This document is an excerpt from the EUR-Lex website
Opinion of Advocate General Bobek delivered on 3 September 2020.
Opinion of Advocate General Bobek delivered on 3 September 2020.
Opinion of Advocate General Bobek delivered on 3 September 2020.
ECLI identifier: ECLI:EU:C:2020:649
OPINION OF ADVOCATE GENERAL
delivered on 3 September 2020(1)
D.-H. T. acting as insolvency administrator in relation to the assets of J & S Service UG
(Request for a preliminary ruling from the Bundesverwaltungsgericht (Federal Administrative Court, Germany))
(Reference for a preliminary ruling — Jurisdiction of the Court — Reference in national legislation to provisions of EU law — Dzodzi case-law — Direct and unconditional renvoi — Interest in conceptual uniformity — Regulation (EU) 2016/679 — Data protection — Restrictions — Article 23(1), points (e) and (j) — Enforcement of civil law claims — Insolvency proceedings — Tax authorities)
1. According to Article 267 TFEU, a request for a preliminary ruling must concern the interpretation of the Treaties or the validity and interpretation of acts of the institutions, bodies, offices or agencies of the Union. One of the conditions for the Court’s jurisdiction under that provision is that the EU act, the interpretation of which is sought, is applicable in the main proceedings, with that applicability normally set out in the relevant EU law act itself.
2. The Court has nonetheless stated, beginning with the judgment in Dzodzi, (2) that the Treaties do not exclude from the jurisdiction of the Court requests for a preliminary ruling on EU provisions which do not apply to the facts of the case directly (that is to say, by virtue of provisions contained in those EU law acts themselves), but where those provisions are made applicable indirectly (that is, through a renvoi (reference) made by national law, in effect extending the scope of EU law). The Court thus held that where, in regulating situations outside the scope of the EU measure concerned, national legislation adopts the same solutions as those adopted in that measure, it is clearly in the interest of the European Union that, in order to prevent future differences of interpretation, provisions taken from that measure should be interpreted uniformly.
3. Although that dictum was then confirmed and applied in a number of subsequent cases (‘the Dzodzi case-law’ (3)), the limits of the Court’s jurisdiction in those situations remain, to date, still quite unclear.
4. The present case brings that line of case-law to its outer limits. In its request for a preliminary ruling, the Bundesverwaltungsgericht (Federal Administrative Court, Germany) asks the Court to interpret Article 23(1), points (e) and (j), of Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation), (4) even though that provision is not directly applicable to the situation at issue in the main proceedings. Indeed, that situation falls, for a number of reasons, outside the scope of Regulation 2016/679. Article 23(1) thereof has been made applicable to the situation before the referring court solely by virtue of a renvoi contained in the applicable national legislation.
5. The present case thus invites the Court to clarify how far the logic of a national renvoi, started in Dzodzi, can reasonably be pushed in a scenario in which, following not one, but in fact several extensions of the scope of an EU rule on the part of the national legislature, the referring court is faced with having to interpret that rule which, in my view, has simply nothing useful to say on the actual question before that court.
II. Legal framework
A. EU law
6. Recitals 2, 4 and 73 of Regulation 2016/679 read:
‘(2) The principles of, and rules on the protection of natural persons with regard to the processing of their personal data should, whatever their nationality or residence, respect their fundamental rights and freedoms, in particular their right to the protection of personal data. …
(4) The processing of personal data should be designed to serve mankind. The right to the protection of personal data is not an absolute right; it must be considered in relation to its function in society and be balanced against other fundamental rights, in accordance with the principle of proportionality. This Regulation respects all fundamental rights and observes the freedoms and principles recognised in the Charter [of Fundamental Rights of the European Union; “the Charter”] as enshrined in the Treaties, in particular the respect for private and family life, home and communications, the protection of personal data, freedom of thought, conscience and religion, freedom of expression and information, freedom to conduct a business, the right to an effective remedy and to a fair trial, and cultural, religious and linguistic diversity.
(73) Restrictions concerning specific principles and the rights of information, access to and rectification or erasure of personal data, the right to data portability, the right to object, decisions based on profiling, as well as the communication of a personal data breach to a data subject and certain related obligations of the controllers may be imposed by Union or Member State law, as far as necessary and proportionate in a democratic society to safeguard public security, including the protection of human life especially in response to natural or manmade disasters, the prevention, investigation and prosecution of criminal offences or the execution of criminal penalties, including the safeguarding against and the prevention of threats to public security, or of breaches of ethics for regulated professions, other important objectives of general public interest of the Union or of a Member State, in particular an important economic or financial interest of the Union or of a Member State, the keeping of public registers kept for reasons of general public interest, further processing of archived personal data to provide specific information related to the political behaviour under former totalitarian state regimes or the protection of the data subject or the rights and freedoms of others, including social protection, public health and humanitarian purposes. Those restrictions should be in accordance with the requirements set out in the Charter and in the European Convention for the Protection of Human Rights and Fundamental Freedoms.’
7. Article 1 (‘Subject-matter and objectives’) of Regulation 2016/679, provides:
‘1. This Regulation lays down rules relating to the protection of natural persons with regard to the processing of personal data and rules relating to the free movement of personal data.
2. This Regulation protects fundamental rights and freedoms of natural persons and in particular their right to the protection of personal data.
8. Article 23 of Regulation 2016/679 (‘Restrictions’) is the concluding provision of its Chapter III dealing with the rights of the data subject. It states:
‘1. Union or Member State law to which the data controller or processor is subject may restrict by way of a legislative measure the scope of the obligations and rights provided for in Articles 12 to 22 and Article 34, as well as Article 5 in so far as its provisions correspond to the rights and obligations provided for in Articles 12 to 22, when such a restriction respects the essence of the fundamental rights and freedoms and is a necessary and proportionate measure in a democratic society to safeguard:
(e) other important objectives of general public interest of the Union or of a Member State, in particular an important economic or financial interest of the Union or of a Member State, including monetary, budgetary and taxation matters, public health and social security;
(j) the enforcement of civil law claims.’
B. National law
‘(3) The provisions of this Law and the tax laws relating to the processing of personal data do not apply where European Union law, in particular Regulation 2016/679 … applies directly in the version valid in each case or in accordance with subparagraph (5).
(5) Unless stated otherwise, the provisions of Regulation 2016/679, of this Law and of the tax laws relating to the processing of personal data of natural persons apply correspondingly to information relating to identified or identifiable
1. deceased natural persons or
2. corporations, associations of persons and corporate funds with or without legal personality.’
10. Paragraph 32b (‘Duty of the financial authority to provide information where personal data have not been obtained from the data subject’) of the AO states:
‘(1) The duty on the part of the financial authority to provide information relating to the data subject in accordance with Article 14(1), (2) and (4) of Regulation 2016/679, in addition to the exceptions laid down in Article 14(5) of Regulation 2016/679 and Paragraph 31c(2), does not exist
1. where the provision of the information
(a) would be prejudicial to the proper performance of the tasks within the competence of the financial authorities or other public bodies within the meaning of Article 23(1)(d) to (h) of Regulation 2016/679, or
and therefore the interest of the data subject with respect to the provision of information has to be of secondary importance. Paragraph 32a(2) applies correspondingly.’
11. Paragraph 32c (‘Right of access by the data subject’) of the AO reads:
‘(1) The right of access by the data subject vis-à-vis a financial authority in accordance with Article 15 of Regulation 2016/679 does not exist where
1. the data subject does not have to be informed pursuant to Paragraph 32b(1) or (2),
2. the provision of information would adversely affect the legal entity of the financial authority in the establishment, exercise or defence of civil law claims or in the defence of civil law claims established against it within the meaning of Article 23(1)(j) of Regulation 2016/679; duties on the part of the financial authority to provide information under civil law remain unaffected,
12. Paragraph 32e (‘Relationship to other rights to access and provision of information’) of the AO states:
‘If the data subject or a third party has a right of access to information vis-à-vis the financial authority in accordance with the Law on freedom of information of 5 September 2005 … or in accordance with corresponding legislation of the Länder, Articles 12 to 15 of Regulation 2016/679 apply correspondingly in conjunction with Paragraphs 32a to 32d. More extensive rights to information relating to tax data are excluded in this respect. …’
13. Pursuant to Paragraph 129(1) of the Insolvenzordnung (Law on insolvency of 5 October 1994), as amended:
‘The insolvency administrator may challenge … legal acts that were performed prior to the commencement of insolvency proceedings and that are detrimental to creditors.’
14. Pursuant to the first sentence of Paragraph 143(1) of the Law on insolvency, the legal consequence of a successful challenge is that the beneficiary of the payment concerned must pay it back into the insolvency estate.
15. According to Paragraph 144(1) of that law, the beneficiary’s original claim against the insolvency debtor is reactivated. That claim must be satisfied from the insolvency estate.
16. Paragraph 2 (‘Scope’) of the Gesetz über die Freiheit des Zugangs zu Informationen für das Land Nordrhein-Westfalen (Law on freedom of information of the Land of North Rhine-Westphalia; ‘the Law on freedom of information’) of 27 November 2001, as amended, reads:
‘(1) This Law shall apply to the administrative activity of public authorities … For the purposes of this Law, a public authority is any body which performs public administration tasks.
17. Paragraph 4 (‘Right to information’) of that law states:
‘(1) Any natural person may claim, against the bodies referred to in Paragraph 2, in accordance with this Law, a right of access to official information held by the body.
(2) Any special legal provisions concerning access to administrative information, the provision of information or the granting of access to the file shall take precedence over the provisions of this Law. …’
III. Facts, national proceedings and the questions referred
18. The applicant in the main proceedings is an insolvency administrator of a company, J & S Service UG. In that capacity, he requested certain information from the tax administration in relation to the tax situation of the insolvent company under his administration, in order to examine the possibility of bringing insolvency avoidance claims against the competent tax office.
19. The applicant made such a request on the basis of the Law on freedom of information. The applicant asked, in particular, for information concerning the potential enforcement measures and the actual enforcement measures taken against the company, the payments received from that company, and the date on which the tax authorities became aware of that company’s insolvent status. The applicant also asked for extracts of the company’s accounts in relation to all taxes for the period between March 2014 and June 2015.
20. The applicant’s access to information was rejected by the tax office. The applicant challenged that decision before the Verwaltungsgericht (Administrative Court, Germany), which essentially upheld the action. The appeal by the Land was dismissed by the Oberverwaltungsgericht (Higher Administrative Court, Germany). That court took the view that the right to information, enshrined in the Land’s Law on freedom of information, was not ruled out by the specific provisions on tax secrecy. Although in general, such type of information could be subject to tax secrecy, that was certainly not the case with regard to the insolvency administrator of the company to which the information related.
21. The Land then brought an appeal on a point of law before the Bundesverwaltungsgericht (Federal Administrative Court). That court, harbouring doubts as to the correct interpretation of Article 23(1) of Regulation 2016/679 — a provision to which the provisions on tax secrecy make reference — decided to stay the proceedings and to refer the following questions to the Court of Justice for a preliminary ruling:
‘(1) Does Article 23(1)(j) of Regulation [2016/679] also serve to protect the interests of financial authorities?
(2) If so, does the wording “the enforcement of civil law claims” also cover the defence of the financial authority against civil law claims and must such claims already have been submitted?
(3) Does the provision of Article 23(1)(e) of Regulation [2016/679] relating to the protection of an important financial interest of a Member State in taxation matters allow a restriction of the right of access under Article 15 of Regulation [2016/679] in relation to the defence of civil law insolvency avoidance claims against the financial authority?’
22. Written observations in these proceedings have been submitted by the Land of North Rhine-Westphalia, the Czech, German and Polish Governments and the European Commission.
23. The three questions referred by the Bundesverwaltungsgericht (Federal Administrative Court) concern the interpretation of Article 23(1), points (e) and (j), of Regulation 2016/679. However, it is common ground that that provision is not directly applicable to the situation at issue in the main proceedings. As the referring court explains, Article 23(1) has been made indirectly applicable by the national legislature, through a renvoi made by the relevant national provisions.
24. In those circumstances, before answering the questions referred, a preliminary procedural issue should be addressed: does the Court have jurisdiction to answer the questions referred in the present case?
25. The parties who submitted observations in these proceedings hold different views in that respect: whereas the Land of North Rhine-Westphalia, the Polish Government and, to some extent, the Commission expressed doubts as to the jurisdiction of the Court, the Czech and German Governments took the opposite stance.
26. Against that background, this Opinion is structured as follows. First, I shall examine the Court’s jurisdiction in the present case (A). To that end, I will set out the Dzodzi case-law (1), recalling its origins and expansion (a), and the limitations that the Court has progressively introduced (b). In the light of the problems which may arise from a broad application of that case-law, I shall then propose that the Court consolidate it (2). Next, I shall apply the proposed legal framework to the present case, concluding that, in my view, no useful guidance can be drawn from the EU legislation invoked for the specific issue faced by the national court (3). Having come to the conclusion that the Court does not have jurisdiction in the present case, I will deal with the substance of the three questions referred only very briefly and within the limits of what is indeed logically possible (B).
A. Jurisdiction of the Court
1. The Dzodzi case-law
(a) Origin and expansion
27. The origins of the Dzodzi case-law may be found in Thomasdünger. (6) In that case, the referring court asked the Court for an interpretation of a specific position in the Common Customs Tariff (‘the CCT’). However, the case concerned the importation of goods into Germany from another Member State and was thus outside the scope of the CCT. The reason behind the reference was that the German authorities had, in the relevant domestic rules, referred to the positions set out in the CCT for other purposes.
28. In his concise Opinion, Advocate General Mancini proposed that the Court decline jurisdiction. In his view, the Treaty did not empower the Court to rule where the (then) Community provisions to be interpreted affected the interests of the parties not directly but per relationem. In his view, the Court lacked jurisdiction with regard to provisions which come into consideration solely because a national authority has decided, freely and unilaterally, to adopt them as a frame of reference in order to determine its own rules. (7)
29. In its judgment, however, the Court did not engage with the issue. The Court made reference to the presumption of relevance of questions referred and then turned to the merits of the questions. (8)
30. An explicit affirmation of its own jurisdiction to answer references in cases such as those at issue came in Dzodzi. (9) The case concerned the right of residence in Belgium of the spouse (of Togolese nationality) of a deceased Belgian citizen who had never made use of his freedom to work or reside in another Member State. The EU legislation referred to by the Belgian court clearly did not apply and the case lacked any cross-border element. However, the referring court pointed out that, under national law, the spouse of a Belgian national was to be treated as if he or she were a Community national. Accordingly, that court asked the Court whether Mrs Dzodzi would have the right to reside and remain in Belgium if her husband had been a national of a Member State other than Belgium.
31. In his Opinion, Advocate General Darmon recommended that the Court decline jurisdiction. In particular, he stressed that the unity and consistency of the Community legal order is unaffected by situations falling outside its field of application, regardless of the possible similarity of the relevant national provisions with the Community provisions whose interpretation was sought. (10)
32. The Court did not follow the Advocate General’s recommendation. The Court based its jurisdiction on three main considerations: (i) the presumption of relevance, (ii) the text of the Treaty not expressly excluding jurisdiction, and (iii) the interest of the Community that ‘in order to forestall future differences of interpretation, every Community provision [is] given a uniform interpretation irrespective of the circumstances in which it is to be applied’. (11)
33. Dzodzi went on to become settled case-law. The approach was applied not only in cases which concerned the fundamental freedoms, but also in cases which dealt with, inter alia, the common agricultural policy, (12) consumer protection, (13) taxation, (14) social policy, (15) and competition law. (16)
34. Even though the factual and legal background of those cases varied, in most of them, two elements seemed to be of particular importance for the Court: (i) the fact that the national rules faithfully reproduced the relevant EU provisions, (17) and/or (ii) the express (or at any rate clear) intention of the national legislature to harmonise national law with EU law. The latter requirement was considered met, inter alia, where, in regulating purely internal situations, domestic legislation adopted the same solutions as those adopted in EU law, (18) in order, in particular, to avoid discrimination against its own nationals or any distortion of competition, (19) or so as to provide for a single legal framework in comparable situations. (20)
35. The subsequent case-law has remained rather ambiguous as to the precise scope of the Dzodzi extension of jurisdiction. The principles stemming from that case-law seem to have been applied very generously in a number of cases.
36. In particular, in some cases the Court went as far as accepting vague, indirect or implied references to EU law. That was the case, for example, in BIAO, where the national provision ‘did not reproduce the [relevant EU provisions] verbatim’ but the national government and the referring court agreed that the Court’s judgment would be binding at national level. (21) In the same vein, in BAT, the Court gave a ruling on the basis of the fact that ‘domestic legislation ha[d] indisputably aligned itself, for the resolution of internal situations, with the solutions used in Community law’, although the specific provision at issue did not make an express reference to EU law in that regard. (22)
37. Similarly, in Kofisa, the Court accepted a reference where the national rule governing internal situations did not refer to a specific provision of EU law, but only to the relevant EU ‘customs legislation’. (23) By the same token, in Schoonbroodt, the Court gave judgment where the relevant national provisions merely referred to ‘principles of Community law’. (24) Finally, in Ostas, the Court declared its jurisdiction under the Dzodzi case-law, subject to verifications by the referring court that there was a direct and unconditional renvoi. (25)
38. Interestingly, in Federconsorzi and in Fournier, (26) the Court affirmed jurisdiction even in the circumstances where the renvoi to the relevant EC provisions were not included in any national legislation but only in private-law contracts or agreements entered into by public authorities.
39. Moreover, in Kofisa and Poseidon, the Court gave its ruling, on the assumption that that ruling would be binding on the referring court. The Court observed that there was nothing in the file to indicate that the national court was empowered to depart from the Court’s interpretation of the relevant EU provisions. (27) The Court went even further in Fournier, where it accepted jurisdiction after having noted that the national provisions at issue fell in an area not covered by the relevant directive and, accordingly, the terms used in the national provisions did ‘not necessarily have to have the same meaning as those used in the [relevant] Directive’. The Court held that it was ‘for the national court, as the only court with jurisdiction to interpret [the national provisions at issue], to give to the terms used in [those provisions] the meaning which it considers appropriate, without being bound in that regard by the meaning which must be attributed to the same expression as used in the Directive’. (28)
40. Finally, in some cases, the Court did not elaborate on why, in line with the Dzodzi case-law, it had jurisdiction to answer the reference. It merely repeated the case-law without explaining how the principles stated were applicable to the facts of the case. (29)
(b) The limits
41. Despite becoming well established, the Dzodzi case-law continued to attract criticism from various Advocates General: Jacobs in Leur-Bloem (30) and in BIAO, (31) Ruiz-Jarabo Colomer in Kofisa, (32) and Tizzano in Adam. (33) In essence, the learned Advocates General (i) found the fact that the Treaties do not expressly exclude jurisdiction unconvincing, given the fundamental principle of attribution of competences enshrined therein, (ii) questioned the existence of an actual interest of the Union to ensure a consistent interpretation, (iii) emphasised the unconventionality of interpreting EU rules outside their proper context, and (iv) doubted the binding nature of the Court’s response in those circumstances. On that basis, those Advocates General invited the Court to abandon the Dzodzi case-law or, at any rate, to apply it restrictively.
42. The Court never followed the suggestion to abandon the Dzodzi case-law. However, it has introduced certain qualifications to it over time.
43. First, in Kleinwort Benson, following the suggestion of Advocate General Tesauro, (34) the Court declined jurisdiction on the ground that the national law at issue did not provide for ‘a direct and unconditional renvoi’ to the relevant provisions of EU law but merely took them as a model, without wholly reproducing their terms. Moreover, the Court noted that the national law was clear in that the interpretation provided by the Court would not necessarily have been binding upon the national court. (35)
44. Although the ‘direct and unconditional renvoi’ test was not always followed in the years subsequent to the delivery of the judgment in Kleinwort Benson, (36) the Court has progressively applied it more rigorously. It thus dismissed cases where the renvoi was unclear, unproven or too generic, (37) and where the renvoi did not imply that the answer given by the Court to the reference would bind the referring court. (38) The Court, instead, accepted references where the documents before the Court made it clear that the relevant EU provisions were made directly and unconditionally applicable by national law. (39) In a number of cases, the Court accepted the reference only after being assured that the interpretation of the EU provisions sought from the Court would be binding upon national authorities. (40) Where necessary, the Court did not hesitate to look at the travaux préparatoires of the national laws to verify that the national legislature did indeed have the intention to treat EU and national situations in the same manner. (41)
45. Second, in Ullens de Schooten the Court emphasised that, in cases which concern the fundamental freedoms, its jurisdiction to answer questions referred in purely internal situations is an exception. It then went on to systematise the matter, by listing four sets of circumstances in which it is nonetheless competent to answer the references (the Dzodzi case-law being one of them). Importantly, the Court also pointed out that in a situation which is confined in all respects within a single Member State, ‘it is for the referring court to indicate to the Court, in accordance with the requirements of Article 94 of the Rules of Procedure of the Court, in what way the dispute pending before it, despite its purely domestic character, has a connecting factor with the provisions of EU law … that makes the preliminary ruling on interpretation necessary for it to give judgment in that dispute’. (42)
46. The Court’s recent case-law (both pre- and post-Ullens de Schooten) appears to confirm that the Court is ready to check more rigorously whether referring courts have duly provided the Court with all the information required to positively establish its jurisdiction. (43)
47. Third, there is Nolan. In that case, the Court found that the Dzodzi case-law did not apply where ‘an EU measure expressly provide[d] a case of exclusion from its scope’. The Court held that ‘it cannot be stated or presumed that there was an interest of the Union that, in an area excluded by the EU legislature from the scope of the measure which it adopted, there should be a uniform interpretation of the provisions of that measure’. (44)
48. True, Nolan could be read as stating that the Dzodzi rule does not apply each time the EU provisions referred to by national law expressly exclude situations such as those in the main proceedings. However, if it were interpreted in that manner, Nolan would be a unicum in the case-law. (45)
49. In fact, in subsequent cases the Court has made clear that such a (far-reaching) interpretation of Nolan is incorrect. In E, the Court declined to follow the same logic, noting that ‘the case which gave rise to that judgment was characterised by particularities which did not apply to the case in the main proceedings’. (46) More recently, in G.S. and V.G., the Court emphasised that its jurisdiction cannot ‘vary depending on whether the scope of the relevant [EU] provision was limited by a definition of the cases to which it refers or by means of certain exclusions from its scope, since both legislative techniques may be used interchangeably’. (47) In particular, the Court explained that the Dzodzi case-law is intended ‘to allow the Court to rule on the interpretation of provisions of EU law, irrespective of the circumstances in which those provisions are applicable, in situations which the authors of the Treaties or the EU legislature did not consider it appropriate to include within the scope of those provisions’. (48)
50. Therefore, I think that Nolan should rather be understood as a case in which the Court declined jurisdiction because the EU provision, whose interpretation was sought, had been ‘used’ by the national legislature in a context that was too removed from the original one. Indeed, not only did the EU provision not apply to the situation at issue ratione personae, but also the legal context in which the national rule operated varied considerably.
2. Consolidating and clarifying the Dzodzi case-law
51. Notwithstanding the criticism to which it has been subject, the Court has consistently confirmed the Dzodzi case-law. However, as is apparent from the overview set out above, the conditions under which a case will be accepted, even if actually outside the scope of an EU measure, and the exact limits to such extension, are far from clear.
52. I believe that the Court should seize the opportunity offered by this case to bring at least some clarity on this matter. (49) In the absence of a more precise framework, national courts lack guidance as to when they can refer questions to the Court on the interpretation of EU provisions that are only indirectly applicable in the main proceedings. It is hardly necessary to point out, in this context, that each reference rejected on procedural grounds inevitably entails a sub-optimal use of time and resources for both the referring court and the Court of Justice.
53. In the following passages, I shall set out the conditions which, in my view, should be satisfied in order for a reference made outside the scope of an EU measure to be acceptable under the Dzodzi approach. Two of those conditions — a twofold substantive condition and a procedural condition — already stem from the Court’s case-law (a). I suggest expressly confirming that there is a third, substantive, condition in addition to those two, which relates to the interest in conceptual uniformity sought by such a reference (b).
(a) A direct and unconditional renvoi and the national court’s duty to explain the reference
54. A first substantive condition — which is actually a twofold condition — was first set out in Kleinwort Benson and has more recently been confirmed in a strand of cases: national law must contain a ‘direct and unconditional renvoi’ to the EU provisions whose interpretation is sought from the Court. That naturally begs the question: when is a renvoi direct and unconditional?
55. The term ‘direct’ should be understood, in my view, as meaning that the reference must be specific and unambiguous, as opposed to a general (or generic) reference. (50) That requirement is certainly met by national provisions which contain an express renvoi to EU provisions that are identified or easily identifiable. However, it cannot be excluded that references which are not found in the domestic provision itself, but in other documents — such as the acts which accompany the legislation (or other preparatory works), or in the implementing legislation — may be deemed sufficiently plain and clear to that end. (51)
56. In turn, the term ‘unconditional’ appears to mean that the EU provisions referred to are ‘applicable without limitation to the situation at issue in the main proceedings’, (52) which implies that the referring court cannot depart from the interpretation provided by the Court of Justice. (53) This reading of the term ‘unconditional’ is borne out by the abovementioned case-law in which the Court stressed the importance that its judgment be binding in the main proceedings. (54)
57. After Ullens de Schooten, it has also become clear that there is a procedural condition to be satisfied in order for the Court to accept a reference in a Dzodzi case. The Court’s jurisdiction in situations where EU law does not directly apply to the facts of the case constitutes an exception and, as such, must be narrowly construed. Moreover, a clear and detailed knowledge of the relevant national laws by the Court is all the more important in Dzodzi cases, as the relevance and necessity of an answer by the Court may not be readily apparent. (55) It is thus the task of the referring court to explain to the Court why, notwithstanding the fact that the EU provisions in question are not directly applicable in the main proceedings, the Court’s jurisdiction is triggered. Failing to provide pertinent information on that point prevents the Court from simply assuming its jurisdiction. (56)
(b) The interest in ‘conceptual uniformity’
58. There is, in my view, an additional condition that must be satisfied to trigger the Court’s jurisdiction in Dzodzi cases. Although some traces of that condition may be identified in the case-law, (57) I must admit that the Court has, so far, not expressly referred to it. Yet, this additional condition appears to stem from the very logic underpinning the Dzodzi case-law.
59. The main justification for the Court’s jurisdiction in Dzodzi cases — which the Court has repeated, almost as a mantra, throughout its case-law — is that, in cases of renvoi, it is in the interest of the Union to ensure a uniform interpretation of the relevant EU provisions ‘in order to forestall future differences of interpretation’. By implication, that ‘uniformity of interpretation’ must, on the structure of those cases, have referred to the interest of the Union to maintain internal uniformity within a Member State and not, what would indeed be the more traditional instance in EU law, the uniform interpretation of EU law across its Member States. Otherwise, it is rather difficult to see exactly how the uniformity of EU law could be threatened by individual Member States unilaterally maintaining different rules outside the scope of EU law at national level.
60. Nevertheless, I fail to see why each and every reference to provisions, principles or concepts of EU law that may be found in Member States’ laws would necessarily give rise to an interest, on the part of the Union, in having (what could be referred to as) ‘conceptual uniformity’.
61. I can understand an interest in conceptual uniformity, and not only for the Union but also, even more so, for the Member State, in not having a situation where the two sets of rules operate, at national level, in a functionally and legally comparable context. That is likely to be the case where those rules pursue the same objective and concern the same subject matter. In those — and only in those — situations, it seems to me both desirable and possible to interpret the two sets of rules consistently.
62. Indeed, several Advocates General have warned the Court about the dangers of interpreting a rule out of its context, or of applying it to a set of facts that differ from those the EU legislature had in mind. (58) I cannot but join in expressing the same reservations. However, where the context of the two sets of rules is essentially analogous, those risks are arguably smaller.
63. That said, some clarification may be needed with regard to the requirement that the domestic and EU provisions concern the same subject matter. Indeed, in so far as the former regulates a matter which falls outside the scope of the latter, that requirement cannot, obviously, be intended to require perfect identity, but more as requiring contiguity, proximity or close similarity of their objects.
64. That requirement is likely to be met where the national authorities decide to extend the reach of the EU rules to ‘neighbouring’ situations, in order to treat EU-governed situations and purely internal situations alike. That is indeed the case, in particular, where the relevant domestic rules are so to say ‘one step removed’, be it ratione materiae, ratione personae, ratione loci or ratione temporis, from the EU legislative framework to which they make a renvoi.
65. Examples from the case-law help to give a better understanding of that concept. In Dzodzi, the relevant EC and national rules clearly concerned the same matter: the acquisition of the right of residence for spouses of, respectively, EU and Belgian nationals. The Belgian court thus asked the Court whether Ms Dzodzi would have been entitled to that right, had she fallen ratione personae within the scope of the relevant EC rules. (59) In Leur-Bloem, the Dutch legislature had, when transposing the EU provisions on taxation applicable to mergers, divisions, transfers of assets and exchanges of shares concerning companies of different Member States, broadened the scope of those rules to also cover mergers between two Dutch companies. (60) In SGI and in Solar Electric Martinique, French legislation had essentially made applicable the EU VAT rules to overseas departments and territories, despite the fact that the latter were expressly excluded from the scope of the relevant EU directives. (61) In Europamur, the relevant Spanish laws had extended the scope of the EU rules on unfair business-to-consumer commercial practices to also regulate business-to-business practices. (62) In several cases, the Court has answered questions on the proper construction of Article 101 TFEU where the agreements or practices in question did not affect competition within the internal market but could have potentially breached national competition rules. (63)
66. Even in many cases where the Court may have generously assessed the existence of a direct and unconditional renvoi — such as in Fournier, Ostas or BIAO (64) — the fact remains that the national rules in question had merely realised a ‘one-step’ extension of the scope of the relevant EU rules. (65)
67. However, the further national legislation moves from the context in which the EU provisions have been conceived and operate, the more tenuous the interest of the Union (and of the Member State concerned) in ensuring conceptual uniformity becomes, and the weaker the basis for the Court’s jurisdiction. The Court could perhaps use its jurisdiction to interpret an EU rule on the road transport of pigs if a Member State were to extend the scope of the rule to include the road transport of sheep. But would the Court then declare itself also competent, with a reference to Dzodzi, if a Member State were to extend those rules, or just a few selected provisions of those rules, based on a clear and unconditional renvoi, to the interstellar transports of humans?
68. Obviously, nothing precludes Member States’ authorities from taking inspiration from existing EU rules and borrowing those rules — or some principles, concepts and terms used therein — in order to regulate other matters. Yet, their creativity in grafting new things onto the EU rules cannot have, as a consequence, an abnormal and limitless extension of the Court’s jurisdiction.
69. Thus, the underlying logic is one of steps. As a rule of thumb, taking just one step outside the existing EU legal framework, while preserving the overall logic of that framework, may not pose a problem. However, a number of small steps for the national legislature suddenly becomes a rather giant leap for the Court, which is then asked, in essence, to give judgment indirectly in a case which has very little to do with the original EU law instrument.
(c) Interim summary
70. In short, in order for the Court to accept jurisdiction in a case in which the EU law at issue is being applied before the national court by virtue of its national extension beyond the originally conceived scope of application, three conditions must be fulfilled.
71. First, national law must contain a direct and unconditional renvoi to the EU provision whose interpretation by the Court is sought, which makes that provision of EU law not only clearly applicable to the case at hand, but also means that the guidance provided by the Court will be binding for the national court in that case.
72. Second, the EU rules extended by the national law must still operate within a functionally and legally comparable context, where there remains an interest in preserving conceptual uniformity, and within which the interpretation of EU law provisions may still be of some practical assistance to the referring court.
73. Third, the referring court is obliged to explain clearly how the two requirements stated above are met in the case at hand, setting out the relevant provisions of national law.
74. It is in the light of this framework that I shall now assess whether the Court has jurisdiction in the present case.
3. The present case
75. First, starting with the last procedural condition, the information required for the Court to rule on its jurisdiction is clearly set out in the order for reference. Indeed, the referring court exhaustively explains why, in its view, the Court has jurisdiction in accordance with the Dzodzi case-law. The Court does not need, therefore, to resort to any assumptions concerning national law in that regard.
76. Second, it is undisputed that the relevant national law, the AO, makes a ‘direct and unconditional’ renvoi to the provisions of Regulation 2016/679. An express renvoi to the provisions of that regulation is included in Paragraphs 2a, 32b, and 32e of the AO. In addition, Paragraph 32c(1)(2) of the AO states that the terms ‘the establishment, exercise or defence of civil law claims or … in the defence of civil law claims’ must be interpreted in the light of Article 23(1)(j) of Regulation 2016/679. (66) The intention to align the two sets of rules also transpires from the explanatory memorandum of that law. Furthermore, it is also common ground that the Court’s answer would be binding upon the referring court.
77. Third, however, I have strong doubts that there is any interest to ensure conceptual uniformity. Indeed, it seems to me that the EU and national provisions in question — when assessed both at macro level (in the context of the legal instrument to which they belong), and at micro level (focusing on the specific provisions alone) — neither pursue the same objective nor concern the same subject matter.
78. To begin with, it is hardly necessary to stress the significant differences, in terms of content and objective, of the different legislative frameworks of which the two provisions form part. Article 23(1) is a provision of the General Data Protection Regulation: a corpus of rules governing the processing of personal data in the single market. (67) That regulation, first and foremost, lays down limits to the use of data processing in order to protect the fundamental rights of data subjects.
79. By contrast, Paragraph 32c(1)(2) is a provision of Germany’s Tax Code, a very different legal instrument altogether. The provisions included in that instrument, including those on data processing, are geared towards ensuring uniform and lawful taxation and the safeguarding of tax revenue.
80. Moreover, the two specific provisions, whether taken together or by themselves, are also different in terms of objectives and content.
81. Article 23 of Regulation 2016/679 provides for situations in which the Union or the Member States are allowed to introduce restrictions to the rights normally granted to data subjects guaranteed in Chapter III of Regulation 2016/679 (for example, the rights of access, to rectification and to erasure), and the corresponding obligations placed on the controllers (such as information obligations). In the situations listed therein (to be interpreted restrictively), certain public or private interests can limit the fundamental right of individuals to the protection of personal data. (68)
82. By contrast, Paragraph 32c(1)(2) of the AO — included in the First Part, Fourth Chapter (‘Data protection and tax secrecy’) of the AO — is part of a set of rules aimed at regulating when the German tax authorities can (or cannot) disclose or make use of taxable persons’ data which they have gained access to in the context of tax-related procedures. Within that dimension, as also evidenced by the present case, those rules in effect act as restrictions or limitations to an individual’s right of access to information held by public authorities.
83. More specifically, the referring court explains that Paragraph 32c(1)(2) of the AO was introduced in order to remedy a disadvantageous status of the tax administration in the context of insolvency proceedings. Under German law, an action for revocation brought in the context of insolvency proceedings constitutes a civil law action which takes place before civil courts. German insolvency law places private law creditors and public law creditors on an equal footing, so that public law claims, such as taxes and social security contributions, do not have priority.
84. Nevertheless, as the referring court further points out, before Paragraph 32c(1)(2) of the AO was enacted, the tax authorities were, in practice, in a worse position than any other private creditor. Indeed, thanks to an access-friendly national case-law with regard to the rules on the freedom of and access to information, insolvency administrators were able to demand access to tax information relating to the insolvency debtor from the tax authorities. That enabled the administrators to decide in full knowledge of the fact whether to bring insolvency avoidance claims against those authorities. That possibility does not exist with respect to private creditors of the insolvency debtor given that those creditors are not subject to laws on the freedom of information.
85. It was in such a legislative context and practice that the German legislature introduced Paragraph 32c(1)(2) of the AO in order to prevent the tax authorities from being in a weaker position than other private creditors in insolvency proceedings.
86. Accordingly, the objectives of the two provisions are equally different: Article 23 of Regulation 2016/679 seeks to strike a fair balance between respect for natural persons’ fundamental rights affected by data processing (for example, private and family life) and the need to safeguard other legitimate interests in a democratic society (for example, national security). By contrast, Paragraph 32c(1)(2) of the AO aims at redressing a perceived imbalance with regard to tax authorities when avoidance claims are brought in the context of insolvency proceedings.
87. In addition, in order to arrive at such an objective, Paragraph 32c(1)(2) of the AO, far from simply extending the reach of Article 23(1) of Regulation 2016/679 to certain ‘neighbouring’ situations, ‘borrows’ a provision from the latter and through an intriguing legislative reference applies it to a rather different set of circumstances. That legislative construction is, however, only possible because there have already previously been, in other elements of the national legislative framework, several other extensions of the scope of Regulation 2016/679, ratione materiae and ratione personae.
88. First, Article 23(1) of Regulation 2016/679 applies only in respect of natural persons, whereas Paragraph 32c(1)(2) of the AO applies in respect of all natural and legal persons. In fact, in the present case access to information is sought in relation to tax data of a legal person. This is, obviously, no small detail: the balancing exercise required by Article 23(1) may not necessarily be carried out in the same manner in respect of data pertaining to a legal person, to which Regulation 2016/679 is not even applicable, thereby implying no balancing or legislative evaluation with regard to such a situation. The interest of a natural person to protect his own privacy and that of his family can hardly be compared to the interest of a legal person who may have to protect data concerning, for example, its business, its organisation or its fiscal position.
89. Second, as the German Government explained, under national law the insolvency administrator is a ‘third party’ (69) for the purposes of access to data of the insolvent debtor. Therefore, the insolvency administrator cannot be regarded as acting on the basis of rights which the data subject (the company under administration) has transferred to him. Yet, Article 23(1) of Regulation 2016/679 is concerned with rights of data subjects and obligations of data controllers. That provision simply does not concern third-party access to information held by public authorities.
90. Third, Article 23(1) of Regulation 2016/679, unlike Paragraph 32c(1)(2) of the AO, is not concerned with restrictions to requests for access to information, held by the public authorities, based on rules on transparency and openness. The EU provision sets out the permissible restrictions to the rights of data subjects (including the right of access), seeking to enforce their privacy rights against data holders and processors (regardless of their private or public nature).
91. Therefore, Article 23(1) of Regulation 2016/679 is a provision which permits specific exceptions to certain rights of data subjects, resulting from the system and the logic of the General Data Protection Regulation. It is intended to strike a fair balance between the fundamental rights of individuals, that is, natural persons, and certain vital public and private interests.
92. Paragraph 32c(1)(2) of the AO operates ‘a transplant’ of this provision seeking to provide a certain balance into a completely different legislative and factual context. The national provision in fact serves to curb an overly broad reach of the domestic rules on access to information in the possession of the public authorities, by depriving certain (third) parties of access to tax-related information, with a view to (re)establishing a balance between the parties in revocation actions in insolvency proceedings.
93. In conclusion, Article 23(1) of Regulation 2016/679 and Paragraph 32c(1)(2) of the AO form part of different bodies of law, pursue different objectives, and follow a completely different logic. I fail to see why there would be any interest, certainly on the part of the Union and possibly also of the Federal Republic of Germany, in ensuring a conceptual uniformity between them.
94. On that basis, I recommend that the Court declare that, in the present case, it lacks jurisdiction to answer the questions referred.
95. However, since it is my duty to (fully) assist the Court, I will briefly deal with the substance of the questions referred, with the significant caveat that, from all the foregoing, I do not think that this Court could provide the referring court with any useful answer for the purpose of the dispute pending before it.
96. Metaphorically speaking, it is as if an ice hockey referee were to be asked whether the exception to the prohibition on icing is applicable in a situation where both teams are shorthanded but only one goaltender leaves his goal crease and moves in the direction of the puck, by someone playing chess in the audience of the ice rink for the purpose of deciding whether he or she could perhaps castle with his or her queen. The answer to such a question is certainly likely to be that the ice hockey rules do not preclude such a move, but I suspect that is because, it is safe to assume, those rules have nothing to say on attempts at castling with a queen in chess.
B. The substance of the questions referred
97. By its three questions, the referring court seeks essentially to know whether a national provision — such as Paragraph 32c(1)(2) of the AO — that limits the right of access to personal data held by the tax authorities, when that data may be used to bring insolvency avoidance claims against those authorities, is compatible with Article 23(1) of Regulation 2016/679.
98. As follows from the considerations developed above, the answer to that question cannot but be in the affirmative. Due to its different content, scope and aim, Regulation 2016/679 has nothing to say on the concrete legislative choice made by the German authorities to limit the scope of the national rules on freedom of information in the context of insolvency proceedings in order to re-establish equality between the private and public law creditors.
99. At any rate, and despite the logical difficulty encountered in ‘fitting’ the EU provision in question into the facts of the case, in the following I shall attempt to interpret it as if it were applicable in a situation such as that at issue in the main proceedings.
1. First question
100. By its first question, the referring court asks essentially whether Article 23(1)(j) of Regulation 2016/679 permits restrictions introduced in pursuance of an interest of the public authorities, as opposed to the protection of private parties’ interests.
101. The referring court points out that such a possibility is evidently assumed by the national legislature: Paragraph 32c(1)(2) of the AO expressly refers to Article 23(1)(j) of Regulation 2016/679. However, that court notes that certain scholars suggest that the situations envisaged in points (i) and (j) of Article 23(1) allow restrictions only in pursuance of a private interest, and thus cannot ‘cover’ the furthering of a public objective.
102. The Land and the Czech and German Governments take the view that the first question must be answered in the affirmative. With regard to the (abstract) issue of construction of points (i) and (j) of Article 23(1), I share their view. It does not follow, either from the wording or the objective of Article 23(1)(j) of Regulation 2016/679, that that provision would be inapplicable in relation to public authorities.
103. Point (j) refers only to ‘enforcement of civil law claims’, without any limitation as to the private or public nature of the parties to the proceedings. Nothing in the wording of Article 23(1) of Regulation 2016/679 (or in recital 73) excludes from the scope of that provision civil law actions in which public authorities take part, as either the applicant or as the defendant.
104. In addition, even if civil law claims are mostly brought between private persons, in pursuit of a private interest, I fail to see any reasonable ground on the basis of which the EU legislature would have wished to treat differently civil law proceedings in which the public authorities are parties. Any such rule, granting more rights to certain claimants over other claimants, would be rather odd in my view.
105. The aim of the rule laid down in point (j) is — I assume — to permit the EU or national legislature to decide that, in the context of proceedings for the enforcement of civil law claims, the specific rules on discovery prevail, in case of conflict, over the general rules stemming from data protection. Indeed, most jurisdictions provide for special regimes of discovery in the context of judicial procedures, including civil law procedures. However, Member States’ regimes vary significantly. Given the importance of such regimes, it is possible to assume that the EU legislature could decide that their application should not be affected by the rules on data protection. Logically, that must be so, irrespective of whether the parties are subjects of private or public law, and regardless of the private or public interest on which their action or defence is based.
106. The Commission, however, puts forward the opposite view. It argues that Article 23(1) of Regulation 2016/679 draws a distinction of principle between the exceptions aimed at safeguarding public interests, listed in points (a) to (h), and those aimed at safeguarding private interests, listed in points (i) and (j) of that article. It explains that point (j) was not in the predecessor of the current provision (Article 13 of Directive 95/46/EC (70)), and was only added by the current regulation. The Commission contends that, because it was added at the end of the list, and not at the beginning of the list, point (j), like point (i), must be aimed at safeguarding private interests.
107. The Commission further submits that the addition of point (j) was meant to codify the Court’s findings in Promusicae, (71) in which the Court had found an apparent gap in the provisions which corresponded to the current Article 23 vis-à-vis the capacity of (private) parties to enforce their rights before national courts. In its view, the amendment had a very specific and narrow scope.
108. However, I see no element, either in the text of Regulation 2016/679 or in the preparatory works, supporting the Commission’s arguments. With regard to the latter, in particular, I find it rather puzzling that the Commission did not submit any such document before the Court, or at least refer to them in its observations.
109. In any event, from the publicly available preparatory works, it seems that point (j) did not appear in the Proposal for a regulation tabled by the Commission in 2012. (72) That point was subsequently introduced by the Council. (73) However, in the documents that I was able to review, I found no trace of the alleged intention of the EU legislature to limit the scope of this amendment to enforcement claims brought by private parties. (74) Nor did I find any specific indication that the EU legislature structured the list of possible restrictions according to the public interest/private interest dichotomy suggested by the Commission.
110. I am also not persuaded by the argument according to which the scope of point (j) is supposed to be rather narrow, in so far as the EU legislature has merely intended to react to the judgment in Promusicae.
111. Point (j) may well have been prompted by the judgment in Promusicae. However, I fail to see why, once that issue had been ‘opened up’ by the judgment, the legislature had necessarily intended to limit the amendment to the specific facts of that dispute. Why limit the amendment to civil law claims brought by private parties in the pursuit of a private interest? As mentioned in points 104 and 105 above, any such rule would seem quite unreasonable.
112. I therefore see no reason to conclude that Article 23(1)(j) of Regulation 2016/679 permits introducing restrictions only where the enforcement of civil law claims is pursued by private individuals.
2. Second question
113. By its second question, the referring court asks whether the expression ‘enforcement of civil law claims’ in Article 23(1)(j) of Regulation 2016/679 also encompasses the defence against civil law claims and, if so, whether the existence of the claim must have already been established.
114. The referring court explains that Paragraph 32c(1)(2) of the AO establishes essentially the principle that, in a situation such as that in the main proceedings, the duties to provide information are to be guided solely by civil law rules. However, those (national) rules provide for discovery obligations only where an insolvency avoidance claim is established on its merits and where the proceedings concern only the further determination of the nature and extent of the claim. Accordingly, until a duty of restitution is established, the insolvency administrator can request information only from the insolvency debtor.
115. The referring court points out that the term ‘Durchsetzung’ used in the German version of Article 23(1)(j) of Regulation 2016/679 traditionally refers to the sphere of the claimant and is mainly used as a synonym for the enforcement or execution of a claim which has already been established on its merits. That term is in fact similar to the terms ‘enforcement’ in the English version of the regulation and ‘exécution’ in the French version of the regulation. On that basis, the referring court wonders whether the defence against civil law claims can be subsumed under the term ‘enforcement’. In that context, it draws attention to the fact that other provisions of the regulation refer to the ‘establishment’, ‘exercise’ or ‘defence’ of legal claims. (75)
116. If the expression ‘enforcement of civil law claims’ includes the defence by the tax authorities against such claims, this raises the further question — according to the referring court — of whether the claims (in this case, the insolvency avoidance claims) must already have been established, or whether it is sufficient for the information to be requested in order to examine such claims. Indeed, the wording of Paragraph 32c(1)(2) of the AO refers to the defence of the legal entity of the tax authority ‘against civil law claims established against it’. (76) That suggests that the claimant has already raised a claim vis-à-vis the opposing party, and that the merits of that claim have already been substantiated. However, a different textual interpretation is also possible. In addition, if the national rule were to exclude the right of access to tax data only in enforcement procedures, that rule would to a great extent be ineffective: the insolvency administrator would have already acquired the necessary information. The referring court would thus be inclined to read Paragraph 32c(1)(2) of the AO to the effect that the wording ‘established’ also encompasses ‘yet to be established’ or ‘possible’.
117. Also with regard to that second question, I find myself in agreement with the views expressed by the Land and the Czech and German Governments.
118. To begin with, it seems to me that the enforcement of civil law claims necessarily involves an assessment of the arguments put forward in support of the existence of the claim, and also the argument in support of the non-existence of the claim. An interpretation to the contrary would be incompatible with the principle of equality of arms. Indeed, it would upset the procedural balance between the parties to the judicial proceedings by favouring the claimant. Accordingly, the concept of ‘enforcement’ must encompass that of ‘defence’ against the claim brought by the claimant.
119. That conclusion is not called into question by the fact that other provisions of that regulation use the phrase ‘establishment, exercise or defence’ of a right before the courts. First, those provisions may have been drafted by different ‘hands’ at different times (as mentioned, point (j) was added by the Council at a later stage). Second, those other provisions do not include the term ‘enforcement’ among those used, and thus do not lend themselves to draw a meaningful comparison.
120. Following the same logic, I also take the view that the limitation provided for in Article 23(1)(j) of Regulation 2016/679 is not conditional on the requirement that civil law claims have already been established.
121. Admittedly, the term used in the regulation, in the various languages, (77) may be read as referring only to the enforcement phase of proceedings: that in which compliance by a party is obtained, if need be through coercion. However, that term may also be understood, more broadly, as referring to the start of proceedings in order to have one’s subjective right recognised and thus protected.
122. I think the latter interpretation should be preferred. I fail to see the logic of why the EU legislature would permit Member States to maintain their specific regimes on discovery only in certain types or stages of civil procedures and not in others. If reasons relating to the protection of the integrity and fairness of civil law proceedings allow Member States to introduce restrictions to the rights of data subjects (and obligations of data controllers), those rules should arguably apply, in principle, at any stage of the proceedings. (78)
123. The opposite interpretation would also seems counterintuitive: why allow Member States to limit access during the final (or enforcement) stage of the proceedings but not before? As the referring court rightly points out, the administrator would have acquired the information sought by then, and the rule restricting access would have become meaningless.
124. I thus conclude that the expression ‘enforcement of civil law claims’ in Article 23(1)(j) of Regulation 2016/679 also encompasses the defence against civil law claims, and is not limited to situations where the existence of the claim has already been established.
3. The third question
125. Finally, by its third question, the referring court asks, in essence, whether a national provision, such as Paragraph 32c(1)(2) of the AO, that limits the right of access to data held by the tax authorities, when that data may be used to bring insolvency avoidance claims against those authorities, is compatible with Article 23(1)(e) of Regulation 2016/679.
126. The referring court points out that Paragraph 32c(1)(2) of the AO does not refer to point (e), but to point (j), of Article 23(1). However, that court does not exclude that point (e) could nonetheless provide a valid basis for the national provision. In that connection, the national court wonders whether the objectives pursued by paragraph 32c(1)(2) of the AO — to place the tax authorities in an equivalent position to that of other creditors in case of insolvency avoidance claims, in order to ensure uniform taxation and safeguard tax revenue — may be considered an ‘important objective of general public interest’ within the meaning of point (e). Moreover, the referring court also asks whether the national provision at issue belongs to ‘taxation matters’, for the purposes of point (e), in so far as the subject matter of the dispute before it is not governed by tax law but by insolvency law.
128. Point (e) of Article 23(1) allows restrictions intended to safeguard ‘other important objectives of general public interest of … a Member State, in particular an important economic or financial interest of the Union or of a Member State, including monetary, budgetary and taxation matters’. (80) It thus follows from the very wording of Article 23(1)(e) of Regulation 2016/679 that Member States are entitled to introduce restrictions to certain rights conferred by that regulation in order to pursue economic interests in taxation matters.
129. Article 23(1)(e) of Regulation 2016/679 reproduces, in substance, the wording of Article 13(1)(e) of Directive 95/46, (81) in respect of which the Court held that ‘a limitation of the data protection conferred by Directive 95/46 for tax purposes is … expressly provided for by that directive’. (82)
130. It is true, as the Commission points out, that Article 23(1)(e) of Regulation 2016/679 lists a number of exceptions and, as such, it should be interpreted narrowly. Admittedly, as a general rule, this is the correct approach. However, point (e) is not in itself a restriction, but just the statement of a legitimate aim. By its very nature, a legitimate aim is open-ended in its wording. That is common to most interests listed in Article 23(1) of Regulation 2016/679 (notably, national security, defence, and public security). The reason is, clearly, that a narrow interpretation and application of the restrictions permitted by Article 23(1) is ensured by requiring fulfilment of the conditions set out in the first sentence of that provision: any restriction must (i) be introduced by way of a legislative measure, (ii) respect the essence of fundamental rights and freedoms, and (iii) be a necessary and proportionate measure in a democratic society.
131. Thus, the only requirement laid down in point (e) of Article 23(1) is that the interest which the Member State intends to protect is an ‘important’ one. Yet again, the regulation does not offer any indication as to what may (or may not) be regarded as ‘important’.
132. Personally, I would read ‘important’ as simply ‘deserving of protection’: an interest that, in so far as it allows for derogation from a number of EU provisions, is recognised as legitimate also in the EU legal system. Therefore, provided the interest pursued benefits the generality (objective of general public interest), and that it is not at odds with any rule or principle of EU law, or at any rate unjust or unfair, that interest clearly falls within the scope of point (e).
133. Having said that, I hardly need to point out that the objective of ensuring uniform taxation and safeguarding tax revenue are recognised as legitimate aims in the EU legal system. (83) However, could the specific manner in which Paragraph 32c(1)(2) of the AO pursues that objective go beyond the ‘safe space’ granted to Member States by Article 23(1)(e) of Regulation 2016/679?
134. I do not think so.
135. I fail to see why EU law, and more specifically Article 23(1)(e) of Regulation 2016/679, should be interpreted as precluding a national provision, such as Paragraph 32c(1)(2) of the AO, that seeks to place the tax authorities in an equivalent position to that of other creditors in the context of insolvency avoidance claims.
136. Apart from the fact, again, that it is rather difficult to infer any position of Article 23(1) of the Regulation 2016/679 on the issue of equality of public and private law creditors in insolvency proceedings, I would also point out that, in a number of other Member States, the tax authorities enjoy a privileged position in the context of insolvency proceedings. It is thus conceivable that the German legislature may take the view that, in those proceedings, its tax authorities should be (at least) no worse off than private creditors.
137. The Commission argues, however, that establishing equal treatment between the tax administration and private-law creditors in actions such as that in the main proceedings does not constitute a general public interest, but an interest proper to the State which cannot be weighed against the fundamental right of the data subject to have access to the data collected concerning him or her. Therefore, the Commission considers that Article 23(1)(e) of Regulation 2016/679 must be interpreted as precluding a national provision such as Paragraph 32c(1)(2) of the AO.
138. The rationale behind the distinction between ‘general public interests’ and ‘interests proper to the State’, and the precise contours of the two concepts, frankly escape me. Failing an explanation from the Commission on this point, and finding no trace of it in the text of Regulation 2016/679, I find the Commission’s arguments unpersuasive.
139. In the light of the above, I conclude that a national provision, such as Paragraph 32c(1)(2) of the AO, that limits the right of access to information held by the tax authorities, when that information may then be used to bring insolvency avoidance claims against those authorities, cannot be said to be incompatible with Article 23(1)(e) of Regulation 2016/679, but in practical terms primarily because the latter provision has nothing to say on that specific issue.
140. I propose that the Court declare that it lacks jurisdiction to answer the questions referred for a preliminary ruling by the Bundesverwaltungsgericht (Federal Administrative Court, Germany).
1 Original language: English.
2 Judgment of 18 October 1990 (C‑297/88 and C‑197/89, EU:C:1990:360) (‘Dzodzi’).
3 The Court itself referred to this strand of cases as ‘the Dzodzi line of cases’ in the judgments of 17 July 1997, Leur-Bloem (C‑28/95, EU:C:1997:369, paragraph 27), and of 17 July 1997, Giloy (C‑130/95, EU:C:1997:372, paragraph 23).
4 OJ 2016 L 119, p. 1.
5 Law that entered into force with effect from 25 May 2018.
6 Judgment of 26 September 1985 (166/84, EU:C:1985:373).
7 Opinion of Advocate General Mancini in Thomasdünger (166/84, EU:C:1985:208, points 1 and 2).
8 Judgment of 26 September 1985, Thomasdünger (166/84, EC:1985:373, paragraph 11).
9 Supra, footnote 2.
10 Opinion in Joined Cases Dzodzi (C‑297/88 and C‑197/89, EU:C:1990:274, points 8 to 11). The learned Advocate General took the same stance in his Opinion in Gmurzynska-Bscher (C‑231/89, EU:C:1990:276), delivered on the same day.
11 Dzodzi, paragraphs 29 to 43.
12 See, for example, judgment of 25 June 1992, Federconsorzi (C‑88/91, EU:C:1992:276).
13 See, inter alia, judgment of 12 July 2012, SC Volksbank România (C‑602/10, EU:C:2012:443).
14 See judgments of 3 December 1998, Schoonbroodt (C‑247/97, EU:C:1998:586), and of 11 January 2001, Kofisa Italia (C‑1/99, EU:C:2001:10).
15 See, inter alia, judgment of 7 November 2013, Isbir (C‑522/12, EU:C:2013:711).
16 See, among many, judgments of 11 December 2007, ETI and Others (C‑280/06, EU:C:2007:775), and of 14 March 2013, Allianz Hungária Biztosító and Others (C‑32/11, EU:C:2013:160).
17 See, to that effect, judgments of 4 December 2014, FNV Kunsten Informatie en Media (C‑413/13, EU:C:2014:2411, paragraph 19); of 26 November 2015, Maxima Latvija (C‑345/14, EU:C:2015:784, paragraph 13); and of 21 November 2019, Deutsche Post and Others (C‑203/18 and C‑374/18, EU:C:2019:999, paragraph 39).
18 Judgment of 16 March 2006, Poseidon Chartering (C‑3/04, EU:C:2006:176, paragraph 17).
19 See, among many, judgment of 17 July 1997, Leur-Bloem (C‑28/95, EU:C:1997:369, paragraph 32).
20 See, to that effect, judgments of 17 July 1997, Giloy (C‑130/95, EU:C:1997:372, paragraph 28), and of 21 July 2016, VM Remonts and Others (C‑542/14, EU:C:2016:578, paragraph 18).
21 Judgment of 7 January 2003 (C‑306/99, EU:C:2003:3, paragraph 92).
22 Judgment of 29 April 2004, British American Tobacco (C‑222/01, EU:C:2004:250, paragraph 41).
23 Judgment of 11 January 2001 Kofisa Italia (C‑1/99, EU:C:2001:10, paragraphs 18 to 33).
24 Judgment of 3 December 1998 Schoonbroodt (C‑247/97, EU:C:1998:586, paragraph 15).
25 Judgment of 14 January 2016 Ostas celtnieks (C‑234/14, EU:C:2016:6, paragraphs 20 and 21).
26 Respectively, judgments of 25 June 1992 (C‑88/91, EU:C:1992:276, paragraphs 2 and 3), and of 12 November 1992 (C‑73/89, EU:C:1992:431, paragraphs 13, 14 and 22).
27 Judgments of 11 January 2001, Kofisa Italia (C‑1/99, EU:C:2001:10, paragraph 31), and of 16 March 2006, Poseidon Chartering (C‑3/04, EU:C:2006:176, paragraph 18).
28 Judgment of 12 November 1992 (C‑73/89, EU:C:1992:431, paragraphs 22 and 23).
29 See, for example, judgments of 3 December 2015, Quenon K. (C‑338/14, EU:C:2015:795, paragraphs 15 to 19), and of 17 May 2017, ERGO Poisťovňa (C‑48/16, EU:C:2017:377, paragraphs 26 to 32).
30 C‑28/95, EU:C:1996:332.
31 C‑306/99, EU:C:2001:608.
32 Opinion in Kofisa Italia (C‑1/99 and C‑226/99, EU:C:2000:498, point 28 et seq.).
33 C‑267/99, EU:C:2001:190, points 22 to 35.
34 Opinion of Advocate General Tesauro in Kleinwort Benson (C‑346/93, EU:C:1995:17, points 18 et seq.).
35 Judgment of 28 March 1995 (C‑346/93, EU:C:1995:85, paragraphs 20 et seq.).
36 For an overview and critique, see Krommendijk, J., ‘Wide Open and Unguarded Stand our Gates: The CJEU and References for a Preliminary Ruling in Purely Internal Situations’, German Law Journal, vol. 18, 2017, pp. 1359 to 1394; and Arena, A., Le ‘situazioni puramente interne’ nel diritto dell’Unione Europea, Editoriale Scientifica, Naples, 2019, pp. 127 to 143 and 180 to 200.
37 See, for example, judgments of 18 December 2014, Generali-Providencia Biztosító (C‑470/13, EU:C:2014:2469, paragraph 25), and of 16 June 2016, Rodríguez Sánchez (C‑351/14, EU:C:2016:447, paragraph 66). See also order of 28 June 2016, Italsempione — Spedizioni Internazionali (C‑450/15, not published, EU:C:2016:508, paragraphs 21 to 23).
38 See, for example, order of 3 September 2015, Orrego Arias (C‑456/14, not published, EU:C:2015:550, paragraphs 24 and 25).
39 See, for example, judgment of 27 June 2018, SGI and Valériane (C‑459/17 and C‑460/17, EU:C:2018:501, paragraph 28).
40 See judgments of 7 January 2003, BIAO (C‑306/99, EU:C:2003:3, paragraph 92), and of 14 March 2013, Allianz Hungária Biztosító and Others (C‑32/11, EU:C:2013:160, paragraphs 18 and 22).
41 Judgments of 7 November 2013, Isbir (C‑522/12, EU:C:2013:711, paragraph 29), and of 21 November 2019, Deutsche Post and Others (C‑203/18 and C‑374/18, EU:C:2019:999, paragraph 40).
42 Judgment of 15 November 2016 (C‑268/15, EU:C:2016:874, paragraphs 47 to 55, especially 53 and 55).
43 See, for example, judgments of 7 July 2011, Agafiţei and Others (C‑310/10, EU:C:2011:467, paragraph 43); of 20 March 2014, Caixa d’Estalvis i Pensions de Barcelona (C‑139/12, EU:C:2014:174, paragraphs 46 and 47); of 24 October 2019, Belgische Staat (C‑469/18 and C‑470/18, EU:C:2019:895, paragraphs 24 and 25); and of 30 January 2020, I.G.I. (C‑394/18, EU:C:2020:56, paragraphs 47 to 54). See also orders of 3 July 2014, Tudoran (C‑92/14, EU:C:2014:2051, paragraphs 41 and 42), and of 12 May 2016, Sahyouni (C‑281/15, EU:C:2016:343, paragraphs 27 to 31).
44 Judgment of 18 October 2012 (C‑583/10, EU:C:2012:638, paragraphs 32 to 57).
45 The only decision that may, perhaps, be read as following Nolan, albeit with an a contrario reasoning, is the judgment of 26 March 2020, Kreissparkasse Saarlouis (C‑66/19, EU:C:2020:242, paragraphs 25 and 26).
46 Judgment of 13 March 2019 (C‑635/17, EU:C:2019:192, paragraph 42).
47 Judgment of 12 December 2019, G.S. and V.G. (Threat to public policy) (C‑381/18 and C‑382/18, EU:C:2019:1072, paragraph 47).
48 Judgment of 12 December 2019, G.S and V.G. (Threat to public policy) (C‑381/18 and C‑382/18, EU:C:2019:1072, paragraph 47). Emphasis added.
49 The truth to told, I actually believe, like my learned predecessors quoted throughout this Opinion, that the Dzodzi line of case-law is an anomaly and should be abandoned. Apart from that case-law being highly questionable in constitutional terms, especially today when, perhaps in contrast to the past, what falls ‘within the scope of EU law’ tends to be more vigorously guarded than before, the Dzodzi approach was also born at a time when the number of references coming from the national courts was much more moderate and the Court apparently did not mind taking on some extra work. But I also suspect that the Court is not yet ready to make such a step.
50 See, to that effect, judgments of 21 December 2011, Cicala (C‑482/10, EU:C:2011:868, paragraph 25), and of 7 November 2013, Romeo (C‑313/12, EU:C:2013:718, paragraphs 19 to 38). Cf. also Opinion of Advocate General Cruz Villalón in Allianz Hungária Biztosító and Others (C‑32/11, EU:C:2012:663, point 29).
51 See, similarly, Opinion of Advocate General Kokott in ETI and Others (C‑280/06, EU:C:2007:404, point 39), and Opinion of Advocate General Pikamäe in Joined Cases Deutsche Post and Others (C‑203/18 and C‑374/18, EU:C:2019:502, points 47 and 48).
52 Judgment of 21 December 2011, Cicala (C‑482/10, EU:C:2011:868, paragraph 27).
53 Order of 9 September 2014, Parva Investitsionna Banka and Others (C‑488/13, EU:C:2014:2191, paragraph 29).
54 Supra, point 44 of this Opinion.
55 See, to that effect, Ritter, C., ‘Purely Internal Situations, Reverse Discrimination, Guimont, Dzodzi and Article 234’, European Law Review, 31, 2006, pp. 690 to 710, at 709; and Iglesias Sanchéz, S., ‘Purely Internal Situations and the Limits of EU law: A Consolidated Case Law or a Notion to be Abandoned?’, European Constitutional Law Review, 14, 2018, pp. 7 to 36, at 31.
56 See, to that effect, Opinion of Advocate General Wahl in Joined Cases Venturini and Others (C‑159/12 to C‑161/12, EU:C:2013:529, points 54 to 62).
57 Indeed, some decisions of the Court may be read as suggesting that the existence of an actual interest of the Union in interpreting the provisions at stake in order to forestall future differences in interpretation must be ‘established’ (see especially judgment of 7 July 2011, Agafiţei and Others (C‑310/10, EU:C:2011:467, paragraph 42); see also judgments of 12 July 2012, SC Volksbank România (C‑602/10, EU:C:2012:443, paragraphs 87 and 88), and of 19 October 2017, Europamur Alimentación (C‑295/16, EU:C:2017:782, paragraph 32)) or at any rate ‘conceivable’ (judgment of 19 October 2017, Solar Electric Martinique (C‑303/16, EU:C:2017:773, paragraph 29)).
58 See supra, point 41 of this Opinion.
59 Dzodzi, paragraphs 29 to 43.
60 Judgment of 17 July 1997, Leur-Bloem (C‑28/95, EU:C:1997:369, paragraphs 31 and 32).
61 Respectively, judgments of 19 October 2017, Solar Electric Martinique (C‑303/16, EU:C:2017:773, paragraph 29), and of 27 June 2018, SGI and Valériane (C‑459/17 and C‑460/17, EU:C:2018:501, paragraph 28).
62 Judgment of 19 October 2017, Europamur Alimentación (C‑295/16, EU:C:2017:782, paragraph 31).
63 See, inter alia, case-law referred to supra, in footnote 17.
64 Respectively, judgments of 12 November 1992, Fournier (C‑73/89, EU:C:1992:431, paragraph 23) (integration of a concept included in the EU directive on insurance against civil liability for the use of motor vehicles in an agreement regulating the same subject matter); of 7 January 2003, BIAO (C‑306/99, EU:C:2003:3, paragraphs 68 to 77) (extension of an EU accounting rule to certain situations falling outside the scope of the relevant EU directive); and of 14 January 2016, Ostas celtnieks (C‑234/14, EU:C:2016:6, paragraphs 17 to 19) (extension of EU public procurement rules to a contract the market value of which was below the threshold set out in the relevant EU directive).
65 But see also judgment of 12 July 2012, SC Volksbank România (C‑602/10, EU:C:2012:443, paragraphs 85 to 93), where the Court accepted jurisdiction to interpret the EU provisions on credit agreements for consumers despite the fact that those provisions did not apply in the main proceedings ratione temporis and ratione materiae. Nevertheless, it is indisputable that both sets of rules applied to very similar situations, as they concerned the same subject matter (credit agreements) and pursued the same objective (to protect consumers).
66 Quoted in full above in point 11 of this Opinion.
67 See, in particular, recitals 2 and 4 and Article 1 of Regulation 2016/679.
68 See especially recitals 4 and 73 of Regulation 2016/679. More generally, see Feiler, L., Forgó, N., Weig, M., The EU General Data Protection Regulation (GDPR) — A Commentary, GLP, 2018, pp. 138 to 140; Moore, D., ‘Comment to Article 23 — Restrictions’, in Kuner, C., Bygrave, L., Docksey, C., Drechsler, L. (eds), The EU General Data Protection Regulation (GDPR) — A Commentary, Oxford University Press, 2020, pp. 543 to 554; and Ehmann, E., Selmayer, M. (eds.), Datenschutz-Grundverordnung: Kommentar. 2nd edn, C.H.Beck, 2018, pp. 467 to 469.
69 For that concept in Regulation 2016/679, see Article 4, point (10).
70 Directive of the European Parliament and of the Council of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on the free movement of such data (OJ 1995 L 281, p. 31).
71 Judgment of 29 January 2008 (C‑275/06, EU:C:2008:54, paragraphs 51 to 55).
72 Proposal for a Regulation of the European Parliament and of the Council on the protection of individuals with regard to the processing of personal data and on the free movement of such data (General Data Protection Regulation) COM(2012) 11 final of 25 January 2012 (pp. 54 and 55).
73 See, inter alia, Council documents 9398/15 of 1 June 2015 (pp. 145 and 146), and 9565/15 of 11 June 2015 (pp. 107 and 108).
74 Leaving aside the broader structural question on how far such an alleged intention should in fact be relevant if it is in no way reflected in either the text of the legal act or its recitals — in detail see my Opinion in BV (C‑129/19, EU:C:2020:375, points 119 to 123).
75 Article 9(2)(f), Article 17(3)(e), Article 18(1)(c) and (2), Article 21(1), second sentence, and Article 49(1)(e) of Regulation 2016/679.
76 Emphasis added.
77 That expression is similar, inter alia, to the Czech (‘vymáhání’), Spanish (‘ejecución’), Finnish (‘täytäntöönpano’), Italian (‘esecuzione’), Portuguese (‘execução’) and Slovak (‘vymáhanie’) versions of the regulation.
78 See, by analogy, the reasonably broader logic embraced by the Court in the interpretation of the expression ‘may be disclosed in civil or commercial proceedings’ as to whether that proceedings must in fact already be pending in the judgment of 13 September 2018, Buccioni (C‑594/16, EU:C:2018:717, paragraph 35).
79 Points 93 to 95 of this Opinion.
80 Emphasis added.
81 Point 106 of this Opinion.
82 Judgment of 27 September 2017, Puškár (C‑73/16, EU:C:2017:725, paragraph 42).
83 See, to that effect, judgments of 13 December 2005, Marks & Spencer (C‑446/03, EU:C:2005:763, paragraph 51), and of 13 March 2007, Test Claimants in the Thin Cap Group Litigation (C‑524/04, EU:C:2007:161, paragraph 68).