which has no legal basis in the state concerned”, but that it would be inconsistent with the rule of law if the state were to “confer immunities from civil liability on large groups or categories of persons.” These statements have been repeated in much of the subsequent case law of the Strasbourg Court. It is not always easy to distinguish between cases in which the petitioner’s problem arose from some difficulty in accessing the adjudicative jurisdiction of the court, and cases where it arose from the rules of law which fell to be applied when he got there. The jurisprudence of the Strasbourg court establishes that, as a general rule, the question whether such cases amount to the creation of “immunities” engaging article 6 depends on whether the rule which prevents the litigant from succeeding is procedural or substantive: see, among other cases, Fayed v United Kingdom, at para 67; Al-Adsani v United Kingdom(2002) 34 EHRR 11, para 47; Fogarty v United Kingdom(2001) 34 EHRR 12, para 25; Roche v United Kingdom(2005) 42 EHRR 30, paras 118-119; Markovic v Italy(2006) 44 EHRR 52, para 94.

The dichotomy between procedural and substantive rules is not always as straightforward as it sounds, partly because the categories are not wholly distinct and partly because they do not exhaust the field. There may be rules of law, such as limitation, which are procedural in the sense that they bar the remedy, not the right, but which operate as a defence. There may be rules of law which require proceedings to be dismissed without consideration of the merits. These may be substantive rules, such as the foreign act of state doctrine, or procedural rules such as state immunity. There may be rules, whether substantive or procedural, which limit the territorial or subject-matter jurisdiction of the domestic courts, and which they have no discretion to transgress. Or the claimant’s right may be circumscribed by a substantive defence, such as privilege in the law of defamation. Or he may simply have no legal right to assert under the domestic law, for example because the law is that no relevant duty is owed by a particular class of defendants although it would be by defendants generally. But these are not refinements with which the Strasbourg court has traditionally been concerned. What the Strasbourg court means by a procedural rule is a rule which, whether technically procedural or substantive in character, has the effect of barring a claim for reasons which do not go to its legal merits; that is to say, rules which do not define the existence or extent of any legal obligation.

State immunity is a mandatory rule of customary international law which defines the limits of a domestic court’s jurisdiction. Unlike diplomatic immunity, which the modern law treats as serving an essentially functional purpose, state immunity does not derive from the need to protect the integrity of a foreign state’s governmental functions or the proper conduct of inter-state relations. It derives from the sovereign equality of states. Par in parem non habet imperium. In the modern law the immunity does not extend to acts of a private law character. In respect of these, the state is subject to the territorial jurisdiction of the forum in the same way as any non-state party. In Germany v Italy: Greece Intervening (Jurisdictional Immunities of the State) [2012] ICJ Rep 99, at para 59, the International Court of Justice observed that the rule

The International Court of Justice has characterised state immunity as procedural: Democratic Republic of the Congo v Belgium (Arrest Warrant of 11 April 2000) [2002] ICJ Rep 3, paras 59-61; Germany v Italy: Greece Intervening (Jurisdictional Immunities of the State) [2012] ICJ Rep 99, at paras 92-97. This is correct, but state immunity is not procedural in the sense that the organisation and practices of the courts are procedural. It is procedural in the same sense as that concept has been used in the case law of the European Court of Human Rights. In other words, it requires the court to dismiss the claim without determining its merits. But it leaves intact the claimant’s legal rights and any relevant defences, which remain available, for example, to be adjudicated upon in the courts of the state itself.

This gives rise to difficulty in a case where the rule goes to the court’s jurisdiction. Proceedings brought against a state entitled to immunity are not a nullity. But the court’s jurisdiction to entertain the proceedings is limited to examining the basis on which immunity is asserted and determining whether it applies. As the International Court of Justice put it in Jurisdictional Immunities of the State (para 60), the question whether the acts relied upon are such as to attract immunity must be determined “before that jurisdiction can be exercised, whereas the legality or illegality of the act is something which can be determined only in the exercise of that jurisdiction.” The impleaded state may consent to the proceedings. Where, however, it does not consent, there is no jurisdiction to proceed to the adjudicatory stage. The court must dismiss the claim. As Lord Bingham observed in Jones v Ministry of the Interior of the Kingdom of Saudi Arabia[2007] 1 AC 270, at para 14, article 6 cannot confer on a court a jurisdiction which it does not have, and a state cannot be said to deny access to its courts if it has no access to give. In Holland v Lampen-Wolfe [2000] 1 WLR 1573, Lord Millett had put the matter in this way, at p 1588:

The Strasbourg court has dealt with many cases involving claims to state immunity since it first grappled with these matters in Waite and Kennedy v Germany(2000) 30 EHRR 261. Although the reasoning has been somewhat modified over the years, its position has remained constant. Notwithstanding the difficulty pointed out in the preceding paragraph, it has always treated article 6 as being engaged by a successful claim to state immunity. But it has applied the Convention in the light of article 31(3) of the Vienna Convention of the Law of Treaties, which requires an international treaty to be interpreted in the light of (inter alia) any relevant rules of international law. Against that background, it has always held that the proper application of the rule of state immunity was justifiable because it was derived from a fundamental principle of international law. The only cases in which it has ever held article 6 to have been violated are those in which it has found that a claim to state immunity was unfounded in international law.

It is convenient to start with three judgments delivered on the same day by a similarly constituted Grand Chamber: Al-Adsani v United Kingdom(2001) 34 EHRR 11, McElhinney v Ireland (2001) 34 EHRR 13 and Fogarty v United Kingdom(2001) 34 EHRR 12.

In Al-Adsani, the applicant had been barred by state immunity from proceeding in England against the government of Kuwait in an action claiming damages for torture. The Court held (para 48) that article 6 was engaged, because “the grant of immunity is seen not as qualifying a substantive right but as a procedural bar on the national courts’ power to determine the right.” It rejected the submission of the British government (para 44) that article 6 could not extend to matters which under international law lay outside the jurisdiction of the state. However, it held that the bar was justifiable, for reasons stated at paras 54-56:

McElhinney v Ireland (2001) 34 EHRR 13 arose out of a claim against the British government in the courts of Ireland for psychological injury arising from an incident at the border with Northern Ireland. The Court rejected the allegation that by upholding the assertion of immunity the Irish court had violated article 6, in language substantially identical to that employed in Al-Adsani.

Fogarty v United Kingdom(2001) 34 EHRR 12 was the first of a number of cases to come before the Strasbourg court involving employment disputes between a state and non-diplomatic staff at one of its embassies. It concerned a sex discrimination claim brought in England against the United States by a secretary employed at US embassy in London. Once again, the Court held in substantially identical language that article 6(1) was engaged but not violated. The importance of the decision for present purposes lies in the additional observations which the Court addressed specifically to diplomatic employment disputes. The Court said at para 37 that:

The first case in which the European Court of Human Rights held that the recognition of state immunity violated article 6(1) of the Convention was Cudak v Lithuania(2010) 51 EHRR 15. The applicant was a secretary and switchboard operator employed in the Polish embassy in Vilnius, Lithuania. The Supreme Court of Lithuania’s decision appears to have been closely based on the Strasbourg court’s decision in Fogarty. It had upheld Poland’s claim to state immunity on the ground that:

Some further explanation is called for concerning the Strasbourg Court’s treatment of the ILC’s draft articles of 1991, since it is criticised by Ms Karen Steyn QC, for the Secretary of State on grounds that I think misunderstand it. The Court began its observations on this question by noting (para 64) that “the application of absolute state immunity has, for many years, clearly been eroded.” This is a reference to the progressive adoption of the restrictive doctrine. The court treated draft article 11 as reflecting the adoption of the restrictive doctrine in the domain of employment. As regards the critical parts of draft article 11, this is plainly correct. The exceptions which were relevant in Ms Cudak’s case were (a) and (b). Of these, (a) directly imported the classic distinction between acts jure imperii and acts jure gestionis. As to (b), the International Law Commission’s commentary on the draft articles suggested that it “confirmed the existing practice of states” by which state immunity extended to the recruitment, renewal of employment and reinstatement of an employee, these being dependent on policy considerations lying within a state’s discretionary power and likely to have been determined as an exercise of governmental authority. A substantial body of domestic case law from various jurisdictions is cited in support of this statement: see Report of the International Law Commission on the work of its forty-third session, 29 April-19 July 1991 [A/46/10], pp 43-44, para (10). The Strasbourg court presumably based its reasoning on the draft articles of 1991 rather than the final text of the Convention because the relevant proceedings in Lithuania occurred in 2000 and 2001, before the final text of the Convention was adopted. But although the final text of article 11 differs in significant respects from the draft article, exception (a) is substantially the same in the final version, and exception (b) (renumbered (c)) is identical. The Court was therefore right to regard these provisions of draft article 11 as applying the restrictive doctrine of state immunity to contracts of employment, and as foreshadowing, in that respect, the terms of the Convention. I do not read the Strasbourg Court as having assumed that everything else in draft article 11 was declaratory of existing customary international law. It did not need to, because the other exceptions in article 11(2) did not affect the issue. Ms Cudak appears to have been a national of or habitually resident in Lithuania, and there was no contractual submission to the local forum. Exceptions (c), (d) and (e) therefore did not arise.

Sabeh El Leil v France(2011) 54 EHRR 14 arose out of another unfair dismissal claim, this time brought in the French courts by the head of the accounts department of the Kuwaiti embassy in Paris. In this case, the final decision of the French courts barring the claim on grounds of state immunity was handed down after the adoption of the United Nations Convention. After reiterating the principle on which the Strasbourg court had always held article 6 to be engaged in such cases, the Grand Chamber summarised its case law as follows (paras 51-52):

The reasoning in Cudak and Sabeh el Leil was subsequently applied by the Strasbourg court in Wallishauser v Austria (Application 156/04, Judgment of 19 Nov 2012) and Radunović v Montenegro (Applications 45197/13, 53000/13 and 73404/13, Judgment 25 Oct 2016), all of them cases involving technical and administrative staff of a foreign embassy.

The Court of Appeal in the present cases thought that it was “questionable” whether article 11 of the draft articles was in fact a definitive statement of customary international law in embassy employment disputes. For my part, I would agree that some of the Strasbourg court’s observations about article 11 have simply served to sow confusion. Article 11 codifies customary international law so far as it applies the restrictive doctrine to contracts of employment. That would have been enough for Ms Cudak’s and Mr El Leil’s purposes. So far as article 11 goes beyond the application of the restrictive doctrine, its status is uncertain. I shall expand on this point below. It would perhaps have been better if the Strasbourg court had simply said that employment disputes should be dealt with in accordance with the restrictive doctrine instead of in accordance with an article of a treaty which is not in force and which a large majority of states have neither signed nor ratified. But this is a point of presentation, not of substance.

Ms Steyn for the Secretary of State has raised a threshold issue. She contends that a decision of a domestic court that a state is entitled to immunity does not engage article 6 at all, because its effect is that there is no jurisdiction capable of being exercised and no access to a court capable of being withheld. As I have pointed out, this is a point which was powerfully made in the House of Lords in Holland v Lampen-Wolfe [2000] 1 WLR 1573 and Jones v Saudi Arabia[2007] 1 AC 270, but was rejected by the Grand Chamber in Al-Adsani. In Jones v United Kingdom(2014) 59 EHRR 1, a chamber of the European Court of Human Rights was invited to depart from Al-Adsani on this point, but it declined to do so, adhering to its long-standing distinction between procedural and substantive bars to the exercise of jurisdiction. Ms Steyn now invites us to resolve this issue in accordance with the views of the House of Lords. In my view, there may well come a time when this court has to choose between the view of the House of Lords and that of the European Court of Human Rights on this fundamental question. But the premise on which the question arises is that there is a rule of international law which denies the English court jurisdiction in the instant case. In both Jones and Lampen-Wolfe, the Appellate Committee had satisfied itself that there was. I would not be willing to decide which of the competing views about the implications of a want of jurisdiction is correct, unless the question actually arose. So the first question which I shall address is what is the relevant rule of international law.

To identify a rule of customary international law, it is necessary to establish that there is a widespread, representative and consistent practice of states on the point in question, which is accepted by them on the footing that it is a legal obligation (opinio juris): see Conclusions 8 and 9 of the International Law Commission’s Draft Conclusions on Identification of Customary International Law (2016) [A/71/10]. There has never been any clearly defined rule about what degree of consensus is required. The editors of Brownlie’s Principles of Public International Law, 8th ed (2012), 24, suggest that “complete uniformity of practice is not required, but substantial uniformity is”. This accords with all the authorities. In the words of the International Court of Justice -

In view of the emphasis placed by the European Court of Human Rights on the United Nations Convention and its antecedent drafts, it is right to point out that a treaty may have no effect qua treaty but nevertheless represent customary international law and as such bind non-party states. The International Law Commission’s Draft Conclusions on Identification of Customary International Law (2016) [A/71/10], propose as Conclusion 11(1):

The Secretary of State’s case is that there is no sufficient consensus on the application of state immunity to a contract for the employment of non-diplomatic staff of a foreign diplomatic mission, to found any rule of customary international law on the point. He submits that two consequences follow from this. The first is that article 6 of the Human Rights Convention is satisfied if the rule of the forum state “reflects” generally recognised principles of international law. For this purpose, it is enough for the forum state to apply a “tenable” view of what international law is, or at any rate that its domestic law applies a solution that is not outside the “currently accepted international standards” treated as a benchmark in Fogarty. He submits that it is not necessary to show that international law “requires” the foreign state to be treated as immune. The second consequence is said to be that in the absence of a rule of customary international law justifying some other solution, the state employer is entitled to absolute immunity. This is because, in the Secretary of State’s submission, the restrictive doctrine of state immunity operates by grafting exceptions onto the principle of absolute immunity, so that unless and until a relevant exception has achieved the status of customary international law, the immunity remains unqualified.

I can deal quite shortly with the suggested distinction between “reflects” and “requires”, for in my opinion it is misconceived. The argument is based on the observation of the European Court of Human Rights in Al-Adsani (para 56) that “measures taken by a High Contracting Party which reflect generally recognised rules of public international law” are within a state’s margin of appreciation. That observation is repeated in most of the subsequent cases: see Fogarty (para 36), Cudak (para 57), Sabeh El Leil (para 49). But in my view the distinction proposed by the Secretary of State is a purely semantic one. International law is relevant to the operation of article 6 of the Human Rights Convention because, in accordance with article 31(3)(c) of the Vienna Convention on the Law of Treaties, the Human Rights Convention is interpreted in the light of “any relevant rules of international law applicable in the relations between the parties.” It is therefore necessary to ask what is the relevant rule of international law by reference to which article 6 must be interpreted. The relevant rule is that if the foreign state is immune then, as the International Court of Justice has confirmed in Jurisdictional Immunities of the State, the forum state is not just entitled but bound to give effect to that immunity. If the foreign state is not immune, there is no relevant rule of international law at all. What justifies the denial of access to a court is the international law obligation of the forum state to give effect to a justified assertion of immunity. A mere liberty to treat the foreign state as immune could not have that effect, because in that case the denial of access would be a discretionary choice on the part of the forum state: see Al Jedda v United Kingdom(2011) 53 EHRR 23; Nada v Switzerland(2012) 56 EHRR 593, paras 180, 195; Perincek v Switzerland(2016) 63 EHRR 6, paras 258-259. To put the same point another way, if the legitimate purpose said to justify denying access to a court is compliance with international law, anything that goes further in that direction than international law requires is necessarily disproportionate. I conclude that unless international law requires the United Kingdom to treat Libya and Sudan as immune as regards the claims of Ms Janah and Ms Benkharbouche, the denial to them of access to the courts to adjudicate on their claim violates article 6 of the Human Rights Convention.

There are circumstances in which an English court considering the international law obligations of the United Kingdom may properly limit itself to asking whether the United Kingdom has acted on a “tenable” view of those obligations. A suggestion to this effect by Sir Philip Sales and Joanne Clement, “International law in domestic courts: the developing framework” (2008) 124 LQR 388, 405-407 was tentatively endorsed by Lord Brown of Eaton-under-Heywood in R (Corner House Research) v Serious Fraud Office[2009] 1 AC 756, at para 68. Thus the court may in principle be reluctant to decide contentious issues of international law if that would impede the executive conduct of foreign relations. Or the rationality of a public authority’s view on a difficult question of international law may depend on whether its view of international law was tenable, rather than whether it was right. Both of these points arose in Corner House. Or the court may be unwilling to pronounce upon an uncertain point of customary international law which only a consensus of states can resolve. As Lord Hoffmann observed in Jones v Saudi Arabia (para 63), “it is not for a national court to ‘develop’ international law by unilaterally adopting a version of that law which, however desirable, forward-looking and reflective of values it may be, is simply not accepted by other states.” But I decline to treat these examples as pointing to a more general rule that the English courts should not determine points of customary international law but only the “tenability” of some particular view about them. If it is necessary to decide a point of international law in order to resolve a justiciable issue and there is an ascertainable answer, then the court is bound to supply that answer. In the present cases, the law requires us to measure sections 4(2)(b) and 16(1)(a) against the requirements of customary international law, something that we cannot do without deciding what those requirements are.

I do not read the Strasbourg court as having said anything very different in Fogarty. The court considered (para 37) that although there had been a “trend” in favour of the restrictive doctrine of state immunity, there was too much diversity of state practice in the specific area of embassy staff to enable them to say that the restrictive doctrine applied to them. In those circumstances they thought it sufficient that the United Kingdom had acted on a view of international law which, although not the only possible one, was within “currently accepted international standards”. But this is not the same point as the one made by the Secretary of State, for it applies only if there is no relevant and identifiable rule of international law. If there is such a rule, the court must identify it and determine whether it justifies the application of state immunity. That is what the Strasbourg court did in Cudak and Sabeh El Leil, and what it criticised the Lithuanian Supreme Court and the French Court de Cassation for failing to do in those cases. For reasons which I shall explain, I find the view expressed in Fogarty that there was no relevant and identifiable rule of international law surprising, but that is another matter.

The fundamental difference between the parties to this appeal concerns the starting point. On the footing that customary international law must require the United Kingdom to treat Libya and Sudan as immune, the Secretary of State submits that it does. This is because in his submission state immunity is absolute unless the case is brought within an internationally recognised exception to it. This submission, if it is correct, would considerably broaden the scope of state immunity in customary international law, by extending it to any group of claimants about whom there was a diversity of state practice. But in my view, it is not correct. The rule of customary international law is that a state is entitled to immunity only in respect of acts done in the exercise of sovereign authority. In the absence of a special rule to some different effect applicable to employees in the position of Ms Janah and Ms Benkharbouche, that is the default position.

It is true that the State Immunity Act 1978 adopts the drafting technique of stating a presumptive immunity subject to exceptions. Section 1 provides that a state is immune except as provided in the following sections of Part I. The same drafting technique is employed in other national legislation, especially in common law jurisdictions, for example the United States, Canada and Australia. As applied to international law, the submission is lent a certain superficial plausibility by the fact that the United Nations Convention has adopted the same drafting technique. Article 5 provides for a general immunity “subject to the provisions of this Convention”. In Jones v Saudi Arabia, Lord Bingham relied on the way that the United Nations Convention was drafted as showing that a state was immune in respect of everything that was not the subject of an express exception, and concluded that the immunity extended to torture because torture was not the subject of any express exception: see paras 8-9 (Lord Bingham), and cf para 47 (Lord Hoffmann).

I do not regard these considerations as decisive of the present issue. No one doubts that as a matter of domestic law, Part I of the State Immunity Act is a complete code. If the case does not fall within one of the exceptions to section 1, the state is immune. But the present question is whether the immunity thus conferred is wider than customary international law requires, and that raises different considerations. In the first place, it is necessary to read the grant of the immunity in article 5 of the United Nations Convention together with the exceptions which follow, as an organic whole. The exceptions are so fundamental in their character, so consistent in their objective and so broad in their effect as to amount in reality to a qualification of the principle of immunity itself rather than a mere collection of special exceptions. Secondly, it is important when doing this to distinguish between a drafting technique and a principle of law. The travaux préparatoires of the United Nations Convention show that the technique of stating a general rule of immunity subject to exceptions, was highly contentious. This was partly because it might be taken as an implicit recognition that absolute immunity was the basic rule, something which many states did not accept; and partly because it was thought that it would lead to undue rigidity and thereby impede the future development of customary international law. These differences are summarised in the Report of the International Law Commission on the Work of its thirty-eighth session [A/41/10] ILC Yearbook (1986), ii(2), 16, and in the valuable commentary of O’Keefe and Tams (ed), The United Nations Convention of Jurisdictional Immunities of States and their Property (2013), 99-101. It is clear that the draftsman’s objective was to remain neutral as between the competing doctrines said to represent the current state of international law. Various proposals were made with a view to achieving this and avoiding undue rigidity. In particular, it was proposed that what became article 5 should provide that a state was immune “subject to the provisions of the present articles [and the relevant rules of general international law applicable in the matter].” This provoked much discord, and the bracketed words were ultimately dropped on the ground that they made no difference. In its commentary on the draft articles of 1991, the International Law Commission explained that this was because

The main difficulty about the Secretary of State’s submission is a more fundamental one, namely that it is not consistent with the way that the law of state immunity has developed. Unlike diplomatic immunity, which is based mainly on an international consensus established by writers and governmental practice over many centuries, state immunity was developed during the nineteenth and twentieth centuries primarily by municipal courts. In the words of the Special Rapporteur of the International Law Commission, presenting in 1980 the first draft of what became the United Nations Convention, their decisions constituted “a great and divergent volume of municipal jurisprudence”: [A/35/10] ILC Yearbook (1980), ii(2), 143. Before the age of state trading organisations, there were few occasions for testing the limits of state immunity. States rarely did acts in peacetime within the territory of other states, other than conduct diplomatic relations, and that was the subject of a distinct and far older international law immunity. Leaving aside the rather special case of the immunity of personal sovereigns visiting the forum state, the only other acts which a sovereign performed in the territory of another state involved the presence of state-owned ships in its ports or the placing of public procurement contracts. The latter were generally for military or diplomatic purposes, and were therefore closely related to the inherently governmental acts of the state, even if they were strictly speaking acts of a private law character.

The earliest notable landmark was the judgment of the Supreme Court of the United States in The Schooner Exchange v McFaddon 11 US 116 (1812), delivered by Chief Justice Marshall. The Exchange was a trading vessel belonging to two American merchants, which was captured at sea by French ships of war, converted into an armed cruiser and incorporated into the French navy. When the ship put in to the port of Philadelphia, its former owners claimed possession. The Supreme Court held that a ship of war in the possession of a foreign state was immune from any proceedings in rem. Marshall CJ founded the rule on an implied exception to the territorial sovereignty of states for certain classes of act done there by a foreign state, which was based on “the usages and received obligations of the civilised world”. At pp 144-145, he drew a distinction between trading vessels and ships of war.

The same assumption was made in the earliest English cases. In Duke of Brunswick v King of Hanover (1848) 2 HL Cas 1 the House of Lords held, in the words of Lord Chancellor Cottenham (p 17), that

The question whether a corresponding immunity applied to a sovereign’s non-sovereign acts arose for the first time in England in The Charkieh (1872-5) LR 4 A & E 59, a collision action brought against a ship which belonged to the Khedive of Egypt and flew the flag of the Ottoman navy, but was employed for ordinary commercial purposes and at the time was under charter to a British trading house. Sir Robert Phillimore, sitting in the Admiralty Court, held that the vessel was not immune because the Khedive was not a sovereign but an officer of the Ottoman Porte. However, he went on to hold that there would have been no immunity in any event, because the use of state property for trading purposes was an implicit waiver of any immunity attaching to the state. At pp 99-100, he stated that

Nonetheless, the Parlement Belge was for many years regarded as authority for the absolute immunity of state property. The extreme point of this tendency was reached with the decision of the Court of Appeal in The Porto Alexandre [1920] P 30. In that case there was no suggestion that the vessel was in use for any public purpose. She was engaged in ordinary trading operations. But the Parlement Belge was treated as warranting the absolute immunity of state-owned ships from actions in rem. The decision provoked controversy well before the Privy Council held in The Philippine Admiral[1977] AC 373 that it had been wrongly decided. In Compania Naviera Vascongada v Steamship Cristina (The Cristina)[1938] AC 485, a Spanish trading ship had been requisitioned by the Spanish government while on the high seas in order to assist the republican government of Spain to put down the nationalist rebellion. Possession of her had then been taken in the port of Cardiff by the Spanish consul there. The speeches need to be read in conjunction with the fuller account of the facts and arguments which are reported at (1938) 60 Lloyd’s Rep 147. It was not in doubt that the Spanish consul had taken possession of it for public purposes. The real issues were whether an action in rem against a state-owned ship impleaded the foreign state; and whether the English courts should recognise an extraterritorial decree of the Spanish state. The House of Lords rejected the argument about the extraterritorial operation of the decree, and dismissed the action on the ground (i) that an action in rem against a state-owned ship indirectly impleaded the state, or indeed (per Lord Wright, at p 505) directly impleaded it; and (ii) that however she had previously been employed by her owners, she was intended for public purposes in the hands of the Spanish government. An action for possession could not therefore proceed. The interest of the case for present purposes lies in the divergence of views about the Porto Alexandre. Lord Atkin and Lord Wright considered that the immunity of states was absolute and applied irrespective of the purpose for which a ship was in the state’s possession. But the other members of the Appellate Committee doubted this, primarily on the ground that it could be correct only if there was a sufficient international consensus to that effect. However, no attempt had been made in the earlier cases to establish that there was. Lord Thankerton observed (pp 495-496) that it “may be argued” that the judgment of Brett LJ in the Parlement Belge did not authorise the extension of state immunity to property in commercial use since proceedings against such property were “not to be regarded as inconsistent with the independence and equality of the state represented by such owner.” He pointed out that the Court of Appeal in the Porto Alexandre had “made no inquiry as to whether such an exemption was generally agreed to by the nations, and it seems to be common knowledge that they have not so agreed.” Lord Macmillan shared these doubts, remarking at p 498:

In this uncertain state of English law, Lord Simon, delivering the advice of the Privy Council in Sultan of Johore v Abubakar Tunku Aris Bendahar[1952] AC 318, 343, observed:

The doubts expressed in The Cristina by Lord Thankerton, Lord Macmillan and Lord Maugham about the international law basis for a rule of absolute immunity were justified, as a review of pre-1945 decisions in different jurisdictions demonstrates. For what follows, I am indebted to the extensive reviews of this large body of material by Sir Hersch Lauterpacht in his influential article “The Problem of Jurisdictional Immunities of Foreign States”, 28 BYIL (1951), 220, 250-272, by the German Bundesverfassungsgericht in Claim against the Empire of Iran (1963), Entscheidungen des Bundesverfassungsgerichts, 16 (1964), 27 (partially translated in 45 ILR 257), and by Fox, The Law of State Immunity, 3rd ed (2013), Ch 6, and Dunbar, “Controversial Aspects of Sovereign Immunity in the Case-law of some States”, (1971) 132 Recueil des Cours, 197.

Broadly speaking, these show that states which adopted the absolute doctrine of state immunity generally did so on one or other of two grounds. One was that the sovereign equality of states implied an entire absence of jurisdiction by the courts on one state over another. The other was that while there was in principle a distinction between the public and private acts of a state, the distinction should depend on the state’s purpose in doing the relevant act and not on its juridical character, so that even trading activities were immune if they were carried on in the public interest. The two approaches are very different but in practice they lead to the same result, except perhaps in the case of the private acts of personal sovereigns.

As far as the common law world is concerned, the English courts, after a period of hesitation, finally opted for the first analysis. In British dependencies and dominions, the absolute doctrine of state immunity was generally adopted in line with what was assumed to have been laid down in The Parlement Belge. In the United States, the absolute doctrine had a more chequered history, but it ultimately adopted the second analysis. The State Department’s traditional approach to the question of state-owned ships was described in a communication addressed by the Secretary of State to the Attorney General in 1918, stating that “where [state-owned] vessels were engaged in commercial pursuits, they should be subject to the obligations and restrictions of trade, if they were to enjoy the benefits and profits.” Instructions to this effect were given to United States diplomatic and consular officers abroad: see Hackworth, Digest of International Law, ii (1941), 429, 439-440. This only changed with the decision of the United States Supreme Court in Berizzi Brothers Co v Steamship Pesaro, 271 US 562 (1926), an action in rem in support of a cargo claim against a trading vessel owned by the Italian state. The State Department had refused, in accordance with its traditional practice, to certify that the ship was immune, on the ground that “vessels owned by a state and engaged in commerce are not entitled, within the territorial waters of another state, to the immunity accorded to vessels of war, and that notwithstanding such ownership these vessels are subject to the local jurisdiction to the same extent as other merchant vessels”: Hackworth, op cit, ii, 437. But the Supreme Court upheld the claim to immunity. The Court adopted The Parlement Belge, as it had been interpreted in subsequent English case law, including The Porto Alexandre. It accepted in principle the distinction between ships operated for public and private purposes which dated back to The Schooner Exchange v McFaddon. But it largely emptied it of substance by applying it according to the state’s purpose in doing the act. As Van Devanter J put it, at p 574,

The position in civil law countries was highly diverse. In France, the absolute doctrine was endorsed by the Cour de Cassation in its celebrated decision in Lambège et Pujol v Etat d’Espagne, 22 Jan 1849, Dalloz (1849), i, 5. But the principle was not consistently applied in its absolute form, and a series of decisions in the first half of the 20th century appeared to recognise a distinction between the public and private law functions of states: see Dunbar, art cit, 212-218. The absolute doctrine was consistently applied until recent times in Spain, Portugal and Japan, but less consistently in Germany and the Netherlands and not at all in Switzerland. It has never been recognised in Italy or Belgium, whose highest courts were among the first to adopt the restrictive doctrine in a recognisably modern form. As early as 1886, the Italian Corte di Cassazione justified its position by observing:

Looking at the position in the years immediately following the second world war, Sir Hersch Lauterpacht concluded that the common assumption that the majority of states were wedded to the doctrine of absolute state immunity was inaccurate. On the contrary, “in the great majority of states in which there is an articulate practice on the subject, courts have declined to follow the principle of absolute immunity” (pp 250-251). It followed, that “so far as the actual practice of states may be said to be evidence of customary international law, there is no doubt that the principle of absolute immunity forms no part of international custom” (p 221). Thirty-five years later the International Law Commission, reporting to the United Nations General Assembly on the difficulties which it had encountered in formulating a basic principle of state immunity, expressed the same view:

The story of the progressive adoption of the restrictive doctrine of state immunity in the past 70 years is well known and can be shortly summarised. The main impetus for this was the growing significance of state trading organisations in international trade. The critical moment was the formal adoption (or readoption) of the restrictive doctrine by the United States government in the Tate Letter, addressed by the legal adviser to the State Department to the Acting Attorney General on 19 May 1952. After reciting the adoption of the restrictive doctrine by a growing number of states, it stated the intention of the executive to act on it. “The widespread and increasing practice on the part of governments of engaging in commercial activities,” it observed, “makes necessary a practice which will enable persons doing business with them to have their rights determined in the courts”: loc cit, 985. Following the Tate Letter, the restrictive doctrine was generally adopted by Federal Courts, a development which was ultimately approved by the Supreme Court in Alfred Dunhill of London Inc v Republic of Cuba, 425 US 682, 701-703 (1976). In Europe, the main landmark was the adoption by the German Bundesverfassungsgericht of the restrictive theory in 1963, and its acceptance that the distinction between acts jure imperii and jure gestionis depended on the juridical character of the act, not the purpose of the state in doing it: Claim against the Empire of Iran(1963) 45 ILR 257. The courts of the United Kingdom, followed suit in the 1970s. Today, the international consensus in favour of the restrictive doctrine is almost complete. While there are a few states whose domestic position is unclear, with the legislative adoption of the restrictive doctrine by Russia in 2015, the only notable state still to adhere to the absolute doctrine is China.

Three points can be derived from this history. The first is that there has probably never been a sufficient international consensus in favour of the absolute doctrine of immunity to warrant treating it as a rule of customary international law. All that can be said is that during certain periods, a substantial number of states, but not necessarily a majority, have adopted the absolute doctrine as part of their domestic law. Some of them have done so on the assumption that it represented international law, but without any real investigation of the rule recognised in other states. Secondly, while there has for at least two centuries been a consensus among nations in favour of some form of state immunity, the only consensus that there has ever been about the scope of that immunity is the consensus in favour of the restrictive doctrine. Thirdly, the adoption of the restrictive doctrine has not proceeded by accumulating exceptions to the absolute doctrine. What has happened is that governments, courts and writers of authority have been prompted by the widening scope of state operations and their extension into commerce and industry, to re-examine the true basis of a doctrine originally formulated at a time when states by and large confined their operations in other countries to the classic exercises of sovereign authority. The true basis of the doctrine was and is the equality of sovereigns, and that never did warrant immunity extending beyond what sovereigns did in their capacity as such. As Lord Wilberforce put it in The I Congreso del Partido[1983] 1 AC 244, 262,

As a matter of customary international law, if an employment claim arises out of an inherently sovereign or governmental act of the foreign state, the latter is immune. It is not always easy to determine which aspects of the facts giving rise to the claim are decisive of its correct categorisation, and the courts have understandably avoided over-precise prescription. The most satisfactory general statement is that of Lord Wilberforce in The I Congreso del Partido, at 267:

In the great majority of cases arising from contract, including employment cases, the categorisation will depend on the nature of the relationship between the parties to which the contract gives rise. This will in turn depend on the functions which the employee is employed to perform.

The Vienna Convention on Diplomatic Relations divides the staff of a diplomatic mission into three broad categories: (i) diplomatic agents, ie the head of mission and the diplomatic staff; (ii) administrative and technical staff; and (iii) staff in the domestic service of the mission. Diplomatic agents participate in the functions of a diplomatic mission defined in article 3, principally representing the sending state, protecting the interests of the sending state and its nationals, negotiating with the government of the receiving state, ascertaining and reporting on developments in the receiving state and promoting friendly relations with the receiving state. These functions are inherently governmental. They are exercises of sovereign authority. Every aspect of the employment of a diplomatic agent is therefore likely to be an exercise of sovereign authority. The role of technical and administrative staff is by comparison essentially ancillary and supportive. It may well be that the employment of some of them might also be exercises of sovereign authority if their functions are sufficiently close to the governmental functions of the mission. Cypher clerks might arguably be an example. Certain confidential secretarial staff might be another: see Governor of Pitcairn v Sutton(1994) 104 ILR 508 (New Zealand Court of Appeal). However, I find it difficult to conceive of cases where the employment of purely domestic staff of a diplomatic mission could be anything other than an act jure gestionis. The employment of such staff is not inherently governmental. It is an act of a private law character such as anyone with the necessary resources might do.

This approach is supported by the case law of the European Court of Human Rights, which I have already summarised. In Cudak, Sabeh El Leil, Wallishauser and Radunović, all cases concerning the administrative and technical staff of diplomatic missions, the test applied by the Strasbourg Court was whether the functions for which the applicant was employed called for a personal involvement in the diplomatic or political operations of the mission, or only in such activities as might be carried on by private persons. In Mahamdia v People’s Democratic Republic of Algeria (Case C-154/11) [2013] ICR 1, para 55-57, the Court of Justice of the European Union applied the same test, holding that the state is not immune “where the functions carried out by the employee do not fall within the exercise of public powers.” The United States decisions are particularly instructive, because the Foreign State Immunity Act of the United States has no special provisions for contracts of employment. They therefore fall to be dealt with under the general provisions relating to commercial transactions, which have been interpreted as confining state immunity to exercises of sovereign authority: see Saudi Arabia v Nelson 507 US 349, 360 (1993). The principle now applied in all circuits that have addressed the question is that a state is immune as regards proceedings relating to a contract of employment only if the act of employing the plaintiff is to be regarded as an exercise of sovereign authority having regard to his or her participation in the diplomatic functions of the mission: Segni v Commercial Office of Spain 835 F 2d 160, 165 (7th Cir, 1987), Holden v Canadian Consulate 92 F 3d 918 (9th Cir, 1996). Although a foreign state may in practice be more likely to employ its nationals in those functions, nationality is in itself irrelevant to the characterisation: El-Hadad v United Arab Emirates 216 F 3d 29 (DC Cir, 2000), at 4, 5. In Park v Shin 313 F 3d 1138 (9th Cir, 2002), paras 12-14, it was held that “the act of hiring a domestic servant is not an inherently public act that only a government could perform”, even if her functions include serving at diplomatic entertainments. A very similar principle has been consistently applied in recent decisions of the French Cour de Cassation: Barrandon v United States of America, 116 ILR 622 (1998), Coco v Argentina 113 ILR 491 (1996), Saignie v Embassy of Japan 113 ILR 492 (1997). In the last-named case, at p 493, the court observed that the employee, a caretaker at the premises of the mission, had not had “any special responsibility for the performance of the public service of the embassy.”

I would, however, wish to guard against the suggestion that the character of the employment is always and necessarily decisive. Two points should be made, albeit briefly since neither is critical to this appeal.

The first is that a state’s immunity under the restrictive doctrine may extend to some aspects of its treatment of its employees or potential employees which engage the state’s sovereign interests, even if the contract of employment itself was not entered into in the exercise of sovereign authority. Examples include claims arising out of an employee’s dismissal for reasons of state security. They may also include claims arising out of a state’s recruitment policy for civil servants or diplomatic or military employees, or claims for specific reinstatement after a dismissal, which in the nature of things impinge on the state’s recruitment policy. These particular examples are all reflected in the United Nations Convention and were extensively discussed in the preparatory sessions of the International Law Commission. They are certainly not exhaustive. In re Canada Labour Code[1992] 2 SCR 50, concerned the employment of civilian tradesmen at a US military base in Canada. The Supreme Court of Canada held that while a contract of employment for work not involving participation in the sovereign functions of the state was in principle a contract of a private law nature, particular aspects of the employment relationship might be immune as arising from inherently governmental considerations, for example the introduction of a no-strike clause deemed to be essential to the military efficiency of the base. In these cases, it can be difficult to distinguish between the purpose and the legal character of the relevant acts of the foreign state. But as La Forest J pointed out (p 70), in this context the state’s purpose in doing the act may be relevant, not in itself, but as an indication of the act’s juridical character.

The second point to be made is that the territorial connections between the claimant on the one hand and the foreign or forum state on the other can never be entirely irrelevant, even though they have no bearing on the classic distinction between acts done jure imperii and jure gestionis. This is because the core principle of international law is that sovereignty is territorial and state immunity is an exception to that principle. As the International Court of Justice observed in Jurisdictional Immunities of the State, at para 57, the principle of state immunity

At the time when the State Immunity Act was enacted, the application of state immunity to contracts of employment had only lately emerged as a potential problem. States had traditionally recruited the staff of diplomatic and representative missions at home. The employment of locally recruited staff in significant numbers was a recent development. The European Convention on State Immunity was one of the first international instruments to make special provision for contracts of employment, which would otherwise have fallen to be dealt with under the general principles of customary international law relating to state immunity.

There was, however, no consistency of state practice capable of founding a special rule of customary international law governing employment. This was recognised during the preparatory sessions of the International Law Commission relating to jurisdictional immunities of states. The working group reviewing the Commission’s draft articles of 1991 observed in 1999:

The travaux leading to article 11 of the United Nations Convention contain no suggestion that existing state practice supported a special rule of international law concerning employment claims, extending beyond the immunity attaching to sovereign acts. On the contrary, it is clear from both the travaux themselves and the impressive body of legal materials assembled by the parties to this appeal that, while many states assert a special jurisdiction over employment disputes extending to the employees of foreign states, there is considerable diversity in this area. The ILC’s Special Rapporteur reported in 1983 (ILC Yearbook (1983), ii(1), 34-8 [A/CN.4/363]) that:

The result is that the State Immunity Act 1978 can be regarded as giving effect to customary international law only so far as it distinguishes between exercises of sovereign authority and acts of a private law character, and requires immunity to be conferred on the former but not the latter. There is no basis in customary international law for the application of state immunity in an employment context to acts of a private law character.

Under the terms of the Act, contracts of employment are excluded from the ambit of section 3, which applies the distinction between sovereign acts and acts of a private law character to other contracts for the supply of services. Section 4 by comparison identifies those contracts of employment which attract immunity by reference to the respective connections between the contract or the employee and the two states concerned. In principle, immunity does not attach to employment in the local labour market, ie where the contract was made in the United Kingdom or the work fell to be performed there: see section 4(1). However, this is subject to sections 4(2)(a) and (b), which are concerned with the employee’s connections by nationality or residence with the foreign state (section 4(2)(a)) or the forum state (section 4(2)(b)). Section 4(2)(a) extends the immunity to claims against the employing state by its own nationals. As I have said, this may have a sound basis in customary international law, but does not arise here. Section 4(2)(b) extends it to claims brought by nationals or habitual residents of third countries. Both subsections apply irrespective of the sovereign character of the relevant act of the foreign state.

Sections 4(2)(a) and (b) are derived from article 5.2(a) and (b) of the European Convention on State Immunity. Like section 4 of the Act, article 5 of the Convention deals with contracts of employment without reference to the distinction between acts jure imperii and jure gestionis which are the basis of the restrictive doctrine of immunity. Contractual submission apart, the availability of state immunity in answer to employment claims is made to depend entirely on the location of the work and the respective territorial connections between the employee on the one hand and the foreign state or the forum state on the other. The explanatory report submitted to the Committee of Ministers of the Council of Europe justified this on the ground that “the links between the employee and the employing State (in whose courts the employee may always bring proceedings), are generally closer than those between the employee and the State of the forum.”

The United Kingdom is not unique in applying this principle. Seven other European countries are party to the European Convention on State Immunity and six other countries have enacted legislation containing provisions similar to section 4(2) of the United Kingdom Act. But this is hardly a sufficient basis on which to identify a widespread, representative and consistent practice of states, let alone to establish that such a practice is accepted on the footing that it is an international obligation. The considerable body of comparative law material before us suggests that unless constrained by a statutory rule the general practice of states is to apply the classic distinction between acts jure imperii and jure gestionis, irrespective of the nationality or residence of the claimant. Indeed, the courts of a significant number of jurisdictions have refused to apply the immunity as between states which are not both party to the Convention, unless they performed functions directly related to the exercise of the state’s sovereign authority, on the ground that the requirements of general international law differed on this point from those of the Convention: see French Consular Employee Claim (1989) 86 ILR 583 (Supreme Court, Austria); British Consulate-General in Naples v Toglia (1989) 101 ILR 379, 383-384 (Corte de Cassazione, Italy); De Queiroz v State of Portugal, 115 ILR 430 (1992) (Brussels Labour Court, Belgium, 4th Chamber); M v Arab Republic of Egypt (1994) 116 ILR 656 (Federal Tribunal, Switzerland); Muller v United States of America 114 ILR 512, 517 (1998) (Regional Labour Court, Hesse); X v Saudi School in Paris and Kingdom of Saudi Arabia,127 ILR 163 (2003) (Cour de Cassation, France - note the observations of the Advocate-General at p 165); A v B Oxf Rep Int L (ILDC 23) (2004) (Supreme Court, Norway); Kingdom of Morocco v HAYearbook of International Law (2008), 392 (Court of Appeal of the Hague, Netherlands).

I conclude that section 4(2)(b) of the State Immunity Act 1978 is not justified by any binding principle of international law.

The Secretary of State has an alternative argument to the effect that section 4(2)(b) may be justifiable as an application of purely domestic policy, on the ground that the United Kingdom’s interest in asserting the jurisdiction of its own courts over the employment of the local labour force does not extend to nationals or residents of third countries. I reject this argument. On the footing that international law does not require a state to be given immunity, I do not see how the absence of British nationality or residence at the time of the contract can be a proper ground for denying an employee access to the courts in respect of their employment in the United Kingdom. They have no territorial connection with their employer, other than that which is implicit in the employment relationship itself. The fact that they may have had no connection with the United Kingdom either before they came to work here does not prevent them from being part of the domestic labour force afterwards. Nor do I accept that the only relevant interest for this purpose is that of the United Kingdom state. The forum state has duties as well as rights, and as a matter of domestic policy they extend to the protection of those lawfully living and employed in the United Kingdom.

Since section 16(1)(a) extends state immunity to the claims of any employee of a diplomatic mission, irrespective of the sovereign character of the employment or the acts of the state complained of, it is plain that it cannot be justified by reference to any general principle of immunity based on the restrictive doctrine. It could be justified only if there were a special rule, in effect an absolute immunity, applicable to embassy staff. I have already pointed out, in the context of section 4(2)(b), that in jurisdictions where the courts determine claims to immunity by reference to customary international law, the test is whether the relevant acts of the state were exercises of sovereign authority. The analysis need not be repeated here. It is inconsistent with any suggestion that immunity can attach to all embassy staff as such.

The Secretary of State submits that there is indeed a special rule applicable to embassy staff. He says that such a rule is implicit in the international obligations of the United Kingdom under the Vienna Convention on Diplomatic Relations, the European Convention on State Immunity, and the state of customary international law reflected in the United Nations Convention. The Vienna Convention on Diplomatic Relations has been ratified by almost every state in the world and may for practical purposes be taken to represent a universally binding standard in international law. Article 7 provides that a sending state may “freely appoint” members of the staff of a diplomatic mission. The staff referred to include the technical, administrative and domestic staff as well as the diplomatic staff: see article 1. The argument is that the freedom to appoint embassy staff must imply a freedom to dismiss them. Article 32 of the European Convention on State Immunity and article 3.1 of the United Nations Convention both provide that they are not to prejudice the privileges and immunities of a state in relation to the exercise of the functions of its diplomatic missions and persons connected with them. In my opinion, however, article 7 of the Vienna Convention has only a limited bearing on the application of state immunity to employment claims by embassy staff. I would accept that the right freely to appoint embassy staff means that a court of the forum state may not make an order which determines who is to be employed by the diplomatic mission of a foreign state. Therefore, it may not specifically enforce a contract of employment with a foreign embassy or make a reinstatement order in favour of an employee who has been dismissed. But a claim for damages for wrongful dismissal does not require the foreign state to employ any one. It merely adjusts the financial consequences of dismissal. No right of the foreign state under the Vienna Convention is infringed by the assertion of jurisdiction in the forum state to carry out that adjustment. Therefore, no right under the Vienna Convention would be prejudiced by the refusal of the forum state to recognise the immunity of the foreign state as regards a claim for damages.

The closest that any international instrument has come to providing for a general immunity of states as regards claims by embassy staff is article 11.2(b) of the United Nations Convention. The article provides, so far as relevant:

In general, article 11 adheres to the restrictive doctrine, confining the immunity in employment disputes to cases where the making of the contract or the acts giving rise to the complaint were exercises of sovereign authority, or the dispute is between a state and one of its own nationals. Article 11.2(b) of the United Nations Convention lists four categories of employee whose claims will attract immunity. The first three categories are diplomatic or consular staff whose functions would normally be regarded as inherently governmental. But the fourth category comprises “any other person enjoying diplomatic immunity.” Under the Vienna Convention on Diplomatic Relations, all members of the staff of a mission who are not nationals of or permanently resident in the receiving state enjoy diplomatic immunity, including (in respect of acts performed in the course of their duties) domestic staff: see article 37(3). On the face of it, therefore, this provision applies state immunity to all claims by embassy staff at whatever level and irrespective of the juridical character of the acts giving rise to the dispute. The Court of Appeal, adopting a suggestion in O’Keefe and Tams (ed), The United Nations Convention of Jurisdictional Immunities of States and their Property. A Commentary (2013), 201-2, have held that it could not have meant this in the light of the travaux preparatoires. These do not explain how article 11.2(b) came to assume its final form. But they do show that the working groups and committees of the International Law Commission intended to limit the immunity to the employment of diplomatic agents. It was suggested to us that sub-paragraph (iv) might also have been intended to cover diplomats at international conferences, and there are passages in the travaux which support that view: see, in particular, Report of the ILC Working Group on Jurisdictional Immunities of States, ILC Yearbook (1999), ii(2), para 105. But since both of these categories are already covered by article 11(2)(b)(i) and (iii), and the language of (iv) is unequivocal, I doubt whether these suggestions can be supported. It is, however, unnecessary to decide the point, because it is in my view clear that if article 11(2)(b)(iv) means what it says, it is legislative rather than declaratory of existing international law. It may one day bind states qua treaty. It may come to represent customary international law if and when the Convention attracts sufficient support. But it does not do either of these things as matters presently stand.

There are judicial decisions in which the court, while limiting the immunity to exercises of sovereign authority, has taken an expansive view of the range of acts relating to an embassy employee which can be so described. Sengupta v Republic of India[1983] ICR 221 was a decision of the Employment Appeal Tribunal under the common law in force before the passing of the State Immunity Act 1978. The Tribunal held that state immunity attached to a claim for the unfair dismissal of an employee of the Indian High Commission in London. He was employed at what Browne-Wilkinson J, delivering the judgment of the court, described (p 223) as “the lowest clerical level”. He was essentially responsible for collating press cuttings. The tribunal’s reasons appear from pp 228-229 of the judgment:

I have already pointed out that in treating article 11 as expressing customary international law, the European Court of Human Rights had in mind those parts of article 11 which reflected the restrictive doctrine. In all of the cases in which it has held the recognition of immunity to violate article 6 of the Human Rights Convention, the applicant appears to have been a national or permanently resident in the forum state. The applicant did not therefore enjoy diplomatic immunity and neither article 11(2)(b)(iv) nor article 11(2)(e) arose for consideration.

Since I have concluded that no principle of international law deprived the Employment Tribunal of jurisdiction in these cases, it follows that the United Kingdom had jurisdiction over Libya and Sudan as a matter of international law, and article 6 is engaged by its refusal to exercise it. The jurisdictional issue raised by Lord Millett in Holland v Lampen-Wolfe and by Lord Bingham and Lord Hoffmann in Jones v Saudi Arabia does not arise.

The employment of Ms Janah and Ms Benkharbouche were clearly not exercises of sovereign authority, and nothing about their alleged treatment engaged the sovereign interests of their employers. Nor are they seeking reinstatement in a way that would restrict the right of their employers to decide who is to be employed in their diplomatic missions. As a matter of customary international law, therefore, their employers are not entitled to immunity as regards these claims. It follows that so far as sections 4(2)(b) or 16(1)(a) of the State Immunity Act confer immunity, they are incompatible with article 6 of the Human Rights Convention.

Ms Janah’s case that the discriminatory character of section 4(2)(b) of the Act is a violation of article 14 of the Human Rights Convention, read in conjunction with article 6, adds nothing to her case based on article 6 alone. Section 4(2)(b) unquestionably discriminates on grounds of nationality. The only question is whether the discrimination is justifiable by reference to international law. If state immunity is no answer to the claim under article 6 alone, then it is no answer to the claim under the combination of article 6 and article 14. In my view, the denial of access to the courts to persons in her position is unjustifiable whether it is discriminatory or not.

Article 47 provides, so far as relevant, that:

I would dismiss the Secretary of State’s appeal and affirm the order of the Court of Appeal. The result is that sections 4(2)(b) and 16(1)(a) of the State Immunity Act 1978 will not apply to the claims derived from EU law for discrimination, harassment and breach of the Working Time Regulations. Subject to any question as to the application of section 4(2)(b) to the particular circumstances of Ms Benkharbouche, the other claims (failure to provide payslips or a contract of employment, unpaid wages, failure to pay the national minimum wage and unfair dismissal) are barred by those sections of the Act. But to that extent they are incompatible with article 6 of the Human Rights Convention, and also, in the case of section 4(2)(b) with article 6 read with article 14 of the Convention. Both cases must be remitted to the Employment Tribunal to determine the claims based on EU law on their merits.