I. INTRODUCTION
Courts are increasingly confronted with the problem of the territorial scope of application of European Union (EU) norms. This issue arises in cases whose fact patterns present some connections to the EU and some to third States—in the parlance of private international law, cases with ‘foreign elements’. In such situations, should the judge apply EU law? Or does the case fall outside its reach? Sometimes the EU norm at stake will provide specific indications as to its geographical perimeter. The task of the judge will be made easier. Other times, however, indications will be vague, incomplete or non-existent. How the Court of Justice of the EU (CJEU) deals with the problem is the object of this article's critical analysis, which seeks to challenge dominant accounts while developing an alternative approach, at the levels of both practice and theory.
Part of the difficulty of this topic, which has attracted significant attention over the last decade, is the fact that it lies at the intersection of several fields of scholarship, each of which has considered it from a particular angle. There are those who have focused only on particular sectors, including data protection,Footnote 1 competition,Footnote 2 the environment,Footnote 3 human rights,Footnote 4 taxation,Footnote 5 or the regulation of purely private relations (such as agency or consumer contracts).Footnote 6 There are also those who have considered the issue in more general terms. Public international lawyers have examined the reach of EU law through the prism of international norms, particularly the limits on the exercise of extraterritorial jurisdiction,Footnote 7 while scholars of EU law have reflected on the particularity of the approaches found in EU legal practice,Footnote 8 and how these approaches connect with the EU's values and goals, and overall legitimacy.Footnote 9
It is perhaps remarkable that, despite the broad diversity of entry points, scholarship on the territorial scope of application of EU law reflects a strong convergence in its general orientation, how it conceives of the EU's place in relation to the wider world, and the basic spatial assumptions on which it draws. This broad position will be referred to as the ‘global actor’ view, which this article will seek to identify and problematise. In short, there are two complementary dimensions to this view. On the one hand, scholarship sympathetically conceives the EU as intervening on the international scene to promote values that it identifies as universal—thus drawing from a certain idea of the EU's legitimacy as a benevolent protagonist in global governance. The EU legal system is therefore called upon to ‘manage the effects of globalization’, as described by Gráinne de Búrca, exercising ‘global leadership’ in tackling challenges common to the world, such as the climate crisis, the rule of law, security or economic prosperity.Footnote 10 This translates into a functionalist approach to the scope of application of EU law, which is determined expansively by reference to whatever substantive goals might be at stake, which in themselves are not spatially defined. Or as put by Alex Mills: even though the EU is ‘acting locally’ (by adopting its own regulations rather than engaging in international norm-making), it is ‘thinking globally’ (as its goal is to project the effect of those regulations externally).Footnote 11 On the other hand, however, spatial limits on that external projection are brought in from outwith EU law, through the constraints set by international law on extraterritorial ‘over-reach’, which derive from basic notions of territorial sovereignty.Footnote 12 This means that the regulated activities must present some nexus to the EU—usually a degree of presence in the physical territories held by the EU Member States. The EU's global leadership will therefore only go as far as the limits of the international rule of law allow, given the perimeter formed by the collection of Member State territories.
In order to challenge this broad position, a reconstruction is offered of the CJEU's case law on this issue, which is dispersed across a number of domains, of interest to both public and private lawyers. It will be argued that the way the CJEU approaches this matter is, in fact, more heterogeneous than suggested by the ‘global actor’ model. Whilst some case law does seem to endorse this model, another vision also emerges. This alternative will be referred to as the ‘local community’ model. Here, the EU's place in the world is defined, not through its role in global governance, but by the construction and preservation of a space that it can call ‘its own’.Footnote 13 That space is not the cartographic perimeter that traditional international law scholarship imagines national territory to be;Footnote 14 not the mere physical launchpad for the EU's global interventions. It is instead a distinct space of social relations, co-constructed by EU law and shaped by the idiosyncratic values that set the EU apart from other polities. From this perspective, the EU legal system appears as a space for inclusion as much as exclusion;Footnote 15 to draw the boundaries of EU law is to determine who belongs and who does not within that particular community, irrespective of any outside limits that international law may impose.
In order to identify such an alternative construction within the case law, a disciplinary tradition that is often neglected in these debates—that of private international law, also known as conflict of laws—will be utilised. Its intellectual resources will be employed on the basis that the territoriality of EU law can be better understood, and its development refined, through the particular concepts and devices that are typical of this discipline. Private international law is traditionally concerned with the specific requirements of justice that apply to, and emerge from, cross-border social relations. Because of this different focus, private international lawyers have tended to be suspicious of the reductive notions of territoriality that public international law has traditionally been more inclined to espouse,Footnote 16 and had less trouble recognising the existence of relational spaces beyond the national State territory.Footnote 17 It is for this reason that the discipline offers, as will be argued throughout, more sophisticated tools to understand EU law as a distinct space of social relations, and operationalise the practice of boundary-drawing that this understanding entails. Thus the influence of private international law in this article is mainly methodological; the purpose is not to examine the particular norms of private international law that EU law has produced (such as the Brussels and Rome Regulations),Footnote 18 or to focus only on the regulation of purely private relations, but to examine the judicial construction of EU law's external reach through a series of key technical concepts well known to private international scholars—what various authors have labelled the ‘conflict of laws style’.Footnote 19
By re-examining this area of judicial practice through the conceptual resources of private international law, this article seeks to make a triple contribution. It aims to contribute, first of all, to the practical development of the case law—in other words, at the most mundane level, to how the territorial scope of application of EU law is concretely delineated by the CJEU, in the very many domains where this issue arises. As will be seen, the dominant paradigm entails a certain approach that prioritises functional considerations—that is, how to make EU law a better tool of (global) governance. The alternative vision that this article seeks to illuminate, which is also already present in the case law but could be further enriched through various doctrines of private international law, focuses instead on the very different question of relational belonging (what type of relations are claimed by EU law as its own, and in what manner) and integrity (how to preserve the coherence and distinctiveness of the space that emerges). Thus, the first objective is to push practice on this issue in this alternative direction, away from the deficiencies that it will be argued are inherent in the dominant functionalism.
The second contribution of this article will be, more tangentially, to debates of EU constitutional law. The global actor paradigm has been developed and promoted by authors such as de Búrca in the context of the deep disorientation suffered by the EU in the aftermath of the financial crisis, which led EU legal studies to begin looking for new sources of legitimacy for the project of European integration.Footnote 20 The global actor paradigm offers a potential solution, by highlighting the global dimension of many of today's key difficulties, while invoking the responsibility of the EU in tackling them for the benefit of humanity (and superior ability to do so vis-à-vis individual States). This article provides a counterpoint to that vision through the examination of this liminal area of legal practice, since it allows the coherence of a global vision for EU law to be questioned and presents a glimpse of a different logic for European integration—based instead, as argued also by other scholars of EU law, on the construction and maintenance of a distinct and concretely situated space of social relations, reflective of an ethos that is claimed as specific to the EU.Footnote 21 In other words, the discussion seeks to bring to light a different ‘spatial identity’ for the EU legal system–that is, how it conceives of itself in relationship to space.Footnote 22
The third and final contribution relates to international law debates about the concept of territoriality and how it relates to non-State entities such as international organisations (including the EU). These debates often remain trapped in the dichotomy between territoriality and functionality—that is, whether jurisdiction must be defined by reference to a purely physical perimeter under the exclusive control of sovereign States (the State territory) or by the pursuit of certain substantive goals (as for international organisations).Footnote 23 Thus, the recent emergence of a variety of international structures of global governance tends to be interpreted as a shift from the former to the latter—in other words, as a form of ‘de-territorialisation’.Footnote 24 This dichotomy makes it hard to give any real content to the expression ‘Union territory’, even if it features officially in the Treaty on the Functioning of the European Union (TFEU)Footnote 25 and in parts of the CJEU's case law,Footnote 26 other than by reducing it to nothing more than the collection of the territories of the Member StatesFootnote 27 or dismissing it as an ‘abuse of language’.Footnote 28 Against traditional understandings of territoriality in international law, the present critique aims to show that the EU's territory exists above and beyond those of Member States.Footnote 29 Various theoretical analyses have already insisted on the distinctiveness of European territoriality (such as Etienne Balibar, with his ‘Europe as Borderland’ thesis).Footnote 30 However, those reflections have tended to focus on the transformation of the EU's internal borders and the emergence of a space of (qualified) circulation.Footnote 31 Instead, this article will examine EU territoriality from the different vantage point of its confrontation to foreign jurisdictions—ie its external borders. In so doing, it seeks to contribute, by vindicating the neglected resources of private international law, not only to a better understanding of the spatiality of the EU's rule, but also to the critical scholarship that, over the last three decades,Footnote 32 has argued for a richer understanding of territory as but a type of ‘bounded social space’Footnote 33 (and which allows for a spatial understanding of non-State entities).Footnote 34
The article will proceed as follows. Section II will begin with an overview of the functionalist approach that we have associated with the ‘global actor’ paradigm. The presence of such an approach in certain instances of reasoning of the CJEU will be highlighted, as will be the attention it has drawn from scholars who take a sympathetic view of EU law's extraterritorial extension. A critique of the functionalist approach will then be developed, examining its underlying assumptions as to the territorial dimension of the EU's rule, the insoluble practical issues that it leads to, and the need to consider EU law alternatively as a space for inclusion and exclusion. Section III will then argue that such an alternative approach can already be found in various lines of cases, even if in a slightly underdeveloped form and in somewhat disconnected settings, which will be analysed through three key doctrines of private international law—the localisation of cross-border relations, the imperativeness of EU law, and the ordre public (or public policy) exception. From this reading emerges a more sophisticated vision of the territory of EU law—one that cannot be reduced to the physical coordinates of its acts of intervention, or the mere sum of the physical spaces under Member State sovereignty, but appears instead as a distinct space of social relations, informed and delineated by the EU's axiological commitments.
II. THE EXTRATERRITORIALITY OF EU LAW AND THE GLOBAL ACTOR PARADIGM
The characteristic features of the widespread approach will first be laid out, as developed in the case law and supported by an important scholarly trend, which views the EU as a ‘global actor’ and on this basis links the scope of application of EU law to its pursuit of universal goals, whilst subjecting it to the external constraints of public international law (A). Critical analysis will then highlight that the EU legal order is, inevitably, also a space for inclusion and exclusion—a view that should lead to a different approach to territoriality (B).
A. Extraterritorial Expansion in Pursuit of Universal Goals
In private international law, functionalism has a long pedigree. One of its most well-known versions, the so-called ‘governmental interests approach’, argues that the scope of application of a particular regulation should be derived from the ‘interests’ that motivated its adoption.Footnote 35 The example that was classically used was a statute that guarantees a certain compensation for victims of accidents. That statute can be said to be geared at protecting the interests of injured parties. From this, it may be deduced that the statute should only apply to injured parties with some meaningful spatial connection to the State who adopted the statute, such as its own residents.
Functionalism is widespread in the context of EU law, in the sense that the territorial scope of application of its various norms tends to be derived from regulatory objectives.Footnote 36 The reasoning is, however, somewhat at odds with traditional functionalist approaches within private international law. Here the targeted ‘interests’ are often not defined spatially, but as universal values. This justifies giving EU law a very expansive applicability. Any territorial connections that may trigger its application are therefore presented as tools to maximise the effectiveness of the pursuit of those values, rather than a reflection of the particular interests targeted by the EU norm at stake. Scholars therefore speak of EU law's ‘global’ or ‘extraterritorial’ reach.
Take the well-known case of Google Spain,Footnote 37 in relation to Directive 95/46/EC on personal data protection (since replaced by the the General Data Protection Regulation). Should the Directive cover the processing of data by operators based in third States, such as Google's parent company? Article 4.1 limited its scope of application to where ‘the processing is carried out in the context of the activities of an establishment of the controller on the territory of the Member State’. The Court nevertheless adopted a very wide reading of the article to rule against Google. It held that the Directive also encompasses the processing of data by operators based outside the EU, provided that operator maintains a branch or subsidiary in a Member State, even where the latter's activities are limited to marketing and publicity and thus do not participate in the data processing itself (as was the case of Google Spain, a subsidiary of the Google group). To justify both points—and this is the key aspect of the judgment for the purposes of this article—the Court looked to the objective pursued by the Directive, identified as ‘ensuring effective and complete protection of the fundamental rights and freedoms of natural persons, and in particular their right to privacy, with respect to the processing of personal data’,Footnote 38 or more simply as guaranteeing ‘effective and complete protection of data subjects’.Footnote 39 The importance of this objective is such that the territorial scope of the Directive should be construed in ‘particularly broad’ terms, in order to ‘prevent individuals from being deprived of the protection guaranteed by the directive and that protection from being circumvented’.Footnote 40 In brief, the Directive, both its substantive provisions and its territorial scope of application, must be interpreted so as to ensure a maximum level of effectiveness.
The same approach appears in other fields. The CJEU will identify the objective pursued and emphasise its importance, in order to broaden the territorial reach of EU law. In AATA, for instance, the Court examined a challenge to the applicability of EU legislation subjecting aviation to the greenhouse gas emission allowance trading regime. It was argued that the application of this regime to all flights arriving at and departing from any aerodrome situated on an EU Member State violated the prohibition of extraterritoriality under customary international law. The Court replied by noting inter alia that the objective of the Directive in question was to ‘ensure a high level of protection’ of the environment. On this basis, the EU legislator is justified in subjecting all such flights to its regulations, even for emissions outside the European airspace, however weak the territorial connection may seem.Footnote 41 Similarly, in the more recent case of Q, R, S v United Airlines concerning the Flight Compensation Regulation, applicable per its Article 3.1 to ‘passengers departing from an airport located in the territory of a Member State’, the Court concluded that it should also cover flights between airports situated in third States, where the flight is a connecting flight that originally departed from a European location. The conclusion was again reached on the basis of the Regulation's substantive goals—notably, to ensure ‘a high level of protection for passengers’.Footnote 42 A similar reasoning is seen in the Zuchtvieh-Export case (interpreting the territorial scope of application of the Regulation on Animal Transport expansively, given the aim of ‘protection of animal welfare’, and more specifically of limiting ‘as far as possible’ ‘the transport of animals over long journeys’),Footnote 43 and in European Federation for Cosmetic Ingredients (where the Court again relies on the objective relating to ‘animal protection’ to adopt a broad reading of the reach of animal testing restrictions contained in the Regulation on Cosmetic Products).Footnote 44
From a technical point of view, these cases have in common the fact that they collapse the distinction between the substance of EU law and its territorial scope of application. The goals that permeate the proper reading of the former also determine how the latter should be constructed. This means that the Court does not consider who the various EU regulations have an actual interest in regulating (or have responsibility to regulate), and who falls outside this zone of concern. It reasons as if there are no inherent limits to EU law's regulatory aspirations, which target types of activities in the promotion of particular values, rather than certain geographical spaces or concretely situated collectives that could be said to ‘belong’ within the EU. The spatial limits to EU law are instead presented as external to the EU, as they derive from public international law jurisdictional limits. The Eva Glawischnig-Piesczek case is explicit in this regard: given the ‘global dimension of electronic commerce’, EU law may potentially extend worldwide, subject only to the constraints set by international law on extraterritorial jurisdiction.Footnote 45 Within those limits—as stated in Poulsen, in relation to regulations whose object was the conservation of protected sea species—the spatial applicability of EU law should be construed as broadly as possible, ‘so as to give it the greatest practical effect’.Footnote 46
It has been pointed out, however, that public international law's jurisdictional constraints are so lax that they actually empower EU law to apply beyond its borders.Footnote 47 Thus, for instance, to go back to the AATA decision, the Court concluded that the extension of the greenhouse gas emissions trading scheme to any flights arriving or departing from an aerodrome on EU soil ‘does not infringe the principle of territoriality or the sovereignty which the third States from or to which such flights are performed have over the airspace above their territory, since those aircraft are physically in the territory of one of the Member States of the European Union and are thus subject on that basis to the unlimited jurisdiction of the European Union’.Footnote 48 From this perspective, if there is a territory of the EU, it is nothing more than the addition of the territories of the Member States. These serve as an enforcement tool, the physical launchpad for EU values and policies to project worldwide.Footnote 49 The notion of EU territory does not therefore serve to define the EU as a space for inclusion and exclusion.
It is common to interpret these cases through the lens of ‘extraterritoriality’, or ‘territorial extension’.Footnote 50 The idea is that EU law here is broadening its scope, to cover global or transnational activities, despite their weak or precarious connections to the territory of the EU (or using those weak connections as an excuse). It is not, however, enough to note that EU law ends up influencing activities beyond the territories of Member States; scholars have also insisted on the underlying universalist aspiration behind EU law's ‘global reach’.Footnote 51 Scott has noted that this aspect of EU law shows that it is ‘imbued with a strong international orientation’,Footnote 52 and reflects an ambition ‘to shape the focus and content of third country and international law’.Footnote 53 Similarly, Ryngaert has said that the broadly extraterritorial scope of application of EU law fits within ‘the classic decentralized enforcement paradigm of international law’, in that the use of specific connecting factors is no more than the means by which the EU, in the absence of centralised global governance structures, aims at ‘realizing global values and interests’.Footnote 54 The fact that there is such exclusive focus on public international law to identify any spatial limits on the regulatory action of its institutions only serves to reinforce this perception of the EU as a world or global actor. More generally, Anu Bradford has popularised the concept of the ‘Brussels effect’ to describe how EU law unilaterally sets global standards, in particular through the sheer economic weight of the internal market and the corresponding need for businesses across the world to obtain access to it through regulatory compliance.Footnote 55 She has also observed how the EU, and in particular the Commission, ‘legitimizes its strategies by claiming that its values and policies are normatively desirable and universally applicable’, so that ‘the EU's externalization of its regulatory preferences reflects altruistic purposes of a benign hegemon’.Footnote 56
B. From a Global Actor to a Space of Inclusion/Exclusion
In reality, the ‘global actor’ approach is largely unable to fulfil the general task that it has been assigned—that of delineating the perimeter of EU law. A justificatory gap is inevitably present. If the goals pursued by EU institutions were indeed universal, it would suggest that the means deployed should also know no territorial bounds. The reach of EU law should thus be ‘unconfined’,Footnote 57 since it could be deployed in any territorial configuration, and EU law would therefore be universally applicable, or at least would apply wherever it did not meet the constraints of international law or the resistance of foreign jurisdictions.
Evidently, this is not the case—EU law reveals itself to be, in practice, highly selective, even in the absence of outside constraints. This became visible in the Google v CNIL decision.Footnote 58 The Court was asked to consider whether search engines such as Google can be requested to de-reference everywhere in the world—that is, to remove the links containing personal information from all versions of the search engine, as opposed to only those specific to the Member States. Having noted that the objective of the EU data protection legislation was to ‘guarantee a high level of protection of personal data throughout the European Union’, the Court admitted that it is ‘true that a de-referencing carried out on all the versions of a search engine would meet that objective in full’,Footnote 59 and that therefore the EU legislature would be competent to impose a global de-referencing order.Footnote 60 Nevertheless, the Court also noted that the EU legislature had only intended to regulate the right to de-referencing of personal information ‘so far as the Union is concerned’, but nothing suggested that it had meant to do so ‘outside the Union’.Footnote 61
Such selectivity is, in fact, a necessary feature of EU law, or indeed of any legal order. A legal order is not only an actor on the international scene, but is also, and perhaps more fundamentally, a concretely localised community. The scope of application of its laws cannot be reduced to a matter of effective policy. Indeed, political communities are inescapably ‘bounded’—for them to qualify as such, they must lay claim to a certain ‘place’ as ‘their own’, as explained by Hans Lindahl, and thus separate an inside, shaped by the community's own values, and an outside, as a space of exclusion.Footnote 62 This applies to the notion of State territory: it is not simply the geographical area enclosed by State borders, but only one of many possible ‘instantiations’ of a larger category, that of ‘bounded social space’.Footnote 63 However, it also applies to international organisationsFootnote 64 and other forms of political communities, such as the EU.Footnote 65 The territorial scope of application of EU law thus serves to distinguish an inside from an outside—what belongs ‘inside the Union’ and what is pushed ‘outside’ of it.
It is therefore possible to reconstruct the case law in this light, even where the reasoning follows a different route. Indeed, the cases examined above can be alternatively read as establishing the sort of connections that are relevant to the EU legal order and those that are not—in effect, drawing the boundaries of the space that the EU claims as its own. Thus, the above cases can be re-interpreted as contributing to the construction of an EU territory in Saskia Sassen's sense of a ‘bounded social space’.Footnote 66 This is not meant to negate that the territorial scope of application of EU law can be usefully analysed in terms of policy-setting, or within the ‘global actor’ paradigm. It is simply to note that there is a separate dimension to the territoriality of EU law—the construction of its identity as a situated local community.
Two cases will be examined. The first is the already discussed Google Spain decision. It can be recalled that this case established that the activities of Google's search engine were subject to the Directive on data protection inasmuch as, on the one hand, they ought to be considered a form of data processing, and, on the other hand, the company possessed an establishment in Spain dedicated exclusively to the promotion and sale of advertising space. What is the basis of the decision, if not to maximise the protection of privacy and the effectiveness of the Directive? It can be argued that the decision contributes to the construction of the EU as a local community in two main ways.
First, to use Paul Schiff Berman's terminology, the decision may be seen as a ‘domestication’ of what might otherwise appear as the highly nebulous world of digital data—that is, constructing that social sphere as domestic or internal to the EU.Footnote 67 In response to common perceptions of the world of digital data as entirely devoid of any connection to an ‘earthly’ legal system and thus as floating in empty legal space, the Court signals that the activities of search engines are in fact territorially grounded. That grounding does not, however, derive from conceiving digital data as being physically present in certain geographical coordinates, but by focusing on the type of relations that are established around it. Indeed, the Court places great emphasis on the fact that Google's search engine is ultimately geared at obtaining profit. Thus, Google's advertising and marketing activities are, in fact, central to its relationship with individuals such as Mr Costeja.Footnote 68 It is because this is so that this relationship, by reason of the establishment in Spain, can be concretely grounded in the territory of EU law.
Secondly, the decision can also be said to be productive with regards to the specifics of community membership. It is asserted that, despite the physical distance between Mr Costeja in Spain and Google Inc. in California, their relation is one that is recognised as belonging in the social space claimed by the EU legal system as its own. From the perspective of the individual applicant, the Court makes it clear that he is a member of that particular community, in that he is not only entitled to its protections against the digital giants, but also that he is placed within a certain social ordering by being characterised as a private figure whose data should not therefore be of interest to the wider public.Footnote 69 From the perspective of Google, it is not only the case that the Court establishes its grounding in the EU's territory by way of its profit-making activities in Spain, but it is also notable that Google is placed in the role of an intermediary, by being required to prevent any offending websites from appearing in its search results. It is therefore ‘deputised’ the responsibility of monitoring compliance with those standards and balancing the interests of the public.Footnote 70 Its membership in the particular community of EU law thus not only entails subjection to its rules but also partaking in the responsibility of their enforcement, alongside and in parallel with national authorities.
This reconstruction of Google Spain thus exemplifies how decisions that would seem to reflect a functional calculation can in fact be read differently, by conceiving the EU as a community of belonging. When viewed in these terms, the bounded space of EU law entails not only an inside, but also an outside. Where Google Spain can be read as a story of inclusion, our second example of such a reconstruction, from the field of asylum law, is a case of exclusion—a dimension of EU law that tends to be hidden from view.Footnote 71 In the case of X & X v Belgium, a Syrian family fleeing the war had travelled to Lebanon and requested at the Belgium consulate that they be granted a temporary visa on humanitarian grounds, based on Article 25 of the EU Visa Code.Footnote 72 The dispute centred on whether EU law was indeed applicable, including the Charter of Fundamental Rights, or if it was instead a matter belonging exclusively to Belgian law. The Court opted for the latter. It insisted on the territorial boundaries of the Visa Code, which it argued aims to govern ‘applications for international protection made in the territory, including at the border, in the territorial waters or in the transit zones of the Member States, but not to requests for diplomatic or territorial asylum submitted to the representations of Member States’.Footnote 73 It also emphasised, relying on a functionalist approach, that to extend the applicability of EU law ‘would undermine the general structure of the system established by Regulation No 604/2013 [the so-called Dublin III Regulation]’.Footnote 74 Thus, rather than explicitly pushing the Syrian refugees away as outsiders, the Court has taken the more convenient route of reverting to a functional reading of the relevant provisions, arguing that the lack of favour shown towards their request is, in fact, the product of regulatory needs.
It is nevertheless obvious that the case can be usefully read through the insider/outsider distinction. This was in fact openly articulated by Advocate General Mengozzi in his conclusions. He argued that, because the Visa Code applies to requests such as the applicants’, they were entitled to the protections granted by the Charter, which applies to all situations governed by EU law, whatever the territorial connection to the EU. To put it differently, the Advocate General's position was that the EU legal system had a responsibility to prevent the outrage of the mass drownings of Syrian refugees in the Mediterranean Sea, and that it was very much a matter ‘domestic’ to it.Footnote 75 By taking the opposite stance, the Court in effect refused to accept such a responsibility. It rejected the applicants as outsiders to the EU legal system—which thus emerges as a space for exclusion as much as for inclusion. This conclusion also helps understand the appeal of approaching such issues in purely functional terms. As has been seen, a notable element of the case is that, rather than articulate exclusion as lack of membership or affiliation, it tends instead to revert to functionalism—it is all a matter of effective policy-making. The Court can therefore avoid admitting that EU law is shutting its doors to outsiders, indeed that there is such thing as an outsider. In the reasoning of the Court, the territory of the EU as a transversal space informed by the EU's constitutive values disappears like some magical kingdom and is replaced by the merely physical substrate for policy-based interventions.Footnote 76
III. THE EMERGING LOCAL TERRITORY OF EU LAW
This article has thus far addressed the tendency to approach the scope of application of EU law in terms of the ‘global actor’ paradigm. It has also been shown that such an understanding is reductive and incapable of doing away with the question of belonging—that of how the delineation of the EU's scope of application contributes to its construction as a local community. It will now be demonstrated that, in reality, the functionalist tendency is only present in some case law. Over the next three sections, an alternative trend will be examined, where the assessment of the territorial scope of application of EU law serves as an opportunity for the Court to articulate, even if somewhat timidly, an idea of the EU legal system as a concretely situated community—whose boundaries determine the territorial reach of EU law.
This is evident in three main types of arguments used by the Court, which can be usefully analysed through three well-known categories of private international law—even if the reasoning found in the case law remains original: first, those where the applicability of EU law hinges on the localisation of the relevant relationship within the EU's territory (A); secondly, those that consist of asserting the applicability of EU law as imperative to relations concretely grounded in its territory, in the face of attempts by parties to international transactions to contract out to the law of a foreign jurisdiction (B); and thirdly, those where EU law applies out of a commitment to preserve the very integrity of its axiological space, which can be analogised to the ordre public exception in private international law (C). From these three lines of cases emerges a different, more sophisticated vision of the EU's territory: not a merely physical substrate, but a restricted space of social relations shaped through its particular values.
A. Localising a Legal Relationship within the EU Territory
The first group of cases to be examined are those where the CJEU takes what will be referred to as a ‘localisation approach’, which is the bread and butter of private international law.Footnote 77 From the latter's traditional perspective, the applicable law will be determined by identifying the centre of gravity of the type of relation at stake, eg contracts will be governed by the law of the place of performance, torts where the harm occurred, etc. In the present context, there is a slight but significant difference, since EU law proceeds by considering, not where the particular relation should be located from a universal perspective, but rather, exclusively from the EU's point of view, whether that relation finds its proper place within its own territory. It is clear, in any case, that under this approach the applicability of EU law is not framed in terms of maximising the reach of its values extraterritorially.
The first example is that of EU competition law, which is only meant to target, according to Articles 101 and 102 of the TFEU, anticompetitive conduct ‘within the internal market’. The CJEU was soon confronted with the difficult issue of where to draw the line, and has held that what matters is either whether the anticompetitive agreements are ‘implemented’ in the territory of the EU (as opposed to where they may have been formed),Footnote 78 or alternatively whether it is ‘foreseeable’ that the anticompetitive conduct in question would have an ‘immediate and substantial effect’ on that territoryFootnote 79 (the so-called ‘qualified effects’ test).Footnote 80 These decisions have de facto given EU competition law a very broad reach.Footnote 81 For the purposes of the present discussion, what matters is, however, that they may be read as an attempt at finding the centre of gravity of the relation at stake, and in so doing they define the perimeter of EU territoriality. The language used in the case law often seems to take a functionalist line, as for instance in the Wood Pulp decision, where the Court reasoned that, ‘[i]f the applicability of prohibitions laid down under competition law were made to depend on the place where the agreement, decision or concerted practice was formed, the result would obviously be to give undertakings an easy means of evading those prohibitions’.Footnote 82 Beneath the emphasis on effectiveness, however, the Court is implicitly stating that it is at the place of implementation or of their qualified effects that relations of competition are properly located (rather than using those triggers as a means to project EU law globally).
Even more prominently than in the domain of competition law, the localisation approach has been developed in the field of employment contracts, in cases presenting some connection to non-EU jurisdictions. As in the previous example, the Court does not seek to locate the relationship objectively on the world map but focuses instead on whether it has sufficient ties to the EU so as to be deemed metaphorically located on its territory. It is therefore indifferent to the possibility of other legal systems also claiming the relationship as their own.
Consider the early Prodest decision.Footnote 83 A Belgian national resident in France was hired by a French company to work temporarily in Nigeria. The case turned on whether he was entitled to maintain his affiliation with French social security during his placement abroad, and in particular whether he could rely on the provisions of Regulation 1612/68 (on freedom of movement of workers, since replaced by Regulation 492/2011). Advocate General Lenz and the Court agreed that he could but reached this conclusion through different paths. The former relied on a typically functionalist argument, by stating that the free movement of workers should be interpreted in a way that maximises its ‘effectiveness’, while underlining that the non-application of EU law would have a negative impact on the internal market, in particular by allowing temporary work agencies to discriminate against EU citizens.Footnote 84 The Court, however, preferred to rely instead on the proposition found in the well-known Walrave decision: the principle of non-discrimination applies where the ‘legal relationship … could be located within the territory of the Community’.Footnote 85
The difference with the approach described in Section II.B. is notable. In that approach, the scope of application of EU law is inextricably linked to the specific goals that it pursues; here the two are entirely separate. Indeed, the assessment of close links operates independently from the substantive considerations that led to the adoption of the EU norm in question. An independent notion of territory is invoked as an enclosing device, rather than the product of regulatory needs. To put it differently, it serves as the explanans of EU applicability, rather than the explanandum.Footnote 86 Instead of starting with a certain provision of EU law and then seeking to determine its scope of application, the localisation approach begins with a relationship (here, an employment contract) which can then be located within or outside the territory of the EU. That relationship may be connected to several jurisdictions (as in Prodest), but it has only one possible localisation—it can be either wholly inside the EU territory or wholly outside of it. This means that it makes no sense to speak of an extraterritorial application of EU lawFootnote 87—despite the connection to third States, the application of EU law is always only territorial since it follows the relationship's localisation within EU territory.
What, however, does it mean to describe a relation as ‘located’ somewhere? It cannot be understood in a physical sense. Even if the various elements of the relation may sometimes occupy a particular geographic location, the relation is what ties those elements together. Indeed, the kind of employment or competition relationships described above involve a variety of elements that find themselves both on the physical territory subject to the sovereign authority of the different Member States, and on that of other jurisdictions. The possibility of territorially locating the link between those elements, ie the relation itself, is only available if traditional conceptions of territory are abandoned,Footnote 88 and it is instead construed as essentially a social space, composed of a series of relations that the EU claims as its own. There are therefore two relations at stake, of a slightly different nature: first, the relation whose localisation is at stake, and secondly, that relation's connection with EU territory. For this reason, the more useful metaphor in describing the latter is not in terms of location, but of distance: a relation can be related more or less closely to the EU, it can stand in proximity or remoteness to it. This is exactly the CJEU's framing in Prodest: whether the employment relation is located in the EU's territory depends on whether it ‘retains a sufficiently close link’ to it.Footnote 89
The key question is, therefore, how does a relation come to be sufficiently close to EU territory to trigger the applicability of its law? What are the concrete boundaries of that social space that EU law claims as its own? Private international law has long been interested in this question through the study of ‘connecting factors’, ie the specific circumstances that tie a relation to a particular legal system and justify the applicability of its laws. A study of the concrete connecting factors that are considered relevant will reveal how that social space is delineated and where its boundaries lie. It will be argued that the CJEU usually understands this to be a matter of social integration. In other words, a relation will be metaphorically located within the EU's territory, where it is integrated within the broader sets of relations that form European societies. Three aspects will be highlighted.
Consider, first of all, the manner in which the CJEU approaches the close links assessment. In traditional private international law, the answer to this question would have required identifying the ‘centre of gravity’ of the type of relation in question, also referred to as the ‘connecting factor’. Thus, contracts could be located in the place of their performance, torts in the place where the tort was committed, etc. Contemporary private international law, however, tends to favour a more open-ended assessment. This involves looking at all relevant circumstances rather than focusing mechanically on a single factor, in order to establish ‘close connections’ or ‘proximity’.Footnote 90 Thus, the localisation of a relation necessarily requires a close examination of social realities that can only be done on a case-by-case basis. This is what the CJEU generally does, as it made clear in the Commission v Spain case: the localisation in EU territory cannot be made to depend on any ‘predetermined criteria’, but requires considering ‘various ties depending on the circumstances of the case’.Footnote 91
Secondly, also revealing is the CJEU's use of ‘effects-based’ connecting factors, the importance of which has grown in private international law over the last few decades.Footnote 92 This shows that the Court does not consider the relation at stake in social isolation, involving only the private parties directly participating in it (contrary again to a traditional private international law approach).Footnote 93 Such relations are instead inserted within broader relational spaces, and it is through their participation in such spaces that they will be deemed located in the EU territory. The example of competition law demonstrates this well. The Court's preference for implementation/qualified effects shows that the relation in that context is not reduced to the bilateral interaction between, say, the parties to a cartel. It is instead a more complex and multilateral relation involving other actors such as consumers, whose web of economic exchanges forms a market. It is for this reason that anticompetitive conduct that is performed outside the EU and by non-EU parties will nevertheless come under EU law's embrace where it participates in the particular relational space that is the internal market.
Thirdly, and finally, attention should also be paid to the importance attached by the Court to certain factors that tie the particular relation to broader collective structures. This trend can also be observed in recent developments in private international law (such as in the shift towards ‘habitual residence’ or ‘centre of interests’ as a dominant connecting factor),Footnote 94 or in the operation of the EU freedoms of movement (where the Court has also tied jurisdictional reach to an examination of social integration).Footnote 95 In the present context, the case law on employment contracts demonstrates this well. It is stated that EU law will apply to relationships that, ‘by reason either of the place where they are entered into or of the place where they take effect, can be located within the European Union’,Footnote 96 which translates into a very open examination. Consideration may be given to the nationality and place of residence of the employee, or the place of establishment of the employer, but also to other connecting factors, including the location of the account where the salary is paid,Footnote 97 the place of affiliation to a social security scheme,Footnote 98 the place where the employee is subject to income tax,Footnote 99 the place of registration of the ship of employment,Footnote 100 and the submission of the contract via a choice-of-law clause to the law of a Member StateFootnote 101 or via a choice-of-court clause to the jurisdiction of its tribunals.Footnote 102 Thus, the specific factors that are thought by the Court to be relevant to determining the territorial scope of application of EU law (paying one's taxes, social security affiliation, maintaining a bank account, continued residence, etc) reveal that it is concerned with establishing that the employee, despite accomplishing their work in the territory of a third State, remains nevertheless socially integrated in the EU, or affiliated to that particular community. Again, as in competition law, the Court is not only interested in the purely bilateral interaction between employer and employee, but also in how that interaction is integrated within the broader social space that EU law claims as its own.
There remains a difficult question. Can social integration be possible in the EU as a whole, thus triggering the applicability of EU law? If the territory of the EU is indeed constructed as essentially a social space, how can a relationship be said to be located on such a common territory rather than in that of the specific Member States?
The easiest answer would be to rely on the idea of intra-EU cross-border elements. Thus, domestic law applies to relations connected to third States and, within the EU, a single Member State and EU law governs transnational relations connected to at least two Member States. The case law on employment contracts could be said to support this idea, inasmuch as it tends to emphasise the circumstance that the employee who is performing work outside the EU has been hired in a Member State different from that of his or her nationality. Thus, the Court stated in Kik that:
a sufficiently close connection between the employment relationship in question and the territory of the European Union derives, inter alia, from the fact that a European Union citizen, who is resident in a Member State, has been engaged by an undertaking established in another Member State on whose behalf he carries on his activities.Footnote 103
Likewise, it was said in Aldewereld that:
a sufficiently close link with the Community … can be found in the fact that the Community worker was employed by an undertaking from another Member State and, for that reason, was insured under the social security scheme of that State.Footnote 104
This reading certainly seems aligned with the traditional understanding of the scope of application of EU law vis-à-vis purely internal situations (ie subject only to Member State law)—and in particular with that of the free movement of workers, which is said to benefit ‘[a]ny Community national who, irrespective of his place of residence and his nationality, has exercised the right to freedom of movement for workers and who has been employed in a Member State other than that of residence’.Footnote 105 However, the type of cases under consideration, involving relations that are partly connected to non-EU jurisdictions, might call this interpretation into question.
Consider the Petersen case, where Mr Petersen, a Danish national resident in Germany, had been employed by an undertaking established in Denmark to work in Benin for three years as part of a development aid project. The German tax authorities argued that, in applying for an income tax exemption in respect of his income obtained in performance of this work, Mr Petersen could not rely on the free movement of workers. Given the work was carried out by a Dane, for a Danish employer and entirely outside the EU, ‘no adequate link’ could be said to exist between Denmark and Germany. Indeed, ‘an employee who carries on an activity in the context of development aid focused entirely in a third State cannot be considered to be simultaneously or even chiefly carrying on cross-border activity within the European Union’.Footnote 106 The argument thus amounted to characterising the relationship as implicating both the German and the Danish territories but without there being any movement between the two, and thus not triggering the applicability of EU law. The Court, however, was not persuaded. It responded by emphasising that the ‘fact that the applicant in the main proceedings carried on his activity in the context of development aid focused entirely in a third State’ is not of itself enough to exclude the applicability of EU law. The Court then listed several elements which demonstrated close links with two Member State territories, and therefore the EU as a whole. In addition to the fact that Mr Petersen was resident in Germany but employed by a Danish company, it was mentioned that he was covered by social insurance in Denmark, his salary was paid into a bank account in Denmark, and his contract was subject to Danish law.Footnote 107 The Court also previously noted that Mr Petersen and his wife owned a house in Germany, where they resided with their daughter, and also had a holiday home in Denmark.Footnote 108
The strong non-EU connections in this case not only did not call into question the applicability of EU law, they in fact appeared to reinforce it. The existence of foreign elements, such as the work being performed in Benin, pushed the Court to consider, over and beyond the existence of an intra-EU employment relationship, whether a meaningful social connection existed with the EU as a whole. The territory of EU law is thus not only a space of mobility across the territories of Member State—Petersen suggests that it is also a space of social integration.
This suggestion aligns with similar trends in relation to EU citizenship and its impact on the law of nationality. Member States are, in theory, exclusively competent unilaterally to bestow their nationality or decide to withdraw it. Since possession of a Member State nationality is a precondition for EU citizenship, who is or is not an EU citizen is a sovereign decision of individual Member States. The Court has, however, begun to erode this power gradually. In the Rottmann case, where the German authorities had decided to remove German nationality from an individual who had previously acquired it, thus leaving him stateless and therefore depriving him of his status as an EU citizen, the Court held that such a decision was governed by EU law.Footnote 109 In effect, this meant that, when considering any decision to withdraw their nationality, Member State authorities have to comply with the principle of proportionality.
The more recent decision in the Tjebbes case further elaborates the meaning of proportionality in this context and, in so doing, advances in the direction of constructing the EU legal order as a social community of belonging.Footnote 110 In relation to the reasons that could justify such a withdrawal by the Dutch authorities, the Court noted that ‘it is legitimate for a Member State to take the view that nationality is the expression of a genuine link between it and its nationals, and therefore to prescribe that the absence, or the loss, of any such genuine link entails the loss of nationality’.Footnote 111 The Court thus conceived of nationality as essentially a presumption of integration in the relevant society. More significantly, it then focused on how such a link may be destroyed. The circumstance that the individual has habitually resided for more than ten years outside the Netherlands ‘may be regarded as an indication that there is no such link’, but only if such residence was also outside ‘the territories to which the EU Treaty applies’.Footnote 112 The Court has therefore fundamentally altered the relationship between Member State nationality and EU citizenship. Here, it is not simply by remaining Dutch that Mr Tjebbes gets to remain European. Instead, it is by participating in European society as a whole, among other things by exercising free movement and integrating in other Member States, that he maintains his Dutch nationality. The decision thus suggests that EU citizenship is not merely a collection of rights added to Member State nationality, but stands for a broader idea of social integration in the EU as a whole.
B. Imperativeness and Territorial Grounding
In private international law, party autonomy refers to the ability of private parties to choose which law will govern their relations.Footnote 113 In the past, this idea was controversial, as it was understood to question the very authority of the State over private relations. From a territorial point of view, an ability to opt out of the governing law challenged the basic assumption that all private relations must be concretely grounded in the territory of a particular jurisdiction. Today, however, the principle of party autonomy is widely accepted, most importantly in contract law but also in other fields (even if subject to the exceptional interference of certain regulations, often termed ‘internationally mandatory’ laws, which, because of their particular importance, apply regardless of party choice).Footnote 114 Nevertheless, party autonomy is usually reserved for parties to international relations, as opposed to purely domestic ones (ie those exclusively connected to one jurisdiction). Thus, a spatial distinction (based on the presence of cross-border elements) triggers a fundamental difference in regime, as it is only parties to international relations that enjoy the privilege of choosing the law that will best suit their interests. This means that, while purely domestic transactions remain grounded in the territory of their respective jurisdictions, international ones are imagined as belonging within a more detached cross-border space, where domestic regulations lose much of their strength through the operation of conflict of laws rules.Footnote 115
EU law has proven very receptive to the idea of party autonomy and the spatial differentiation that it carries, through the operation of the market freedoms (which protect the ability of individuals to move to a more favourable jurisdiction within the EU)Footnote 116 and, most importantly, its various regulations on choice of law (in particular the Rome I Regulation on the law applicable to contractual obligations, which provides that contracts ‘shall be governed by the law chosen by the parties’Footnote 117). That Regulation also distinguishes, in the effects given to party autonomy, between contracts that are purely internal to the EU and those that also present connections to third States, by implying in Article 3(4) that the latter will be able to opt out of certain mandatory provisions of EU law, but not the former.Footnote 118
By contrast, however, it will now be argued that, when it comes to delineating the external boundaries of EU law, the CJEU follows a very different path, one that shows significant resistance to the principle of party autonomy.Footnote 119 As explained below, the Court has stated that the imperativeness of EU law is indifferent to the presence of connections to non-EU jurisdictions. Its applicability will depend, not on the wishes of parties engaging in cross-border relations, but on whether those relations are concretely located on the territory of the EU. This approach negates any distinction between a space that is purely internal to the EU and one that is only partially or more loosely connected to its territory, leading to different regimes for domestic and cross-border relations. This again suggests that the EU's territory is not construed, against those of third States, as an abstract space of pure circulation and individual choice,Footnote 120 but instead as a concretely situated perimeter of social relations.
The Walrave decision already contained this idea, where it stated: ‘By reason of the fact that it is imperative, the rule on non-discrimination applies in judging all legal relationships in so far as these relationships, by reason either of the place where they are entered into or of the place where they take effect, can be located within the territory of the Community’.Footnote 121 However, it is in the Ingmar decision where the link between imperativeness and territorial localisation is most clearly borne out.Footnote 122 In that case, the contract signed between an agent (a UK company) and its principal (a California company) contained a choice of law clause that expressly designated as applicable the law of the State of California. Following termination of the contract, the agent sued to obtain damages pursuant to UK regulations transposing the EU Directive on self-employed commercial agents (which mandate that such agents are entitled to compensation for termination). Two questions were put to the Court. First, do cases where the principal is established outside the EU fall within the territorial scope of application of the Directive? Secondly, even if the answer to the first question is positive, are the relevant mandatory provisions also of an internationally mandatory nature, in the sense of preventing parties to cross-border transactions from agreeing to exclude their application by submitting to the law of a foreign jurisdiction?
From a current private international law perspective, accustomed as it is to offering parties to international transactions the benefit of party autonomy, these two questions should be kept separate. Even if the relation in question falls within the scope of application of the Directive, this does not necessarily mean that its mandatory provisions cannot be overridden through a choice of law clause. That is not, however, the reasoning followed by the Court.Footnote 123 Having found, through the localisation approach described in the previous section, that the ‘situation is closely connected with the Community’,Footnote 124 it immediately held that all of its mandatory provisions should govern this particular transaction, regardless of its connections to a third State, and ‘irrespective of the law by which the parties intended the contract to be governed’.Footnote 125 There is thus no distinction between domestically mandatory provisions, and internationally mandatory ones.
The Ingmar decision therefore conveys the idea that the presence of non-EU connections does not justify the application of a differentiated regime. It is true that the decision remains the only one to have addressed the relationship between party autonomy and the territorial scope of application of EU law directly (it is for this reason that it remains ‘emblematic’Footnote 126 in private international law scholarship), and that it can be seen to stand in tension with Article 3(4) of the Rome I Regulation (as stated above). However, a similar thinking can be found in other decisions addressing related aspects of private autonomy, which suggests that Ingmar is not an outlier. In relation to international commercial arbitration, the Court has ruled that arbitral awards should be subject to appropriate judicial review, in order to ensure that parties do not, through recourse to arbitration, compromise the full effectiveness of EU law.Footnote 127 This again suggests that EU law does not allow for an in-between space, specific to international cross-border transactions, where the imperativeness of its regulations may be downgraded. A similar conclusion can be drawn from Commission v Spain, in relation to consumer law. The case concerned Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer contracts, which applies to all contractual relationships with ‘a close connection with the territory of the Member States’.Footnote 128 The Court somewhat laconically stated that this ‘deliberately vague’ expression ‘seeks to make it possible to take account of various ties depending on the circumstances of the case’, and should therefore be interpreted broadly and open-endedly.Footnote 129 Advocate General Geelhoed also explained, in his conclusions for that case, that the ‘broad criterion’ found in the Directive should be seen as seeking to ‘prevent the protection granted … from being undermined as a result of the parties to a contract having declared that the law of a non-Member country is to be applicable to the contract’.Footnote 130 The idea is thus very similar to Ingmar: EU law projects itself as applying equally to all relationships that, given their proximity to the EU as a whole, can be deemed to be located on its territory. There is no ‘grey area’ in the confines of EU law, where its applicability could, in the face of meaningful connections to third States and a choice of law clause, be reduced to being merely ‘semi-imperative’. In other words, EU law does not recognise a separate social space, governed by a lighter version of itself, which would be situated between its own territory and that of third States. In relation to EU law, a social relation can fall inside or outside, but not somewhere in between.
Moreover, Ingmar and Commission v Spain are also important in that they shed light on the particular relationship that is established between certain social relations (such as the commercial relation involved in an agency contract, or the relation between a business and a consumer) and EU law, by virtue of their localisation on EU territory. The vision of territoriality that emerges is not that of an abstract space, which may be left through sheer will. It encompasses instead a series of social relations that happen to be, as seen in the previous section, and again emphasised by the Court in Commission v Spain, concretely and meaningfully connected to the EU's territory. There is no individual autonomy beyond the one that derives from the EU legal system, for those concretely grounded on its territory.
Finally, it should also be emphasised that the question of imperativeness arises in relation to whether deference should be shown, by giving effect to party autonomy, towards the law of a non-EU State (Californian law, in the Ingmar case). This circumstance is key. As will be further elaborated in the next section, there is a qualitative difference between deferring to the law of a Member State and that of a non-EU jurisdiction. The CJEU insisted later in the Unamar decision on the need to distinguish between the two: ‘unlike the contract at issue in the case giving rise to the judgment in Ingmar, in which the law which was rejected was the law of a third country, in the case in the main proceedings, the law which was to be rejected in favour of the law of the forum was that of another Member State’.Footnote 131 Stéphanie Francq explains why:Footnote 132 it is because the different Member States all share in the constitutive values of the EU, and can therefore be described as a ‘community of law’, that EU law may more freely defer to their respective laws. Conversely, non-EU laws lie outside this axiological space, and party autonomy may only work in their favour under stricter conditions. Once again, and similarly to the earlier analysis of the Petersen case, it is the presence of foreign elements (that is, connections to non-EU jurisdictions) that allows the EU territory to emerge as a unitary notion.
C. The Integrity of the EU's Territory
A third category of cases will now be considered, where the territorial perimeter of EU law is drawn, not through the localisation of cross-border relations, or through the assertion of its territorial imperativeness, but by appeal to the very integrity of its legal system. To illustrate the distinctive logic followed by these cases, yet another mechanism of private international law can assist—the so-called ordre public or public policy exception. This traditional doctrine allows the judge to refuse to apply the normally applicable foreign law, or to recognise a foreign judgment, where to do so would lead to a violation of a fundamental principle or core value of its own legal system. Against private international law's tendency to show respect towards the different choices and values present in foreign legal systems,Footnote 133 the applicability of the judge's own legal system is here compelled out of a concern for the maintenance of its very ‘integrity … and the society that it represents’.Footnote 134 The ordre public exception is thus linked to an existential idea, in that the continuing identity and preservation of the judge's legal order as a whole is seen as at stake, were it to allow the foreign jurisdiction to govern.
It may seem that the public policy exception is irrelevant to the present analysis, inasmuch as it appears to operate regardless of any territorial configuration: the normally applicable foreign law is replaced with the forum's own, not because of any particular spatial connections, but due to that law's intolerable content. The reality is, however, more complicated. The doctrine's operation does incorporate a spatial dimension, in a way that, as will be argued, can help reveal also the close relationship between the territoriality of the EU legal system and its integrity. Indeed, private international law scholarship and judicial practice have gradually constructed a less well-known and somewhat controversial version of the ordre public exception, known in German by the expression Inlandsbeziehung and in French as ordre public de proximité.Footnote 135 This doctrine subjects the application of public policy to the presence of certain meaningful ties to the forum. Conversely, the integrity of the system is not deemed to be under threat where the situation at stake is perceived to be too remote. In this way, the doctrine of Inlandsbeziehung serves to ‘relativise’ the values held to be fundamental by the forum, by connecting them to a certain space of belonging, in relation to which specific relationships can be positioned as internal or external.Footnote 136
In a similar manner to the public policy exception, the integrity of EU law is also mobilised to justify its intervention in situations at least partially connected to external legal systems. This is however often done under the banner of the principle of autonomy. This principle is a key doctrine of EU law, as it has served as the basis for some of the most memorable pronouncements of the CJEU. At a very general level, the principle of autonomy means that EU law constitutes a ‘self-contained and self-referential legal system distinguishable and independent from national and international law’.Footnote 137 In the hands of the CJEU, the principle has led to the assertion of the priority of EU law over competing legalities, such as Member State law,Footnote 138 international lawFootnote 139 or the European Convention on Human Rights.Footnote 140 Autonomy can therefore be seen as directly relevant to drawing the boundaries of EU law vis-a-vis alternative legal orders. Nevertheless, much like the traditional doctrine of ordre public, the principle of autonomy seems to follow a strictly existential logic, rather than operate in a spatial manner. As the Court regularly insists, autonomy ‘stems from the essential characteristics of the European Union and its law’.Footnote 141 In other words, the very existence and preservation of EU qua legal system would be compromised, if it were to defer to an external source of authority, where doing so would jeopardise its own nature and axiological commitments.Footnote 142 Thus, EU law's self-assertion seems to derive from the very structure of its legal system and the content of the values on which it is founded, rather than any particular territorial configuration.
The present thesis, however, is that, comparably to the doctrine of Inlandsbeziehung in private international law, the appeal to the autonomy or integrity of the EU legal systems should be seen as involving a form of territorial boundary-drawing—rather than simply a matter of projecting EU values extraterritorially.Footnote 143 In certain cases, the applicability of EU law is justified by reason of the intolerable outcome that deferring to external legal systems or norms would lead to, even if that deference would normally be justified. As with the ordre public exception, it is the very integrity of the EU legal system that justifies its intervention. By taking this approach, the decisions that will be analysed define the spatial boundaries of the EU's own axiological commitments and connect those commitments to the EU's territory. The values of the EU legal system are not, therefore, conceived of in purely abstract terms, but linked to a concretely situated space. This idea will be illustrated through cases from three different areas.
First, the recent line of cases on the compatibility with EU law of treaty-based investor–State arbitration, which allows foreign investors to bring claims against host States before international arbitral tribunals, may be considered. In the landmark case of Achmea,Footnote 144 involving an investment treaty concluded between the Netherlands and the Slovak Republic, the CJEU was concerned that this mechanism might potentially lead to disputes involving questions of EU law to be adjudicated outside the EU judicial system. It therefore ruled that it had ‘an adverse effect on the autonomy of EU law’,Footnote 145 first, by threatening the full effectiveness of EU law (given the absence of appropriate judicial supervision), and secondly, by flouting the principle of mutual trust between Member States (since a different regime would be carved out for those two countries). The same conclusion was reached in the further cases of Komstroy Footnote 146 and PL Holdings.Footnote 147
Whilst the substantive meaning of these attributes of autonomy has been amply discussed and critiqued, much less attention has been paid to their spatial perimeter, which is implicitly affirmed in this line of cases. Indeed, autonomy is only deemed to be at play with regard to intra-EU investment relations—that is, as the Court describes them, those ‘between a Member State and an investor of another Member State concerning an investment made by the latter in the first Member State’.Footnote 148 The boundaries of autonomy are therefore clearly also spatial, inasmuch as the Court distinguishes, by reason of the territorial position occupied by the various parties and the cross-border relations established between them, what lies inside the EU's territory (the relational space of intra-EU investment) from what lies outside of it (which includes relations between an EU party and a non-EU one). The Court's decisions can consequently be read as guaranteeing the integrity of that particular space, since they reason that the EU legal system's very existence would be called into question, were the EU legal system's values and structural commitments not enforced equally through the social perimeter that it claims as its own (which is exactly the type of logic that sustains, in private international law, the application of the forum's law through Inlandsbeziehung).
A second example concerns the transfer of personal data outside the EU, which has generated a number of important pronouncements by the CJEU.Footnote 149 The most important for present purposes is the Schrems I decision, where the Court concluded that such transfers come within the scope of application of EU law on personal data protection, and are thus also subject to the requirements of the EU Charter of Fundamental Rights. National authorities were therefore held responsible for evaluating whether non-EU countries to which the data transfers are intended ensure an ‘adequate level of protection’. This decision has been read as extending the ‘extraterritorial’ or ‘global’ reach of EU data privacy law.Footnote 150 Such a reading, however, does not fit the reasoning followed by the Court, which chose to focus on the values on which the EU is said to be founded, rather than consider the territorial reach of its legislation or the regulatory goals that it pursued. The Court also emphasised that ‘the European Union is a union based on the rule of law in which all acts of its institutions are subject to review of their compatibility with, in particular, the Treaties, general principles of law and fundamental rights’.Footnote 151 It is thus the very nature of the EU that mandates that any transfer of data to third States complies with the basic requirements of justice that irrigate its legal system. Where the consequences of deferring to foreign authorities (by allowing the data to fall under their control) would be too fundamentally at odds with core EU values, the authorities of Member States should prevent the transfer. To proceed otherwise would compromise the very integrity of EU law.
How does this case contribute to constructing the EU legal order as a concretely situated community? First, it is important to underline that the decision allows the EU legal system to assert a specific axiological identity vis-à-vis those of non-EU jurisdictions. Such an approach necessarily involves a comparison between the values of the forum and those of the foreign legal system concerned, leading the former to refrain from deferring to the latter where the contrast between the two is too stark. By assessing the adequacy of the level of privacy protection that the US legal system guarantees, the CJEU is not only constructing the values that are so fundamental to EU law to warrant application regardless of any territorial connection, but it is also distinguishing the EU legal system from others through their respective axiological commitments.
What, however, is the relationship of those values to territoriality? After all, the EU legal system seems to assert itself regardless of any territorial configuration, or indeed of any spatial connection to it. In reality, however, the notion of EU territory does emerge as unitary and distinctive, as these cases help to illustrate. To begin with, deferring to a foreign legal system would involve, quite literally, an actual exit from the geography of the EU (the transfer of data to servers in the United States). It is therefore possible that such an exit may involve a risk of a breach of the fundamental axiological commitments of the EU legal system that warrant that EU authorities exercise a form of control—and therefore, on the basis of those values, that territorial mobility may be blocked. A strong connection is thus established between the values of the EU and its territory. As Hans Lindahl has noted, ‘values are a constitutive feature of territoriality as such; indeed, space becomes the area—the EU's own place—normatively mediated in terms of values deemed to be relevant’.Footnote 152 In this sense, it is the fact of no longer being within the territory of the EU that prevents those values from being realised through the medium of EU law, thus conjoining a physical and normative space.Footnote 153 In the same way that, as described in the previous sections, the Ingmar and Prodest cases served to illustrate that the applicability of EU law is based on concrete social grounding, here again the Court demonstrates that the values of the EU are not abstract ideas but inextricably linked to a specific and limited geography.
The third and final example comes from the field of extradition law, in relation to extradition requests addressed to Member States to surrender an EU citizen to be judged or serve a sentence outside the EU. Article 19(2) of the EU Charter of Fundamental Rights, under ‘Protection in the event of removal, expulsion or extradition’, provides that ‘No one may be removed, expelled or extradited to a State where there is a serious risk that he or she would be subjected to the death penalty, torture or other inhuman or degrading treatment or punishment’.Footnote 154 In the Petruhhin case, concerning the extradition to Russia from Latvia of an Estonian national, the Latvian authorities were required to consider whether there was ‘a real risk of inhuman or degrading treatment of individuals in the requesting third State’.Footnote 155 Later cases have reaffirmed this requirement.Footnote 156
Again, the case can be read through an analogy with the public policy exception in private international law. The extradition process is one of cooperation between two legal systems, where the extraditing State agrees to surrender an individual to the receiving State, resulting in a change in the governing legal system. Indeed, as a consequence of the extradition decision, the prosecuted or convicted individual ceases to be subject to the legal system of the former and falls instead under that of the receiving State. Conversely, a refusal to extradite will entail that they remain subject to the local authorities and their law. Just as the recognition of a foreign judgment or the application of a foreign law will be refused, in private international law, where it will lead to results that would compromise core values of the judge's own forum, here the extraditing authorities also refuse to defer to those of the requesting State, after examining the content of the latter's legal system, where the outcome would be too obvious in contrast to local values. Petruhhin can be understood as relying ultimately on this contrast between local values and foreign ones. The CJEU has thus emphasised that, ‘in its relations with the wider world, the European Union is to uphold and promote its values and interests and contribute to the protection of its citizens’.Footnote 157
This case may thus be read very similarly to Schrems I. Indeed, Petruhhin should not be reduced to an example of extraterritoriality, as is sometimes suggested of extradition law.Footnote 158 The decision should instead be read as based on the idea of the integrity of the EU legal system, which would be compromised if the EU citizen were surrendered to the authorities of a third State that does not respect basic EU values. Integrity is again presented in spatial terms, as in the case of Schrems I, since the preservation of the EU's own axiology is linked to the continuing presence of the EU citizen on EU soil.
However, Petruhhin can also serve to illustrate a further aspect—how the presence of ‘foreign elements’ (here the extradition request by the Russian authorities) may allow for the emergence of a unitary EU territory as a space of belonging for EU citizens, rather than as a byword for the collection of the various territories of Member States. It has previously been seen that, in cases such as Petersen or Ingmar and Unamar, the presence of non-EU connections pushes the Court to consider whether the said relationship is connected to the EU territory as a whole rather than to a particular Member State. A similar effect can be seen in Petruhhin. As argued by Coutts,Footnote 159 the Court in that case seemed to suggest, on the basis of EU citizenship, that a prosecution anywhere in the EU would be preferable to extradition to a non-EU State. The Latvian authorities (to whom the extradition request had been addressed by Russia) ought therefore, before agreeing to the extradition, to offer the possibility of prosecuting, not only to the Member State of Mr Petruhhin's nationality (Estonia), but also to any other Member State that might potentially be interested. Thus, the possibility of extradition from the EU serves to assert a certain territorial unity between its Member States (their different territories being all interchangeable), as well as to underline the connection between the EU territory and EU citizenship.
By way of contrast, it is useful to note the case of Pisciotti,Footnote 160 concerning the possible extradition to the United States by German authorities of an Italian national, which took a different approach to Petruhhin. Here the Court seemed to move away from Petruhhin by giving priority only to a potential prosecution in Italy, instead of in the EU territory. A similar approach was followed in the more recent case of Generalstaatsanwaltschaft Berlin.Footnote 161 The very idea of an affiliation to the EU as a whole vanishes, and the EU is once again seen as a functional actor, with the Member States as sole territories of belonging. As Coutts notes, one can ‘witness the absence of this protection offered to the individual qua Union citizen and an emphasis placed on the right of the Member State of nationality to assert its jurisdiction over its nationals’.Footnote 162 In other words, ‘[t]he relationship shifts from one of the Union citizen, deriving protection from the Union legal order, to national, one over whom the state claims a privileged relationship of authority, before and above the claims of other (third) States’.Footnote 163 Because the integrity of the EU legal system ceases to be a relevant consideration in these cases, the EU cannot really be said to be presented as a bounded social space, and appears instead as the mere aggregate of physical territories of the Member States.
These ideas can help us better understand other cases, such as the well-known decision of Ruiz Zambrano, where the Court famously stated that EU citizenship ‘precludes national measures which have the effect of depriving citizens of the Union of the genuine enjoyment of the substance of the rights conferred by virtue of their status as citizens of the Union’.Footnote 164 One such measure would be a ‘refusal to grant a right of residence to a third country national with dependent minor children in the Member State where those children are nationals and reside, and also a refusal to grant such a person a work permit’.Footnote 165 The case is usually read as concerning primarily the relationship between EU law and Member State law, since it unexpectedly extended EU law to an issue that was previously thought to be internal to Member States (that of the rights of the non-EU parents of EU children, in situations without an intra-EU cross-border element).Footnote 166 It can, however, alternatively be read as asserting a connection between EU values and its concrete geography. Indeed, the applicability of EU law may be said to have been triggered by a risk of violation of its fundamental values (referred to here as ‘the substance of the rights’ attached to EU citizenship). The exception is not based on a comparison between the values of the EU and those of the Member State involved in the case, but instead with those of the non-EU jurisdiction to which the children would be subject if their father were not granted a right of residence. As argued by Azoulai, leaving the territory of the EU:
means not only leaving Europe in the geographical sense; it means leaving a community of ideals and values; it means being deprived of a certain mode of existence corresponding to the standards of European society. As stated in this case, the territory of the Union … stands for the mix of material and immaterial things that determines the sustainability of individual existence; what we may call a ‘European way of life’.Footnote 167
IV. CONCLUSION
This article has examined an issue of considerable practical difficulty, which is made even more complex by the broad variety of settings and legal areas in which the territorial scope of application of EU law needs to be determined. In so doing, an attempt has been made to connect what may seem like a somewhat technical and peripheral question to foundational questions about the very notion of territoriality, the place of the EU legal system in its relationship to the wider world, and the spatial dimension to its broader constitutional identity.
The overview of the case law has enabled the identification of two approaches that stand in tension with one another, at both practical and theoretical levels, and which reflect two competing visions. The first, the dominant one in scholarship, is a functionalist approach that links the territorial reach of EU law to its pursuit of universal goals, subject to the external limits set by public international law. The second approach, much less visible but identified here with the help of three doctrines of private international law, localises relations within a space that the EU legal system claims as its own, and whose territorial integrity it sets out to preserve. Whereas the first approach portrays the EU as a benign actor on the world stage, the second imagines it as a situated community of belonging. While the former reduces the EU's territory to the cartographic surface from which it can accomplish its goal-oriented interventions, the latter conceives it as a relational space that is informed by the EU legal system's particular axiology and structure. In this way, the examination of the case law has allowed glimpses of a landscape much more complex than is usually thought to exist, from the particular vantage point of how it engages with external spaces. The EU is both a global actor and a local community, it self-identifies as paragon of universal values but also, and simultaneously, as the creator of boundaries of inclusion and exclusion.
The contribution of this article is therefore threefold. First, legal practice in this area has been methodologically enriched, pointing out the insufficiencies of the widespread functionalist method, while exposing and developing an alternative with the help of doctrines of private international law. Secondly, a more sophisticated understanding of territoriality has been offered, through the case study of the EU legal system, thus contributing to a growing public international law literature that is moving away from traditional conceptions of territory, to reconceive it as a form of bounded social space, one that is relevant to our understanding of jurisdiction beyond the State. Thirdly, and finally, the analysis is directed at debates about the constitutional identity and broader legitimacy of the EU legal system. The tension that has been identified in the case law around this somewhat technical issue can be seen to reflect a larger opposition between competing visions about the EU's place in the world. Offering a counterpoint to the narrative that ties the project of European integration to ambitions of global governance, an alternative vision has been brought to light, where the basis for that project is found in the inextricable link between the EU's particular axiology and its own bounded space.
Acknowledgements
I am grateful to Roxana Banu for her comments on an earlier version of this piece, to Gail Lythgoe and Stavros Makris for their suggestions, and to the two anonymous reviewers for their constructive feedback. For their generous exchanges, special thanks go to the other members of the FOLIE research project (Forms of Life and Integration in Europe), as part of which I first developed the main arguments of this article: Lola Avril, Loïc Azoulai, Veronica Corcodel, Edouard Dubout and Hanna Eklund.