In a decision of November 2022, the Swedish Labour Court held that Qatar had, at least in part, the right to invoke state immunity from jurisdiction in connection with proceedings relating to an employment matter.
An employee at the Qatari embassy in Stockholm was summarily dismissed in September 2021. He filed a lawsuit against the State of Qatar in the Stockholm District Court requesting invalidation of, and compensation for, unfair dismissal. Qatar objected to Swedish jurisdiction by invoking State immunity under international law.
In a preliminary decision, the Stockholm District Court granted Qatar State immunity regarding the invalidation claim, but denied it with respect to the claim for compensation. The Labour Court, which was the court of appeal and last instance in the matter, agreed with the District Court’s conclusion.
In its decision, the Labour Court held that the dispute had such a “close connection” to Sweden that Swedish courts should be able to assert jurisdiction, unless Qatar had the right to invoke state immunity. The court continued by stating that whether or not state immunity from jurisdiction should be respected is dependent on customary international law. In this regard, the court held that the 2004 United Nations Convention on Jurisdictional Immunities of States and Their Property is representative of customary international law.
Article 11 of the 2004 UN Convention deals with employment law matters. The provision contemplates several exceptions to State immunity in this field, i.e., situations where a State, as a defendant before the courts of another State, are not entitled to immunity.
Article 11(2)(c) sets out that States are immune from jurisdiction whenever “the subject-matter of the proceeding is the recruitment, renewal of employment or reinstatement of an individual”. Although this provision does not explicitly deal with the invalidation of a dismissal, the Swedish Labour Court concluded that such invalidation is equivalent to reinstatement. With references to the European Court of Human Rights’ (ECtHR’s) judgments Naku v. Lithuania and Sweden and Cudak v. Lithuania, the Labour Court found that granting State immunity does not amount to a violation of the employee’s right to a fair trial under Article 6 in the European Convention of Human Rights (ECHR).
In its examination of the ECtHR’s judgments, the Labour Court noted that Naku had requested both reinstatement and damages and that the ECtHR had found that it was wrong to grant State immunity in that case. The Labour Court noted that the ECtHR made a point that an employee under Lithuanian law could be given severance pay instead of reinstatement. There is no possibility for a court to do so under Swedish law when a former employee has pleaded invalidation of an employer’s dismissal. Therefore, the Labour Court concluded that granting State immunity for Qatar was not a violation of Article 6 in the ECHR.
The Swedish Labour Court’s decision is, in my opinion, problematic and can be criticized for at least three reasons. First, the Labour Court makes no references to EU private international law. Second, the Labour Court’s formalistic approach to public international customary law is erroneous. Third, the Labour Court only – selectively – referred to ECtHR case law on the balance of state immunity and the right to a fair trial. In the following, I will develop those three points of critique.
EU Private International Law Determines Jurisdiction
Whether or not a Swedish court has jurisdiction in an international employment law matter follows from the Brussels I bis Regulation.
Under Article 1, the Regulation shall apply in “civil and commercial matters.” This concept has, on several occasions, been interpreted by the Court of Justice of the European Union (CJEU) as including actions by public authorities, as long as they do not exercise public powers. Further, it is clear that this also extends to embassy employment cases (see, e.g., ZN, C-280/20, paras 26–28, and Mahamdia, C-154/11, para 56). Even under the old Brussels I Regulation, it was clear that the regulation was applicable to third countries’ embassies, as they are to be considered “establishments” (Mahamdia, p. 41). Under the Brussels I bis Regulation, Article 6 has been extended to apply to all matters in which an employee habitually carries out his or her work in the EU, regardless of where the employer is domiciled. In such cases, the Regulation must be applied in the EU Member States (Roi Land Investments, C-604/20 para 48).
With this in mind, it is misleading that the Swedish Labour Court seemingly applies the jurisdictional standard of “Swedish interest of adjudication” that is used against defendants domiciled in third countries, without mentioning the Brussels I bis Regulation.
The 2004 UN Convention Does Not Reflect International Customary Law
The Labour Court oversimplifies customary international law by referring mechanically to the 2004 UN Convention.
Customary international law is defined by its constantly evolving nature and the fact that it must be derived from state practice. Therefore, international customary law cannot ever really be codified. The convention may serve as one of many pieces of public international customary law, but cannot be applied mechanically.
When using the 2004 convention, one shall bear in mind that there were some controversies about the content of the convention when it was adopted and that it still has not gathered enough support to enter into force. Further, since the convention was drafted nearly 20 years ago, important case law development has been made by i.a. both the CJEU and the ECtHR regarding State immunity in embassy employment disputes. As case law from these courts is to be applied by Swedish courts, irrespective of customary international law, it is questionable that this case law was not taken into consideration. This leads me to my third point of critique: that the Labour Court did not correctly refer to relevant EU and ECtHR case law.
The Labour Court Did Not Correctly Refer to Relevant ECtHR and CJEU Case law
Over the last 15 years, the ECtHR has developed a unanimous approach to State immunity in embassy employment law cases. This approach is made clear in the 2019 judgment Ndayegamiye-Mporamazina v. Switzerland. In this judgment, the ECtHR makes clear that an important feature for granting state immunity from jurisdiction is that the employee can access courts elsewhere. In Ndayegamiye-Mporamazina v. Switzerland, the employee could initiate proceedings in the embassy state of Burundi. Access to courts elsewhere in the world has long been a cornerstone in the evaluation of upholding state immunity (see Prince Hans-Adam II of Liechtenstein v. Germany). In the Swedish Labour Court’s decision, the access to courts in Qatar was not assessed at all.
In my opinion, it is a pity that the Labour Court did not pay attention to neither the EU private international law aspect nor the relevant case law developments from the ECtHR. For the development of customary international law regarding state immunity in embassy employment matters, well-motivated case law is needed. Due to the flaws described above, the Swedish Labour Court’s decision can hardly be seen as a contribution to the development of customary international law.
very interesting, as well as alarming…
Dear Erik, I fully agree that merely relying on the UN Convention to determine customary international law is simplistic. The problem is that it is exactly what the ECtHR does, and that the court finds that contracting states which do not follow article 11 are in breach of the ECHR. Contracting States are trapped, and are essentially forced to adopt the corrupt view of the Strasbourg court. And the Strasbourg virus is spreading throughout Europe: France, and I understand also Belgium and Italy …
Dear Gilles, thank you for your comment. I think that the ECtHR goes a little further than the 2004 Convention by e.g. stressing the relevance of a possible court procedure elsewhere.
National courts could perhaps avoid the ‘trap’ by relying on Art. 53 ECHR? It is true that the ECtHR uses Art. 11 UNCSI as a yardstick to assess the compatibility of state immunity with Art. 6 ECHR, but that is only a minimum standard. Should a national court refuse to apply an Art. 11 immunity ground because it does not reflect custom, the decision would be more favorable to the employee and thus fully compatible with the ECHR. I do realize it is unlikely for these subtleties to gain traction in domestic case law. It is much easier to simply parrot the ECtHR.
If a national court offers a better protection by going further than Art. 11, I also trust the ECtHR would not find a violation. Most judgments which have insisted that a national court had not followed Art. 11 and was in breach of Art 6 also insisted that the national court had not given reasons for that. So, an interesting question is whether it would change something if a national court explained that Art 11 does not reflect customary international law. This is what Lord Sumption did in Benkharbouche v Secretary of State for Foreign and Commonwealth Affairs  UKSC 62, and the ECtHR has indeed blinked in Benkharbouche and Janah v. UK:
[the ECtHR] does not consider it necessary to itself examine the substantive issues raised by the applicants’ complaints, or to resolve any potential differences between the Supreme Court’s view as to what was required by customary international law (see also the view taken by the Court of Justice of the European Union in Ahmed Mahamdia v. People’s Democratic Republic of Algeria, summarised in paragraph 39 above) and the view expressed by the Court in its case-law (see, for example, Cudak, Sabeh El Leil, Wallishauser, Oleynikov, Radunović and Others, Naku, all cited above, and Ndayegamiye-Mporamazina v. Switzerland, no. 16874/12, 5 February 2019).
For an example of said Belgian case law, one may refer to the judgment of the Belgian Court of cassation of 4 march 2019, which held that “Il résulte du droit coutumier international en matière d’immunité de juridiction tel qu’il figure également à l’article 11, alinéas 1er et 2, a, de la Convention des Nations Unies du 2 décembre 2004 sur les immunités juridictionnelles des États et de leurs biens, que, lorsqu’une affaire concerne un contrat de travail entre un autre État et une personne physique et que cet autre État se prévaut de l’immunité de juridiction, cet État ne peut invoquer cette immunité que si un certain nombre de critères sont remplis, notamment l’engagement afin de s’acquitter de fonctions particulières « dans l’exercice de la puissance publique », à savoir les « acta iure imperii ».” (translation in French available here: https://juportal.be/content/ECLI:BE:CASS:2019:ARR.20190304.3/FR , the original court language of the case is Dutch)
Frédéric Dopagne commented rather aptly on the judgment, stating that “l’identification du régime coutumier d’immunité est largement centrée sur la Convention, qui se trouve d’emblée mobilisée à cette fin” (Journal des tribunaux 2020, 584).
The French Supreme Court has offered the same analysis in 2019 and 2020: Art 11 reflects customary international law, and should thus be applied as such, without any additional analysis.
The result is that several higher courts in Europe are now stating that Art 11 reflects customary international law. State practice is changing because of the ECtHR. And the position of the ECtHR will soon be a self fullfiling prophecy.
At least a connection is made in the Belgian judgment between the 2004 Convention and the underlying idea of “acta iure imperii”.
This is an interesting case and an interesting debate. I would like to make the following remarks: The 2004 draft convention on state immunity appears me to be a good starting point to assess the content of customary international law in this area of law. The fact that his has not yet ratified by a sufficient number of states is not an unusual situation. Similarly, the 1969 Vienna Convention on the Law of Treaties took several years (less than its twin on immunities, I am aware) until it entered into force. Moreover, the International Court of Justice, in its 2012 judgment Germany v. Italy no 1, referred to the draft convention, too. One should not forget that the International Law Commission (ILC) worked on its elaboration during almost two decades. The aim of the convention is to assess the content of customary international law in the field. Although this text contains compromises, the 2004 draft convention appears as an authority in the field and should be used as a starting point to assess the content of customary international law.
Regarding the case law of the European Court of Human Rights on the immunities of foreign states and international organizations, I see many merits there: The Court cautiously developed a case law that respects the immunities but requires reasonable alternative means to settle the dispute (already in appl. 26083/94 White and Case v Germany, para 68). I do agree that this case law is not always consistent as the 2019 judgment in Ndayegamiye-Mporamazina v. Switzerland (appl.16874/12) demonstrates. In this case, the Strasbourg Court said that the granting of immunity does not depend on the availability of reasonable alternatives in order to settle the dispute. However, this statement appears as an obiter dictum in the case at hand, as alternative relief was available in the Burundi (the home state of the applicant). This concept appears also in article 11 (2)(e) of the 2004 draft convention on state immunity. It excludes labor disputes of nationals of the sending state with the latter from the immunity exception for labor disputes. The underlying idea, expressed by the ILC , is that these disputes should be heard before the courts of the sending state. The Swedish court even went further by permitting the claim for compensation. In my 2017 Hague Lecture, I addressed these tendencies in the case law of supranational and domestic courts, RdC 388, paras 220 ff.).
Finally, there is another aspect that seems to be interesting. The case demonstrates the (growing) tensions between global rules (state immunity) and regional approaches (found in regional human rights protection – article 6 ECHR – and regional private international and procedural law – BX 1bis). During the last decades, there is an ongoing and growing tendency of solving cross-border issues not at the global level, but by more elaborated rules at the regional level, in specific frameworks. Labor law disputes with employees of embassies fall within this category as there regional standards in Europe evolving. It is not easy for national courts to deal with these complicated conflicts between regional and global rules. However, the case law of the Strasbourg Court opens up avenues to overcome the broad concept of immunity and to strengthen the position of the individual. Unfortunately, I could not read the judgment of the Swedish court, but overall, it seems me to be in line with the general approach cautiously developed in Strasbourg.
Thank you for your comment, Burkhard!
One of the reasons for my critique on the mechanical application of the 2004 Convention for these issues, is that it blurs the rationale behind the idea of state immunity and the modern principle exceptions to that. The core idea of state immunity is that a foreign state shall not be subject to court proceedings in another state (in line with the old principle “par in parem non habet imperium”). Historically, this principle was granted for states whenever they called for it. In the modern customary international law, state immunity has been made relative. Nowadays, state immunity is granted for states when they exercise state authority (“acta iure imperii”) but not for other acts (“acta iure gestionis”) (Cf here e.g. article 1 p. 1 of the Brussels I bis Regulation). In my opinion, the 2004 Convention can play a role for distinguishing state acts from other acts, but it shall not be applied mechanically. In the case at hand, the mechanical application of the 2004 Convention leads to a somewhat inconsistent conclusion when the invalidation claim is subject to state immunity at the same time as the claim for compensation due to the same alleged unfair dismissal is not. I have a hard time to understand the rationale behind this split. Is there something that I am missing?
Dear Erik, thank you for your interesting post! This decision looks very similar to what the Italian Court of Cassation has been doing since 2014: citing the ECtHR without ever referring to state practice, opinio juris or the ILC works. I agree with you that considering the whole of Art. 11 UNCSI as a codification of custom is a gross oversimplification. It is quite clear that many of the immunity grounds it lists are not borne out by state practice (the ILC itself said so in 1991). The ECtHR overlooked that, by applying those grounds, the scope of state immunity is unduly expanded at the expense of employees. That said, I am wondering how the use of a proper methodology of custom ascertainment would have affected the outcome of this particular case. It seems to me that some provisions of Art. 11 do correspond to custom, and I would argue that Art. 11(2)(c) is one of them. Hence the outcome seems ok to me in the end of the day – I don’t know if you agree with this assessment.
I also agree that courts should factor in the availability of alternative means of redress. I would be more hesitant to take the Ndayegamiye-Mporamazina judgment as an example. After all, there the ECtHR said that state immunity does not depend on whether the employee has access to the courts of the employer state (para 64). Its remarks on the availability of the Burundian courts were made in passing and without even checking whether that was actually the case. Incredibly, those were based just on generic ‘reassurances’ on the part of Burundi. I find the ECtHR’s attitude disturbing, all the more when compared to its Waite and Kennedy jurisprudence. I cannot make sense of the disparity of treatment between employees of foreign states and international organizations: both can be left remediless by immunity, not just the latter.
Dear Pierfrancesco, as I write in my reply to Burkhard’s comment above I have a hard time to understand why one legal consequence of an alleged unfair dismissal is subject to state immunity whereas another aspect is not. I think that it shall be all or nothing in this part.
Thank you also for your comments on the Ndayegamiye-Mporamazina case. The aspects you mention must surely be taken into account. Another piece in this ECtHR puzzle that is hard to interpret because of its oddities is the recent Benkarbouce and Janah v UK judgment (Applications nos. 19059/18 and 19725/18). Here the UK agreed before the ECtHR that the applicants’ rights to fair trials were violated. I guess that position can serve as some sort of opinio iuris for customary international law purposes.
Dear Erik, thank you for your reply! I have also read your other reply above. To be fair, I am not persuaded by the ‘all or nothing’ approach. The subject-matter of the claim (claims for reinstatement barred, compensation for unfair dismissal permitted) is one of the most common yardsticks used by national courts to apply restrictive state immunity in the context of labor disputes. This is true for many jurisdictions. And indeed the provision now contained in Art. 11(2)(c) was described by the ILC commentary as ‘designed to confirm the existing practice of states’ – one of the very few openly codificatory provisions of Art. 11.
I see your point that this approach is inconsistent with the iure imperii/iure gestionis distinction. Undoubtedly so: courts draw lines between immune and non-immune claims even though the underlying state act is the same. But I don’t think that we should dismiss this practice based on preconceived notions on restrictive state immunity (ie that the distinction between public and private acts must always be determinative). Since custom emerges from state practice, the question should be: what does this practice tell us about the status of the customary law of state immunity? My answer is simply that the application of restrictive state immunity to employment does not necessarily follow the iure imperii/gestionis divide (which after all was developed in the context of commercial claims and, by all accounts, is ill-suited to labor matters). It seems to me that the test based on the subject-matter of the claim is underpinned by a functional rationale, because it shields employers from the most intrusive forms of litigation (orders for specific performance). At the same time, it seeks to strike a balance between the needs of employers and the rights of employees by giving the latter a (compensatory) remedy. If you’re interested, I developed these arguments at length in a book I wrote on immunities and employment claims (Hart Publishing 2021).
Thanks a lot Pierfrancesco!
I have put both your book and Burkhard’s above-metnioned Hague lecture on my reading list and I am much looking forward to read both books.
Why is it so ill-suited to distinguish state acts from other acts in labour law matters? I have been thinking that the assessment of the employee’s working tasks that is often emphasized in these cases is a result of this classification (see e.g. CJEU’s Mahamdia judgment [p. 56]).
The idea of shielding employer states from intrusive forms of litigation – how far can that be stretched? Can an employer state invoke state immunity from another state’s jurisdiction in any employment matter regardless of the position of the employee?
Many thanks, Erik! Of course, the test based on the tasks performed by the employee is another common way of applying restrictive state immunity in labor claims. Sometimes the two tests (tasks and subject-matter) are used together, sometimes not. However, there is virtually no agreement among domestic courts on which employees perform tasks which benefit the state’s sovereign functions (I would argue that all of them do, to some degree, hence the difficulty as I see it). To me, practice is too diverse to say that there is a bar on jurisdiction over employees performing ‘sovereign’ tasks – whatever their meaning. I read state practice as demonstrating that jurisdiction can be exercised regardless of the individual tasks of the employee, except only for personnel having diplomatic and consular status and military personnel. This is just my interpretation; it is a complex issue.
My understanding as to subject-matter is that claims for reinstatement are not admissible also in the case of low-level employees (but conversely, compensation claims would be ok for high-level administrative employees). I would also point out that, in the case of diplomatic and consular missions, reinstatement of any employee would violate the VCDR and VCCR (Arts. 7 and 19 respectively).