THE “LOGIC OF GLOBALIZATION” VERSUS THE “LOGIC OF THE INTERNAL MARKET”: A NEW CHALLENGE FOR THE EUROPEAN UNION

Globalization confronts the European Union with many new challenges. One of these concerns the applicability of harmonized EU law to cross-border situations involving third countries. In its recent judgment in Google/CNIL (C-507/17), on the territorial reach of the EU data protection rules and the “right to be forgotten”, the CJEU introduces a new “logic of globalization” which must be distinguished from the traditional “logic of the internal market”. While the latter justifies extraterritoriality in case internal market interests are affected, restraint characterizes the former. The global horizon does not diminish pertinent EU interests and objectives, but their effective implementation is threatened by the absence of the ensured enforcement of EU law and potential countermeasures. In context of globalization, it is international collaboration rather than unilateralism that would enable the EU to protect its interests and those of its citizens more adequately.


20
ject to comments and analysis in several online blogs that examined the judgment from various perspectives 2 and specifically attempted to bring some nuance, from the broader perspective of EU law, to the media headlines that Google had won a landmark case. 3 Our aim is not so much to focus on the precise contribution of Google/CNIL to the interpretation of the pertinent EU legislation on data protection, but rather to use this judgment and the preceding Opinion of Advocate General (AG) Szpunar as a starting point to reflect on the global role that the EU can assume, the position of (EU) private international law in that respect and the concept of globalization as such. 4 In Google/CNIL, the EU Court of Justice responded to the preliminary questions submitted to it by the French Council of State (Conseil d'État) on the interpretation of Directive 95/46. 5 Google had applied to the Council of State to seek annulment of the fine that the French Data Protection Authority -the Commission nationale de l'informatique et des libertés (CNIL) -had imposed on it, after Google had refused to comply with its earlier formal notice that, when granting a request from a natural person for links to web pages to be removed from the list of results displayed following a search conducted on the basis of that person's name, Google must apply that removal to all its search engine's domain name extensions. Google, on the contrary, wished to confine itself to removing such links from the results displayed following searches conducted from the domain names corresponding to the versions of its search engine in the EU Member States, complementing this refusal by a "geo-blocking" proposal.
As from 25 May 2018, Directive 95/46, the so-called Data Privacy Directive, which was based upon former Art. 100A EC Treaty (today Art. 114 TFEU), has been repealed and replaced by the new EU General Data Protection Regulation (GDPR), 6 which is based upon Art. 16 TFEU and which the CJEU also involved in its preliminary judgment.
Essentially at stake in this case was the territorial reach of the "right to de-referencing", also called the "right to erasure" or the "right to be forgotten", which had been examined earlier in the CJEU's 2014 judgment in Google Spain 7 and has been laid down in Art. 17 GDPR. The CJEU was asked, through several preliminary questions, whether, if all pertinent conditions are fulfilled, EU law requires de-referencing at national, European (i.e., all EU Member States) or worldwide level. In its judgment, the CJEU gave a (solely) EU-wide reach to the right to de-referencing. It held more precisely that the 2 See e.g., https://europeanlawblog.eu/2019/10/29/google-v-cnil-case-c-507-17-the-territorial-scope-of -the-right-to-be-forgotten-under-eu-law/, https://verfassungsblog.de/the-judgment-that-will-be-forgotten/, https://www.huntonprivacyblog.com/2019/09/26/cjeu-rules-right-to-be-forgotten-on-google-limited-to -the-eu-in-landmark-case/. pertinent provisions of the Data Privacy Directive and the GDPR must be interpreted in the sense that where a search engine operator grants a request for de-referencing pursuant to their provisions, the operator "is not required to carry out that de-referencing on all versions of its search engine, but on the versions of that search engine corresponding to all the Member States, using, where necessary, measures which, while meeting the legal requirements, effectively prevent or, at the very least, seriously discourage an internet user conducting a search from one of the Member States on the basis of a data subject's name from gaining access, via the list of results displayed following that search, to the links which are the subject of that request". 8 The Court thus followed AG Szpunar's plea for "a European de-referencing". 9 In order to come to its conclusion, the CJEU referred to the objective of the GDPR and the earlier directive (paragraph 54), the EU legislature's competence (paragraph 58), the scope that should be attributed to de-referencing "beyond the territory of the Member States" (paragraph 62) and, last but not least, the "globalized world" in which access to the internet (i.e., a "global network without borders") takes place (paragraphs 56-57). While the Court interpreted EU law as not requiring a search engine operator to carry out de-referencing on all, worldwide versions of its search engine, it added that EU law does not prohibit this either, and that the Member States hence have the power to order a search engine operator to do so anyway, after weighing the data subject's right to privacy and the protection of personal data on the one hand, and the right to freedom of information on the other hand (paragraph 72). In this judgment, as well as in other recent cases, 10 the CJEU tackled contemporary issues of "cyberspace" that affect important societal interests. Still, our contribution will not so much focus on the Court's interpretation of substantive EU law, but rather on the way it approaches a subject-matter that has, or could have, global ramifications that extend well beyond the territory of the EU Member States. Google/CNIL indeed constitutes a prime example of the way that the CJEU involves issues and concerns of globalization within its interpretation of EU law and, when doing so, appears to make use of particular conflict-of-laws methodology.

GLOBALIZATION AND THE LAW: A MULTITUDE OF QUESTIONS FOR EU LAW AND PRIVATE INTERNATIONAL LAW
Few topics of legal research are as trending today as the impact of globalization on the law. For a number of years, scholars in a great variety of fields have explored whether and, if so, to what extent and in what sense, legal systems are transformed by or should adapt to the ever increasing, and apparently ever faster, globalization of personal, social and economic relations and the legal means through which private and public, individual or institutional actors worldwide respond to it. They have also explored new paths of legal theory that integrate the (assumed) structural change of the law in this era of globalization.
Although the debate on globalization affects many fields, one may readily assume that it is particularly relevant for both EU law and private international law. While the debate on the EU as a "global actor" to a large extent focuses on more traditional topics of EU external relations law, though approached from the new institutional perspectives provided by the Treaty of Lisbon as they have been interpreted recently by the CJEU, 11 it appears that the phenomenon of globalization truly renews traditional academic approaches to private international law.
Undoubtedly, private international law today has a "global horizon", that manifests itself in very diverse forms. 12 Groundbreaking work in that regard has been accomplished recently by Horatia Muir Watt, who, mostly in collaboration with select international colleagues, has both identified its most pertinent issues, questions and cases and, starting from a clear dissatisfaction with current conflict of laws doctrine, proposed new academic thinking on "global" private international law. 13 At the occasion of receiving a doctorate honoris causa from the University of Silesia in Katowice, in 2019, prof. Paul Lagarde expressed some further ideas on this topic and pointed in particular to the effects, through the shifting and weakening of the territorial and temporal connection of international situations that it implies, of globalization on the methodology of private international law. 14 He illustrated his position on the best way for private international law to remedy possible problems linked with globalization with references to the recent collection by Muir Watt et al. of globalization cases. 15 He specifically referred to cases that arose through specific technological evolution of various kinds: the Yahoo! v. Licra case on illegal internet sales (nazi memorabilia) and the Blood and Gomez cases on the cross-border transfer of sperm samples for post mortem insemination. 16 While these cases testify of the important effect that new technological evolutions can have on globalization, they also raise questions on the concept of globalization itself.
While Yahoo! involved a French-American legal dispute, Blood and Gomez related to UK-Belgian and French-Spanish legal diversity respectively. The examination of the underlying fact situations makes one wonder about the definition of globalization. Aren't the latter cases foremost characterized, and hence interesting for further academic debate, by the challenges that new technological evolutions pose for the law? Why characterize such intra-EU disputes on assisted reproductive treatment, which is the subject of unharmonized, different legislative approaches in the respective Member States involved, as prime examples of globalization? In the Blood case, the legal debate even centered specifically on the interpretation of the free movement of services under Articles 56 and 57 of the Treaty on the Functioning of the European Union (TFEU). In that sense, these cases remind us, to name but one example, of the well-known 1991 judgment of the Court of Justice in Grogan on the application of the Treaty's free movement provisions on the distribution in Ireland of information regarding abortion in the United Kingdom. 17 Of course, contrary to Grogan, the Blood and Gomez cases were characterized, as Paul Lagarde rightly emphasized, by a very specific time element, as they related to the legal possibilities for posthumous insemination. 18 Still, this doesn't take away that such cases can be categorized under a pure internal market analysis. Actually, a closer look reveals that the cases examined under the title of "global private international law" include those that involve third countries but fall within the reach of EU legislation (e.g., Owusu, one of the most controversial ECJ judgments on the EU jurisdiction rules, involving a UK-Jamaica fact situation 19 ) as well as those such as Centros 20 and Laval 21 which are almost archetypical internal market cases that, however, cannot be separated from their conflict-of-laws dimensions.
Yahoo!, for its part, raised the issue of extraterritoriality as regards jurisdiction (in France, over Yahoo.com and non-French websites) and enforcement (in California, of the French court order). Such case obviously transgresses EU law, as it would -todaynot be covered by the EU legislation on that matter (in particular, the "Brussels Ibis" regulation 22 ). While the issues at stake -extension of jurisdiction to "cyberspace" and the interaction with constitutional norms such as freedom of expression -could have arisen within a purely EU-internal context as well, and then probably have triggered debate as well, they would in that hypothesis have fallen within the scope of Brussels Ibis. That such a broad variety of cases, with divergent European and/or international connections, can be categorized under a joint term "globalization" can only be explained by a substantive, rather than spatial, understanding of this concept. In the book project on "global private international law" (supra), globalization is understood as "the specific compression of time and space which coincides with late modernity; the coming of risk society, global neo-liberal economics (...); the paradoxical 'return of science' in a period of increasing disbelief in the values of modernity; and (...) the 'liquidification' of sovereignty". 23 While this approach at least has the advantage of transparency -as the definition of globalization is all too often considered self-evident or perhaps even deliberately omitted -the added value of such substantive understanding as compared to a more traditional, spatially oriented definition of globalization is subject for debate. At the very least it is clear, somewhat surprisingly, that there is no real doctrinal consensus on the concept of globalization.
The recent CJEU judgment in Google/CNIL, as well as AG Szpunar's Opinion in that case, offer food for thought, both with respect to the concept of globalization and as regards the analysis of the interests at stake. More particularly, there appears to exist a new and specific "logic of globalization", that appears to inspire (to a certain degree, at least) the Court of Justice when interpreting the application of EU law to particular cross-border disputes.

A CLOSER LOOK AT THE COURT'S AND THE AG'S REASONING: THE "LOGIC OF GLOBALIZATION", AS APPROACHED THROUGH EU-STYLE INTEREST ANALYSIS
Rather than focusing upon the search for an appropriate, comprehensive concept of globalization, which would require an in-depth analysis that cannot be done within the contours of this brief contribution, it seems worthwhile to take a closer look at the CJEU's reasoning in those cases that at least appear to be "globalized", as they require the examination of the applicability of harmonized EU law to cross-border situations that are not intra-EU but involve third, European and/or non-European, countries as well. Google/CNIL is such a case, and its analysis sheds some light on the importance that the CJEU attaches to a "logic of globalization", which apparently can be differentiated from an internal market logic.
In Google/CNIL, the CJEU is actually very cautious when interpreting the (extra-) territorial reach of the relevant EU legislation on privacy and data protection. According to the CJEU, the objective of the GDPR, and of the earlier Data Privacy Directive, is to guarantee a high level of protection of personal data throughout the Union (paragraph 54). The Court admits that "a de-referencing carried out on all the versions of a search engine would meet that objective in full" (paragraph 55) and that internet access outside the Union to the referencing of a link referring to information regarding a person whose centre of interests is situated in the Union is likely to have "immediate and substantial effects on that person within the Union itself" (paragraph 57), which justifies legislative competence on the part of the EU (paragraph 58). While these considerations appear to announce that a broad reach should be granted to the pertinent EU legislation, the Court refuses to take the next step, which actually would have been understandable in light of those considerations, to interpret the GDPR as imposing an obligation to worldwide de-referencing (paragraph 64). On the contrary, this de-referencing is only imposed in respect of all EU Member States (paragraph 66) and the CJEU does not go further than a reminder that EU law does not prohibit worldwide de-referencing (after weighing the fundamental rights involved) (paragraph 72).
In order to explain its prudent approach, the Court refers to the fact that "numerous third States do not recognize the right to de-referencing or have a different approach to that right" and that the balance struck between the two fundamental rights involvedthe right to privacy and the right to the protection of personal data on the one hand and the freedom of information of internet users on the other hand -"is likely to vary significantly around the world" (paragraph 59), as well as to the absence of a (clear) decision in that respect of the EU legislature (paragraphs 61-62). While the latter justification is not that convincing in view of the CJEU's much more active, often controversial or according to some even "activist" approach to the interpretation of EU law in many other cases, 24 the former argument is also quite remarkable. It testifies to a great reticence on the part of the EU to impose its policy beyond its territory, at least in a case such as Google/CNIL. The reasons given by the CJEU for its cautious approach make sense. And yet, it is not unlikely that the true explanation of the Court's reticence should rather be found in the practical power limits that globalization confronts even the EU with. Suppose that the CJEU orders global de-referencing on the basis of a more affirmative interpretation of the GDPR, would the EU really be able to enforce such order in all circumstances, given -as the Court explicitly mentions -the absence of "cooperation instruments and mechanisms as regards the scope of a de-referencing outside the Union" (paragraph 63)? In that sense, one must assume that the particularities of globalization require the EU to be modest, in particular with respect to extra-EU situations.
This new, "globalized" perspective and its impact on the interpretation (of the reach) of EU law was referred to as well, and more explicitly, by AG Szpunar. Interestingly, the Advocate General points to the differences between, on the one hand, the internal market, "a territory that is clearly defined by the Treaties", and, on the other hand, a global phenomenon such as the internet, which "is present everywhere". This fundamental difference makes it in his view difficult to draw analogies with fields such as EU competition and trade mark law where the CJEU exceptionally admits extraterritorial effects in view of the effects on the internal market (points 50-53). Szpunar's reasoning is further inspired by particular aspects of the globalization that characterizes the internet and the limiting effects flowing from it for the reach of EU law. As regards the balancing of the fundamental rights involved -the right to data protection and privacy versus the interest of the public in access to information -he underlines that Art. 11 of the Charter of Fundamental Rights of the EU (hereinafter: the Charter) protects, with respect to the latter right, "not the worldwide public but the public that comes within the scope of the Charter, and therefore the European public" (point 59). To this he adds, in a consideration which has been echoed later by the Court in its judgment (supra), that the public interest in access to information necessarily varies, depending on its geographic location, from one third State to another (point 60). Preventing individuals in third countries to get access to information, through an EU de-referencing order, could result in reciprocal de-referencing orders and a race to the bottom (point 61). For those reasons, AG Szpunar concluded that the issues at stake in Google/CNIL do not require the application of the pertinent EU legislation outside the EU's territory (which does not exclude the existence of situations in which the EU's interest would require such application) (point 62). Further in his Opinion, AG Szpunar pleads for de-referencing at EU, rather than national level, which suits the nature of the GDPR as a directly applicable EU regulation but in his view holds true as well with regard to the earlier Data Privacy Directive, in view of "the logic of the internal market" (points 75-77).
At times, AG Szpunar's Opinion in Google/CNIL carries a far, but noteworthy echo of the reasoning that was made, more than half a century earlier in a very different context of US interstate conflict of laws, by Brainerd Currie when he introduced his so-called governmental interest analysis.
Szpunar insists on the inherent limitations on the reach of the freedom of information in Art. 11 of the Charter, which in his view is self-evidently meant to protect "not the worldwide public" but "the European public", in view of the "connection with EU law and its territoriality" that would be required (point 59). Although the Charter has indeed a limited scope of application, as defined by Art. 51, the AG's focus on a territorial limitation is not as obvious as he suggests, as is evidenced by the Court's admission, in principle, of a competence for the EU legislature to lay down a global de-referencing obligation (point 58 of the judgment). Still, Szpunar's interpretation reminds us of Currie's warning that while lawmakers "are accustomed to speak in terms of unqualified generality", these "extravagantly general terms" must not be taken literally. 25 To quote his example: when the legislature of Massachusetts adopts legislation that without further qualification protects "married women", one should understand that this legislation in reality only serves to protect "those [women] with whose welfare Massachusetts is concerned, of course -i.e. Massachusetts married women". 26 Szpunar further makes the decision on the reach of the EU de-referencing rules explicitly dependent on "the interest of the European Union" (point 62). A decision in a different sense would carry a genuine risk of a race to the bottom, to the detriment of the freedom of expression, on a European and worldwide scale (point 61), which obviously wouldn't be in the European interest either.
Last but not least, the AG's plea for moderation 27 in view of the genuine risk for countermeasures and a race to the bottom echoes Currie's rejection of "the ruthless pursuit of self-interest by the states", his recognition of "rational altruism" and his plea for "restraint and enlightenment" with respect to the identification of the policies and interests involved. 28 Neither the particular brand of interest analysis that characterizes AG Szpunar's Opinion nor the Court's reasoning along those same lines result in any conclusion that the EU would not have an interest in worldwide de-referencing, quite to the contrary. Both, however, refrain from pushing the enforcement of that interest to its limits, particularly in view of the risk that such extraterritorial application of EU law would entail as regards potential countermeasures and probably taking into account as well a realistic assessment of the limited chances of full enforcement of a global European de-referencing order (supra).
To gain full insight into this reasoning, which appears to rest on a particular "logic of globalization", it is also very interesting to compare the Court's and the AG's prudent approach in Google/CNIL to the Court's reasoning, which was based on some type of interest analysis as well, in its earlier judgments in Ingmar and Agro Foreign Trade on the interpretation of the agency directive.
In Ingmar, the Court ruled in favor of the application of the protective rules of the Commercial Agency Directive 29 where the commercial agent carried on his activity in a Member State (the United Kingdom) although the principal was established in a non-member country (the United States) and a clause of the contract stipulated that the contract was to be governed by the law of that country (and more particularly the law of California). This interpretation was based upon that directive's objective to protect commercial agents, ensuring freedom of establishment and the operation of undistorted competition in the internal market. 30 In its more recent judgment in Agro Foreign Trade, on the other hand, the Court logically arrived at the opposite conclusion with regard to the applicability of the same directive to a commercial agent carrying out activities outside the EU (i.e., in Turkey) while the principal was established in a Member State (i.e., Belgium). Where a commercial agent carries out his activities outside the EU, the fact that the principal is established in a Member State does not present a sufficiently close link with the EU for the purposes of the same Directive 86/653 in the light of its objectives as explained in earlier case-law such as Ingmar. 31 Comparing the Court's approach in Ingmar and, in particular, Agro Foreign Trade, with its reasoning in Google/CNIL, is illuminating. Both in Agro Foreign Trade and Google/CNIL, the Court limited the reach of EU law. The difference between both cases is, however, that, while the Court in Agro Foreign Trade denied any interest in the applicability of EU law -as measured by the concerns of the internal market -in the given fact situation, this was different in Google/CNIL where the Court refused to order global de-referencing in spite of having indicated that such action would meet the objectives of the pertinent EU legislation "in full". Very probably, the nature of the internet as a global network serves as the essential distinguishing factor between the situations covered by these respective judgments. The EU is confronted with the limits of its powers to enforce its rules on a worldwide scale, in contrast to its power over intra-EU situations, including the internal market, in which de-referencing can indeed be ordered or the freedom of establishment and undistorted competition protected.
This comparison also sheds more light on the identification of those cases and issues that raise interest from the perspective of globalization and its effects on the law. For the purpose of identifying the legal challenges created by globalization, one should indeed look further than the mere selection of extra-EU cases in the sense of those cases that are connected to both the EU and third countries. Technology, as evidenced in Google/ CNIL by the Court's references to the internet as "a global network without borders" and the internet users' access "in a globalized world", clearly constitutes a distinguishing factor to define pertinent globalization cases. This is not meant as a confirmation that technology is the essential, necessary and sufficient factor to define pertinent globalized cases -other factors are pertinent as well -but the Court's recent case-law at least makes clear that the impact on the law, including EU law and conflict of laws, of technological evolution that has the effect of diminishing or even eliminating the pertinence of territorial borders as we have traditionally known them, must be watched closely and requires a particular legal analysis that follows the "logic of globalization".
If that is true, however, one cannot escape the thought that such "logic of globalization" also implies additional duties for the courts, in particular their responsibility to adopt a global perspective. In Google/CNIL, neither the Court nor the AG really do so, at least not explicitly. While their reasoning examines extraterritoriality and the actual (legal) power of the EU with respect to a global phenomenon such as the internet, they do so from a sole EU perspective, apart from the Court's unsubstantiated consideration that "numerous third States do not recognise the right to de-referencing or have a different approach to that right" (paragraph 59). Certainly, it is the Court's task to interpret EU law only. But wouldn't it be wise for the Court, in such cases with clear global anchor points, to adopt a more global and comparative perspective as well, albeit only for information and interpretation purposes? The "right to be forgotten" is obviously not a mere concept of EU law, but has a broader European reach through the case-law of the European Court of Human Rights and has also been recognized in various ways by the legal systems of many third States beyond the European borders. 32 Taking that into account when considering the extraterritorial effect of pertinent EU legislation or, as the AG does, contemplating the risk of countermeasures and a race to the bottom, would have given additional weight to their eventual interpretation of EU law.