The EU’s Human Rights Obligations in Relation to Policies with Extraterritorial Effects: Rejoinder to Enzo Cannizzaro

This post is a continuation of the EJIL:Debate! in EJIL vol. 25: 4 between Lorand Bartels and Enzo Cannizzaro on “The EU’s Human Rights Obligations in Relation to Policies with Extraterritorial Effects.”

I thank Professor Cannizzaro warmly for his thoughtful reply to my article, which extends it in several interesting directions. Professor Cannizzaro’s main point is that Article 3(5) and Article 21 TEU do not add anything to the EU’s existing human rights obligations insofar as they apply to conduct with mere extraterritorial effects. In some respects I agree with him. However, I would argue that there are some gaps that these provisions plug.

I agree, first of all, that there is some conduct with mere extraterritorial effects that is already covered by the EU’s fundamental rights obligations. Most obviously, there is a Soering-inspired obligation not to remove a person to a third country in which he or she would be at risk of harm (NS v. Secretary of State for the Home Department). But this is a special case, and I do not think that it follows, as does Professor Cannizzaro, that other conduct with mere effects in third countries is also covered. There is to my mind something particular about the EU’s (or a state’s) obligations in relation to a person that is at the relevant time on its territory, and this cannot so easily be translated into an obligation to prohibit exports of death penalty drugs or not to adopt an economic embargo on third states.

At a more general level, the problem is that in practice the high level of fundamental rights protection applicable domestically cannot automatically be extended to policies with mere extraterritorial effects. As my article details, with the exception of Soering scenarios, there have not been any CJEU decisions on whether fundamental rights obligations apply to measures with mere extraterritorial effects, and while the European Court of Human Rights has dealt with this, it has done so very inconsistently (Cf the contrasting decisions in Kovačič (ECtHR, admissibility, 9 Oct 2003) and Ben El Mahi (ECtHR, 11 Dec 2006)). My suspicion is that the CJEU will avoid the problem of extending domestic levels of protection to measures with extraterritorial effects not by applying a dual standard, but rather by not applying fundamental rights obligations to measures in the first place. On my reading, this is supported (or at least not contradicted) by the cases analysed in Moreno-Lax and Costello, ‘The Extraterritorial Application of the EU Charter of Fundamental Rights: From Territoriality to Facticity, the Effectiveness Model’ in Steven Peers et al (eds), The EU Charter of Fundamental Rights – A Commentary (Oxford: Hart, 2014), a chapter to which for timing reasons it was unfortunately not possible to refer in my article. This is why, in my view, Article 3(5) and especially Article 21 TEU have a gap to fill. Indeed, there could also be a reverse effect, in that this legal possibility could also make it less important to apply fundamental rights obligations to such measures.

Professor Cannizzaro takes a different approach to the EU’s fundamental rights obligations. He would allow that these apply to measures with mere extraterritorial effects, but subject to certain inherent limits. He states that the EU’s acts are limited by international rules on extraterritorial jurisdiction, the implication being that measures that would breach these rules would either not be adopted, or, if adopted, would fall at this hurdle. I am less certain that this is a real constraint. In the first place, in my view these rules do not significantly limit measures with mere human rights effects in other countries; to the contrary, they might enable such measures (See, further, Bartels, ‘Article XX of GATT and the Problem of Extraterritorial Jurisdiction: The Case of Trade Measures for the Protection of Human Rights’ (2002) 36 Journal of World Trade 353, 368-76). In practice, these rules are irrelevant to the scope of human rights obligations (Marco Milanovic, Extraterritorial Application of Human Rights Treaties (Oxford: OUP, 2011), 23). And even if these rules did apply to such measures, the CJEU has a relatively non-standard approach to what these rules mean, so they are unlikely to limit EU measures in any case (Scott, ‘The New EU “Extraterritoriality”’ (2014) 51 Common Market Law Review 1343).

Professor Cannizzaro also argues that to the extent that measures with mere extraterritorial effects are subject to fundamental rights obligations, in practice such measures will only have ‘indirect’ effects on human rights and will therefore not be affected by such obligations. He states that:

[T]he EU can hardly be held responsible for violations of human rights occurring beyond its borders as an indirect consequence of conduct performed by its Institutions within its boundaries. Neither under Soering and its progeny, nor under the general law of international responsibility, can a violation of human rights be attributed to an entity when such violation, although prompted by the initial conduct of that entity, constitutes the ultimate and unforeseeable product of a chain of events unfolding outside its control.

I am also not certain about this. If one takes the ECHR jurisprudence standard as a guide, at least sometimes the causal standard is one of reasonable foreseeability, particularly in the context of positive obligations (Stubberfield, ‘Lifting the Organisational Veil: Positive Obligations of the European Union Following Accession to the European Convention on Human Rights’ (2012) 19 Australian International Law Journal 117). Indeed, I would go so far as to suggest that if not Soering itself then perhaps its progeny (Eg NS v Secretary of State for the Home Department [2011] ECR I–13905) could be explained by reference to such a standard, and so could the other examples of measures (eg an embargo) that Professor Cannizzaro considers already to be covered by the EU’s fundamental rights obligations. More generally, while it is by no means easy, it is also far from impossible to establish a direct causal link between, for example, subsidies and negative effects on the economic and social welfare of persons in third countries. This is not uncommon in the WTO, where, incidentally, all that is required is that act be a ‘genuine and substantial cause’ of the injury suffered (WTO Appellate Body Report, EC and Certain Member States – Aircraft (Airbus), WT/DS316/AB/R, para 1234).

I am therefore not convinced that these limits to the application of EU fundamental rights obligations to measures with mere extraterritorial effects are as realistic as Professor Cannizzaro proposes. The result, for me, is that it is unlikely that these obligations will be apply to such measures in the first place, and this opens up a gap in protection that, in my view, Articles 3(5) and especially Article 21 TEU can fill.

I also differ from Professor Cannizzaro insofar as I think that these two provisions do something original. Our main difference in this respect concerns Article 21 TEU, both in relation to its material scope and its normative effects.

As to its scope, Professor Cannizzaro states that Article 21 TEU is limited to the Common Foreign and Security Policy (CFSP). This is because ‘Article 23 [a CFSP provision] assigns the pursuit of the political objectives laid down by Article 21(1) and (2) to the primary competence of the CFSP’. But while the CFSP ‘shall pursue the objectives’ set out in Article 21, it is in my view a non sequitur to infer that the CFSP can be the only means by which these objectives are to be pursued. Nor do I this that the CFSP is capable of being the only means of pursuing these objectives, given that these objectives (set out in Article 21(2)) go well beyond foreign policy, extending to development, trade and environmental policies. Professor Cannizzaro rightly observes that Article 40 TEU ring-fences the CFSP from the EU’s other policies, but it also ring-fences the EU’s other policies from the CFSP. Finally, I would point out that Article 23(3)(1) states that:

The Union shall respect the principles and pursue the objectives set out in paragraphs 1 and 2 in the development and implementation of the different areas of the Union’s external action covered by this Title and by Part Five of the Treaty on the Functioning of the European Union, and of the external aspects of its other policies. (emphasis added)

If Article 21 were limited to the CFSP, this reference to the EU’s non-CFSP policies would seemingly make no sense.

A more particular disagreement concerns the normative features of Article 21(3)(1) TEU, just quoted. According to Professor Cannizzaro, this provision ‘indicates that the objectives of Article 21(1) and (2) will be pursued through “the development and implementation of the different areas of the Union’s external action […] and of the external aspects of its other policies”.’ But being required to ‘respect … principles’ that include ‘democracy, the rule of law, the universality and indivisibility of human rights and fundamental freedoms, respect for human dignity’ is not the same thing as being required to pursue these ‘objectives’. Rather, this imposes on the EU an obligation to respect inter alia human rights in third countries in relation to the policies mentioned. My article goes into this point in more detail.

Finally, I differ from Professor Cannizzaro on a point not covered in the article, namely, the implications of Article 3(5) and 21 TEU for the EU’s competences. Professor Cannizzaro states that ‘[i]n spite of the broad set of objectives laid down by Articles 3(5) and 21, these two provisions do not confer new competences on the EU.’ This is technically true, but in my article I add two qualifications. The first is that Article 3(5) TEU is now covered by the EU’s implied competence in Article 352 TFEU, the ‘flexibility clause’. While Article 352(4) confirms that Article 352 ‘cannot serve as a basis for attaining objectives pertaining to the common foreign and security policy’, Declaration 41 states that Article 352 TFEU can serve as a basis for measures pursuing the objectives set out in Article 3(5) TEU. Moreover, in practice, Article 352 TFEU has supported external human rights measures, and the proposition that it can has received academic support (de Witte, ‘The EU and International Legal Order: The Case of Human Rights’ in Panos Koutrakos and Malcolm Evans (eds), Beyond the Established Legal Orders: Policy Interconnections between the EU and the Rest of the World (Oxford: Hart, 2011), 137; Piet Eeckhout, EU External Relations Law, 2nd ed (Oxford: OUP, 2011), 99). The second is the proposition that every obligation binding on the EU comes with an implied functional power to do what is necessary to comply with that obligation (For discussion and further references, see Lorand Bartels, Human Rights Conditionality in the EU’s International Agreements (Oxford: OUP, 2005), 214-219. Cf a similar argument in Cannizzaro, ‘The Scope of the EU Foreign Power: Is the EC Competent to Conclude Agreements with Third States Including Human Rights Clauses?’ in Enzo Cannizzaro (ed), The European Union as an Actor in International Relations (The Hague: Kluwer, 2002), 316-9, where Professor Cannizzaro proposes that the EU’s international human rights obligations result in an inherent (or implied) power to enforce compliance with those obligations by other actors). So both Article 3(5) and Article 21 TEU can affect the EU’s existing competences, even if strictly speaking they do not confer on it any new competences.

I would like to conclude by expressing my sincere gratitude to Professor Cannizzaro for taking the time to engage with my article and for raising these points, and I look forward to further discussion.