download in pdf

 

Martina Melcher*

 

Substantive EU Regulations as Overriding Mandatory Rules?

 

Introduction

The main research question as expressed in the title of this paper is primarily a question of coordination. It addresses the peculiar relationship between directly applicable, uniform substantive EU law and the EU Private International Law (PIL) regulations, i.e. the Rome Regulations.

On the one hand, PIL rules indicate the applicable substantive law and should therefore naturally be applied before any substantive law rule. On the other hand, certain substantive EU regulations want to be applied irrespective of the (substantive) law indicated by PIL rules. Moreover, PIL rules, such as are contained in the Rome I and Rome II Regulations, typically refer to the substantive law of a state rather than to substantive EU law. The latter would thus be applicable only as part of the law of a Member State, but could not be applied if the relevant EU PIL Regulations referred to a third-state law.

This paper examines whether substantive EU regulations qualify as overriding mandatory provisions pursuant to Article 9 Rome I Regulation and Article 16 Rome II Regulation to ensure their application irrespective of the otherwise applicable third-state law. For this purpose, this paper first confronts the idea that uniform substantive EU law enjoys a ‘categorical precedence’ over uniform EU PIL rules (part II). The following part of the paper explores whether and under what circumstances substantive EU regulations may materially qualify as overriding mandatory provisions, pursuant to Article 9 Rome I and Article 16 Rome II Regulation (part III). Finally, the paper discusses the possible application of such rules as overriding mandatory provisions and the role of special conflict-of-law provisions in more detail (part IV). The General Data Protection Regulation (GDPR),[1] the Passengers’ Rights Regulation[2] and the (abolished) draft Common European Sales Law (CESL)[3] serve as examples.

 

I Categorical Precedence of Uniform Substantive EU Law?

The question whether unified or non-unified substantive law (possibly of a third state) applies is a typical conflict-of-law question, which has to be asked whenever there is no worldwide unified law.[4] Hence, a conflict-of-law rule is required to determine the coordination between substantive EU regulations and the EU PIL regulations, such as the Rome I and II Regulations.

While it is true that substantive EU regulations would take precedence due to the principle of primacy before any national law, substantive EU regulations and EU PIL regulations are on the same step of the normative hierarchy.[5] For this reason, the precedence of EU substantive law cannot be based on the general principle of primacy.

Instead, EU substantive law regulations are often accorded a ‘categorical precedence’ in the way international uniform law is often considered to take priority over (national) conflict-of-law rules.[6] The justification of such a categorical precedence, if one is provided, varies. Zweigert and Drobnig suggest that (international) uniform substantive provisions generally supersede conflict-of-law provisions in their territorial and material scope of application due to their superiority in substance (sachliche Überlegenheit).[7] In particular with regard to EU uniform law, an assumed categorical precedence of substantive regulations is based on the principle of efficiency, which demands a primacy in application of uniform substantive law before uniform conflict-of-law rules in order to respect the special scope of application of such substantive rules.[8] Moreover, the categorical precedence is hardly ever explained or discussed methodologically. Often, uniform substantive law is simply applied without prior consultation of the conflict-of-law rules by way of ‘principle’. Given that the ‘principle of categorical precedence’ provides the resolution to a conflict between uniform substantive law and uniform PIL rules, methodologically, it must be qualified as a conflict-of-law rule. Hence, in an EU context, it does not give precedence to uniform substantive law over conflict-of-law rules as such, but rather represents a conflict-of-law principle which gives priority to uniform EU substantive law over the general EU conflict-of-law rules contained in the EU PIL regulations. If one follows this approach, there would be no need for substantive EU regulations to qualify as overriding mandatory provisions (or any other EU PIL rule to apply) as they would take precedence in any case.

However, while such a precedence of uniform substantive law is indeed teleologically convincing as a result, in an EU context, there is no need for a stand-alone conflict-of-law principle of categorical precedence, which must in essence be based on a lex specialis reasoning. In fact, the general EU conflict-of-law rules contained in the EU PIL regulations already contain several provisions which can possibly be used to resolve a conflict with substantive EU regulations, namely the rules on overriding mandatory provisions (see part III) and the lex specialis rules (see part IV).

 

II Material Qualification as an Overriding Mandatory Provision

According to Article 9 (1) Rome I, overriding mandatory provisions are ‘provisions the respect for which is regarded as crucial by a country for safeguarding its public interests, such as its political, social or economic organization, to such an extent that they are applicable to any situation falling within their scope, irrespective of the law otherwise applicable to the contract under this regulation’.[9] In accordance with the EU PIL system, the applicable law is determined in such situations by the general conflict-of-law rules, namely the rules of the Rome I or II Regulation, but is then superseded by the respective overriding mandatory provision of the law of the forum (or the lex causae) in its scope of application. With regard to EU regulations, several issues deserve a more in-depth consideration.

 

1 Individual Private Law Rules of an Internationally Mandatory Character

First, it is always an individual rule, the overriding mandatory character of which needs to be established rather than the character of the regulation as a whole. This follows already from the wording (‘provisions’ rather than ‘law’) on the one hand and on the other hand from the exceptional character of overriding mandatory provisions, which requires a strict interpretation,[10] and a case-by-case analysis. Thus, any suggestions that qualify whole EU regulations, such as the GDPR,[11] as overriding mandatory laws are (much) too sweeping.[12]

Second, Article 9 Rome I and Article 16 Rome II do not specify whether they refer to private law and/or public law rules. In the literature, this issue is often sidestepped in so far as it is argued that the public or private law nature of an overriding mandatory provision is irrelevant and such a distinction is not to be drawn.[13] This is particularly convincing with regard to EU law, which does not formally know a public/private divide. In fact, EU regulations often contain public and private law provisions side by side.

However, if one takes the material scope of application of the Rome Regulations into consideration, it is evident that they apply only to (contractual and non-contractual) obligations in civil and commercial matters [see Article 1 (1) respectively]. Hence, only substantive rules that deal with or haven an effect on civil and commercial matters are at risk of being disapplied due to the application of another substantive law. Public law rules that do not have such an impact are not affected by a divergent choice of law; they are applied pursuant to their own (public) conflict-of-law rules (e.g. principle of territoriality). Article 83 GDPR, regarding administrative fines imposed by a supervisory authority, might serve as an example for such a (purely) public law provision.

Given the scope of application of the Rome Regulations, particular private law rules (Sonderprivatrecht)[14] may qualify as overriding mandatory rules as long as they do not serve exclusively or predominantly to balance the interests between private persons (see below).[15] What this means precisely, is still subject to much discussion (and will probably continue to be so). National provisions that have been considered by the ECJ and/or national courts to have an overriding mandatory character concern, for example, the termination of a commercial agency contract by notice and related entitlements to indemnities (ECJ Unamar) and the reduction of allowances and remuneration of officials and other employees of public authorities (ECJ Nikiforidis). As regards, for example, the GDPR, this means that provisions such as Article 6 and 7 or 82, which deal with the lawfulness of data processing based on consent and the right to compensation and liability, address questions of a civil or commercial nature and therefore risk being put aside if a non-EU Member State law were chosen to be applicable. Similarly, the Passengers’ Rights Regulation contains only rules and obligations regarding civil and commercial matters, which could thus possibly qualify as overriding mandatory provisions due to their private law nature.

Apart from the above mentioned private law provisions and public law provisions dealing with the exercise of public power,[16] there is a huge grey area. In particular, ‘supporting’ provisions that do not grant a right themselves or cannot be used as a basis for a particular claim are sometimes difficult to categorise as public or private and are often of particular relevance to a private law issue despite their rather public law nature, e.g. public restrictions regarding the purchase of land. To summarize, unfortunately, the public/private divide, which is in itself quite ambiguous, can hardly be used to identify overriding mandatory rules in an unequivocal way – notably, in an EU context where the public/private divide is particularly blurred.

Third, irrespective of their private or public law character, a mandatory rule must be of such importance that it justifies a departure from the (otherwise) applicable law.[17] In other words, the rule demands to be applied as an internationally mandatory rule.[18] A strict interpretation is required.[19] The internationally mandatory character of a rule is to be determined not only according to the wording of the rule but also by taking into account the general structure of the law, its objectives and the circumstances under which it was adopted.[20] In the Da Silva case, which concerns a particular limitation rule, the ECJ emphasises that such rules are principally governed by the law applicable to a non-contractual obligation pursuant to Article 15 (h) Rome II Regulation.[21] Pursuant to this logic, convincing arguments are needed in order to qualify substantive provisions as overriding mandatory provisions if they are expressly part of the scope of the law applicable. If one considers EU rules to fall within the scope of Article 9 Rome I and Article 16 Rome II, it is particularly difficult to determine the internationally mandatory character of such a rule. In essence, one has to ask whether the rule in question requires supremacy also regarding third country law. In view of Article 3 (4) Rome I, it is certainly clear that EU law is not per se of an internationally mandatory nature. However, EU regulations often require a very broad territorial application. For example, Article 3 GDPR envisages the application of the Regulation to the processing of personal data in the EU and, in certain situations, also to their processing by a controller or processor not established in the EU. Such a broad scope of application often hints at an internationally mandatory character.[22]

A broad scope of application is not sufficient in itself, however. An overriding mandatory provision must have an internationally mandatory character ‘to safeguard public interests’. In other words, it must serve (primarily) a public interest purpose that goes beyond a mere legislative interest. Both elements are closely linked, as a special public interest is often derived from an express international scope of application and a broad scope of application is often thought to suggest a particular public interest. In general, rules that are of an internationally mandatory character due to public interest are also referred to as ‘public order legislation’.[23] Given its arguments in the Unamar case, the ECJ would probably qualify Article 82 GDPR (right to compensation and liability) as an overriding mandatory provision. It requires an international application within the (rather broad) confines of Article 3 GDPR[24] and there is a particular public interest in the right to data protection, given that it even qualifies as a fundamental right.[25] In contrast, Article 7 Passengers’ Rights Regulation also contains a right to compensation in the event of denial of boarding or cancellation and an international scope of application, but it is doubtful whether respecting this provision is crucial for safeguarding the EU’s public interests.

As can be seen, both, the public/private divide as well as the particular public interest character of a provision, as ‘identifying features’ of an overriding mandatory rule are rather ambiguous concepts. They give much scope of interpretation to courts and legislators, so that (almost) any provision that shall be applied mandatorily also in cross-border situations to provide an appropriate legal frame may be identified as such.[26]

 

2 Are Article 9 Rome I and Article 16 Rome II Applicable to EU Law?

Moreover, irrespective of their substance, it is still unclear whether mandatory rules in EU regulations may even qualify as overriding mandatory provisions in the sense of Article 9 Rome I and Article 16 Rome II.[27] Actually, the wording of these provisions seems to cover national law only, as they refer to ‘provisions of the law of the forum’ [Article 9 (2) Rome I], and to ‘provisions the respect for which is regarded as crucial by a country’ [Article 9 (1) Rome I]. Whereas EU law could be qualified as part of the ‘law of the forum’ in a Member State, this was not necessarily intended by the legislator, given that the Rome Regulations clearly distinguish between substantive national law of the forum and substantive EU law in other places, such as Article 3 (3) and (4) Rome I. Furthermore, the EU is certainly no country but a regional institution sui generis, so that – prima facie – it does not matter that the EU regards the respect for a certain EU provision as crucial. However, one could argue that the Member States are bound by the principle of loyalty and that therefore any EU regulation, the respect for which is regarded as crucial by the EU, is necessarily as essential to the Member States as it is to the EU.

Furthermore, from a teleological and systematic point of view, there is no need for EU legislation to be addressed by Article 9 Rome I and Article 16 Rome II. Especially with regard to substantive EU regulations, the literature often regards the application of the exception for overriding mandatory rules as pointless, because the substantive EU regulations ought to supersede otherwise applicable national rules anyway, due to the EU principle of supremacy of application.[28] While this is certainly true for the national law of the Member States, which is superseded by (substantive) EU law, a supremacy-based application of directly applicable secondary EU law is less certain with regard to the law of third countries.[29] The assumed non-applicability of an automatic supremacy in these cases is also supported by Article 3 (4) Rome I, which addresses EU law in all its forms including that of an EU regulation. Such a rule would be superfluous if the principle of supremacy applied to directly applicable EU law also in the above situations. Even if the general principle of supremacy is inapplicable, however, there is no need for Article 9 Rome I or Article 16 Rome II to be applied to EU regulations (and also direct transpositions of EU directives) if one agrees that Article 23 Rome I and Article 27 Rome II cover such situations (see below part IV).

As regards national rules that implement EU directives, the ECJ case law ostensibly clarified that such rules may qualify as overriding mandatory provisions.[30] At a closer look, however, the Unamar and Da Silva cases, in which the ECJ explicitly referred to Article 7 Rome Convention and Article 9 Rome I respectively, actually concerned national provisions that had been drafted in the context of minimum harmonisation and were based on the legislative autonomy of the Member States, i.e. they did not (precisely) reflect corresponding EU stipulations. In Unamar, the ECJ stipulated that, with regard to national law transposing the EU rules in question, ‘reference must be made to Article 7 of the Rome Convention’.[31] This is however a rather general instruction at the beginning of its analysis of the specific situation and not an analysis of the situation at hand. In fact, the national provisions under scrutiny for conformity with Article 7 Rome Convention did not only transpose EU law but went beyond (‘goldplating’) the demands of the EU directive, which did not contain very precise instructions to begin with, thereby constituting (genuine) national law rather than EU law. In other words, the final assessment and ruling of the ECJ, which suggests a link to Article 7 Rome Convention actually concerns national (overriding mandatory) provisions rather than EU (overriding mandatory) provisions. Similarly, in Da Silva, the ECJ effectively deals with a national rule in a minimum harmonisation context, which represents the Member State’s reaction to a general allowance of the EU Directive in question to maintain or bring into force provisions that are more favourable rather than a simple prescribed implementation of EU law. Hence, the Unamar and the Da Silva cases do not support the idea that EU regulations may qualify as overriding mandatory law in accordance with Articles 9 Rome I and 16 Rome II.

It is only in the Ingmar case that the national rule in question actually reflects the EU specification to a significant extent. Due to the temporal setting of the Ingmar case, however, the ECJ[32] did not (have to) refer to Article 9 Rome I or Article 7 Rome Convention, but merely argued that the purpose of the rules in question requires that they be applied ‘irrespective of the law by which the parties intended the contract to be governed’.[33] Contrary to the prevailing view in academic literature, this is certainly not an unequivocal qualification as overriding mandatory law.[34] Hence, although the ECJ case law links the Ingmar and Unamar cases argumentatively (i.e. by citation), thus suggesting their coherence,[35] (at least) some ambiguity remains as to whether the ECJ would have relied on these rules if the case had fallen into their scope rationae temporis.

 

III Application as Overriding Mandatory Provision

Irrespective of the material character of an EU provision as an overriding mandatory provision, its application as such is a different question. If a special conflict-of-law rule takes precedence and results in the application of the directly applicable, unified substantive EU law instead of the (otherwise) applicable (national) law, there is no need for this law to be applied on the basis of Article 9 Rome I or Article 16 Rome II – even if it would (also) qualify as overriding mandatory provision.[36] In this sense, Article 9 Rome I and Article 16 Rome II represent only the ‘second exit’ for rules that cannot ‘leave’ the Rome Regulations through the ‘first exit’, namely Article 23 Rome I or Article 27 Rome II.[37]

 

1 Overriding Mandatory Provisions and the Lex Specialis Principle

Article 23 Rome I and Article 27 Rome II contain a lex specialis principle for rules on contractual and non-contractual obligations respectively. According to these provisions, the respective general PIL regulation (i.e. Rome I or Rome II) shall not prejudice the application of provisions of [Union] law which, in relation to particular matters, lay down conflict-of-law rules’ for contractual/non-contractual obligations.

As regards traditional reference rules (Anknüpfungsnormen), there is little doubt that they qualify as conflict-of-law provisions in this sense. Examples include Article 3 Regulation 3921/91[38] and Article 129 (2), 130 (2) EU trade mark Regulation[39]. However, it is disputed whether Article 23 Rome I and Article 27 Rome II also apply regarding the coordination with uniform substantive EU rules. For them to be applicable, the substantive EU regulations in question have to contain (scope of) application rules (also called Abgrenzungsnormen [delimitation rules][40]) and these application rules have to be encompassed by the autonomous EU notion of ‘conflict-of-law rule’, which is used in Article 23 Rome I and Article 27 Rome II.[41]

 

2 Application Rules as Special Conflicts Rules

(Scope of) Application rules have a dual function:[42] On the one hand, they (unilaterally) determine the (international territorial) scope of application of uniform law. They are similar to what is commonly understood as overriding mandatory law, insofar as the question posed is not which law is applicable but rather whether the rule in question is applicable according to its application rule. The applicable law is determined on the basis of the particular substantive rule rather than on the basis of the situation, which is characteristic for traditional conflict-of-law rules (i.e. reference rules) according to Savigny. On the other hand, application rules contain a coordination mechanism regarding conflicting rules. In this sense, traditional conflict-of-law rules and applicability criteria address exactly the same problem. Unlike reference rules, however, the conflict-of-law character of rules that determine the (territorial) scope of application of the instrument of which they are part is disputed.

First, the nature and character of the principal provisions of the Rome Regulations as reference rules argue in favour of a strict understanding of the term ‘conflict-of-law rule’, which does not cover application rules.[43] However, the wording of Article 23 Rome I and Article 27 Rome II is not limited to reference provisions and does not distinguish between various types of conflict-of-law rules. Furthermore, EU law contains conflict-of-law rules that deviate from the standard reference rules, such as Article 9 Rome I, which allows a national provision to apply unilaterally.

Second, application rules are often only ancillary provisions.[44] However, Article 23 Rome I and Article 27 Rome II do not refer to instruments consisting solely of conflict-of-law rules, but also encompass EU regulations that contain conflict-of-law rules, as they refer to (individual) provisions rather than whole instruments.[45] Thus, a single conflict-of-law rule could suffice, e.g. older consumer law directives containing only a single conflict-of-law rule are often referred to as examples.[46]

Third, the purpose of Article 23 Rome I and Article 27 Rome II is to allow an amendment for substantive rules that require the application of a different law.[47] Such a demand also characterises substantive EU regulations (or of individual provisions therein) that determine their territorial/international scope of application independent from the otherwise applicable law.[48] As a test, one could ask whether the substantive provisions in question want to be applied without reference to the general conflict-of-law rules. In this regard, the ECJ pays particular attention to the wording and objectives of the EU legislation in question, insofar as it asks whether the legislation intends to lay down conflict-of-law rules.[49] If this question is answered in the affirmative, such application rules must be qualified as (accessory) unilateral conflict-of-law rules (akzessorische Kollisionsnormen).[50] Primacy of uniform law can be justified, as application rules take precedence due to the lex specialis-principle.[51] With this in mind, Article 23 Rome I and Article 27 Rome II are not applicable if an EU legal instrument does not display any conflict-of-law character but rather requires the national legislator to devise protective measures in general.[52] Therefore, Article 28 Directive 2009/103, which allows the Member states to maintain or bring into force more favourable provisions, was not regarded as a conflict-of-law rules by the ECJ.[53]

Besides express provisions regarding the scope of application, Article 23 Rome I and Article 27 Rome II ought to encompass implicit application rules as well, in spite of referring to ‘provisions’.[54] To distinguish formally between express and implied application rules would result in an artificial distinction that cannot be justified. Moreover, Article 23 Rome I and Article 27 Rome II probably should not be limited to directly applicable EU rules, but should be understood to encompass also national law that transposes EU law.[55] Article 20 Rome Convention, the predecessor of the Rome I Regulation, even mentioned harmonised law explicitly; however, one must keep in mind that the legislator altered the wording of the provision. Also, EU Directives (and national law implementing such EU Directives) pose particular challenges, as they rarely contain unambiguous provisions and differing Member State transpositions cause further divergences (e.g. Article 3 E-Commerce Directive). [56]

 

3 Some Examples

Contrary to – for example – Article 14b Directive 2009/103,[57] which does not qualify as a conflict-of-law provision for subrogation between insurers according to the ECJ,[58] Article 3 GDPR expresses a request for the substantive rules of the GDPR to be applied despite the otherwise applicable national law. It defines independently a particular international scope of application for EU data protection law. Such a rule would be redundant if the general PIL provisions were to be nonetheless applied, as data processors could choose a foreign substantive law to apply and thus deselect the (probably) more protective EU law. Actually, the ECJ considers Article 4 (1) (a) Data Protection Directive 95/46, which precedes Article 3 GDPR, to be a conflict-of-law rule, as it refers to this rule in order to determine the applicable law in the Verein für Konsumenteninformation/Amazon Case.[59] Similarly, the Austrian Supreme Court considers Article 4 (1) (a) Data Protection Directive to be ‘a special rule’ in the context of Article 6 (1) Rome I.[60] Although it is true that – contrary to Article 4 Data Protection Directive – Article 3 GDPR does not indicate the applicable national law but rather the applicable EU law, both rules determine the scope of application of the substantive rules contained in the respective instrument. As such, they have a conflict-of-law character although they do not qualify as traditional all-sided reference rules and thus qualify as special conflict rules according to Article 23 Rome I and Article 27 Rome II.[61] To differentiate between both rules on account of their wording would seem artificial.

Similarly, the Passengers’ Rights Regulation contains, in its Article 3, an application rule that should be qualified as special conflict-of-law provision pursuant to Article 23 Rome I.[62] It stipulates in Article 3 (1) a that it shall apply to ‘passengers departing from an airport located in the territory of a Member State to which the Treaty applies’. It demands the application of its rules, irrespective of a dissenting choice of law or the substantive law applicable to the transportation contract. By qualifying Article 3 as a special conflict-of-law provision, one also avoids the categorisation of all substantive rules of the Passengers’ Rights Regulation as overriding mandatory provisions and an accompanying overextension of this notion.

As a final example, the – discarded – proposal for a Common European Sales Law (CESL) can be named. It also contains a provision that determines its international and territorial scope of application. However, according to its recitals 10 and 15, it was not intended that the CESL should be applied before the determination of the applicable law, but was meant to be subject to the applicable conflict-of-law rules. Notwithstanding the (convincing) critique issued by some scholars regarding the intended link between the applicability of a Member State’s law and the CESL as an optional 2nd regime[63] and the likely competence-inspired nature of this approach,[64] this shows that application rules should not qualify as conflict-of-law rules in a categorical manner, but subject to a case-by-case analysis only.

 

Conclusion

As this paper illustrates, a coordination of traditional PIL rules that determine the applicable law ‘from the facts’ (vom Sachverhalt aus) and substantive provisions, which lay down their own (international) scope of application (‘from the rule’; von der Norm aus) is challenging. In this paper it is suggested that application rules should qualify as leges specialis, which penetrate the general PIL connecting system (also) in EU law. This approach respects normative hierarchies and avoids an overloading of Article 9 Rome I and Article 16 Rome II and of the (ambiguous) notion of an ‘overriding mandatory provision’, which require a strict interpretation.[65] At the same time, it provides a way to deal with an increasing number of EU legislative acts that determine their own scope of application.

Despite these advantages, this approach also provides some challenges, in particular as regards questions of competence. If (scope of) application provisions may be qualified as conflict-of-law rules, one might argue in favour of Article 81 TFEU as an (additional?) competence base. Given the ancillary nature of these provisions, the assumption of an ancillary competence might however resolve this issue in a simple manner. It might prove more complex, though, with regard to the external competence of the EU if one applies this extended understanding of conflict-of-law provisions – as seems only reasonable – also to Article 25 Rome I and Article 28 Rome II regarding uniform substantive law in international conventions.

This research question therefore certainly merits further attention and a more in-depth discussion. In the end, however, the question of whether uniform substantive law is to be applied as overriding mandatory law, due to special conflict-of-law rules, or enjoys a principled ‘categorical precedence’ remains primarily a methodological one – the result is the same: substantive EU rules that demand to be applied, irrespective of the otherwise applicable (national) law, take precedence.

 

Theses

1. A (principle of) categorical precedence which gives priority to uniform substantive law over uniform PIL laws must be qualified as a conflict-of-law rule. Although, a methodological account is often missing, it must in essence be based on a lex specialis reasoning. References to such a general principle are superfluous, if explicit EU conflict-of-law provisions already ensure such a precedence.

2. The material qualification of a rule as an overriding mandatory provision should be distinguished from its application as an overriding mandatory provision.

3. Whether a provision is of an overriding mandatory nature has to be determined individually for each provision and cannot be generalised for a legislative act as such.

4. Article 9 Rome I and Article 16 Rome II encompass only overriding mandatory provisions that qualify as rules relating to civil and commercial matters. Other rules fall outside the scope of application of the Rome Regulations, so there is no actual conflict of (different substantive) laws as regards rules which do not relate to civil and commercial matters. However, the dividing line is blurred and a distinct and precise categorisation of rules on this basis is hardly possible.

5. Substantive rules that (primarily) address and regulate private interests cannot qualify as overriding mandatory provisions according to Article 9 Rome I and Article 16 Rome II. Only rules that are of an overriding mandatory nature due to the public interest they address are covered; but, unfortunately, the term ‘public interest legislation’ leaves much discretion and EU law generally seeks to advance particular social and economic interests.

6. Article 9 Rome I and Article 16 Rome II focus on the substance and intention of a rule and address national provisions rather than EU provisions.

7. Even if a rule materially qualifies as an overriding mandatory provision, it is not necessarily to be applied as such. Several coordination/conflict-of-law rules exist in the context of the Rome Regulations, such as Article 9 Rome I and Article 16 Rome II regarding overriding mandatory provisions and Article 23 Rome I and Article 27 Rome II regarding leges speciales.

8. Application rules that form part of EU substantive regulations (usually) qualify as leges speciales according to Article 23 Rome I and Article 27 Rome II. A case-by-case analysis is necessary.

9. If an overriding mandatory rule is backed up by a special conflict-of-law rule, Article 23 Rome I and Article 27 Rome II already stipulate their precedence, so there is no need to refer to Article 9 Rome I or Article 16 Rome II (or to employ a general principle of categorical precedence).


* MMag. Dr. Martina Melcher, M.Jur (Oxon), is assistant professor at the University of Graz, Faculty of Law, Institute of Civil Law, Foreign Private Law and Private International Law (e-mail: martina.melcher@uni-graz.at).

[1] Regulation (EU) 2016/670 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 [2016] OJ L119/1.

[2] Regulation (EC) No 261/2004 of the European Parliament and of the Council of 11 February 2004 establishing common rules on compensation and assistance to passengers in the event of denied boarding and of cancellation of long delay of flights [2004] OJ L 46/1.

[3] Proposal for a Regulation of the European Parliament and of the Council on a Common European Sales Law, COM (2011) 635.

[4] Jan von Hein, ‘Einheitsrechtliche Anwendungsnormen und Internationales Vertragsrecht’ in Normann Witzleb and others (eds), Festschrift für Dieter Martiny zum 70. Geburtstag (Mohr Siebeck 2014, Tübingen, 365–390) 375; Peter Mankowski in Ulrich Magnus, Peter Mankowski (eds), European Commentaries on Private International Law – ECPIL II (Otto Schmidt 2017, Köln) Article 25 Rome I para 21.

[5] Similarly, Johannes Schilling, ‘Materielles Einheitsrecht und Europäisches Schuldvertrags-IPR’ [2011] EuZW 776–781, 780.

[6] See, for example, Richard Bader, Koordinationsmethoden im Internationalen Privat- und Verfahrensrecht (Mohr Siebeck 2019, Tübingen) 57.

[7] Konrad Zweigert, Ulrich Drobnig, ‘Einheitliches Kaufgesetz und Internationales Privatrecht’ (1965) 29 RabelsZ 146–165, 148–150. See also Horst Eidenmüller and others, ‘The Proposal for a Regulation on a Common European Sales Law: Deficits of the Most Recent Textual Layer of European Contract Law’ Max Planck Private Law Research Paper No. 12/14, 21 who refer to a ‘general principle that uniform substantive law takes priority over uniform conflicts rules’.

[8] Schilling (n 5) 781.

[9] This definition applies also to the notion used in Article 16 Rome II, see Case C-149/18 Agostinho da Silva Martins ECLI:EU:C:2019:84, para 28.

[10] Case C-149/18 Agostinho da Silva Martins, para 29; Case C-184/12 Unamar, para 49.

[11] Marian Thon, ‘Transnationaler Datenschutz: Das Internationale Datenprivatrecht der DS-GVO’ (2020) 84 RabelsZ 24–61, 41.

[12] Similarly, Björn Steinrötter, ‘Kollisionsrechtliche Bewertung der Datenschutzrichtlinien von IT-Dienstleistern – Uneinheitliche Spruchpraxis oder bloßes Scheingefecht?’ [2013] MMR 691–694, 693.

[13] Dieter Martiny in Münchener Kommentar zum BGB (7th edn, C.H. Beck 2018, München) Article 9 Rom I-VO para 12; Ansgar Staudinger in Franco Ferrari and others (eds), Internationales Vertragsrecht (2nd edn, C.H. Beck 2011, München) Article 9 para 8.

[14] See, for example, Mario Giuliano and Paul Lagarde, Report on the Convention on the law applicable to contractual obligations, OJ [1980] C282, 1, Article 7 para 4. Rightly critical Robert Freitag, ‘Einfach und international zwingende Normen‘ in Stefan Leible and others (eds), Grünbuch zum Internationalen Vertragsrecht (Sellier 2004, München, 167–191) 190 ff; Felix Maultzsch, ‘Rechtswahl und ius cogens im Internationalen Schuldvertragsrecht’ (2011) 75 RabelsZ 60–101, 88 ff; Bonomi in Ulrich Magnus, Peter Mankowski (eds), European Commentaries on Private International Law – ECPIL II (Otto Schmidt 2017, Köln) Article 9 Rome I para 73 ff; Martiny (n 13) Article 9 Rom I-VO para 12.

[15] Axel Thorn in Thomas Rauscher (ed), Europäisches Zivilprozess- und Kollisionsrecht III (4th edn, Otto Schmidt 2016, Köln) Article 9 Rom I-VO para 11 with further references. See also Andreas Köhler, Eingriffsnormen – Der „unfertige Teil“ des europäischen IPR (Mohr Siebeck 2013, Tübingen) 22 ff.

[16] See Case C-645/11 Land Berlin ECLI:EU:C:2013:228, para 33.

[17] Case C-149/18 Agostinho da Silva Martins, para 31.

[18] Brigitta Lurger and Martina Melcher, Handbuch des Internationalen Privatrechts (Verlag Österreich 2017, Wien) paras 1/49, 4/50.

[19] Case C-149/18 Agostinho da Silva Martins, para 29; Case C-184/12 Unamar, para 49.

[20] Case C-149/18 Agostinho da Silva Martins, paras 30 and 31; Case C-184/12 Unamar, para 50.

[21] Case C-149/18 Agostinho da Silva Martins, para 25.

[22] See also Jan Lüttringhaus, ‘Das internationale Datenprivatrecht: Baustein des Wirtschaftskollisionsrechts des 21. Jahrhunderts‘ (2018) 117 ZVglRWiss 50–82, 72.

[23] Case C-184/12 Unamar ECLI:EU:C:2013:663, para 47.

[24] In contrast, Article 3 GDPR itself cannot qualify as an overriding mandatory provision because it only expresses the request of the substantive rules of the GDPR to be applied internationally. Dissenting Thon (n 11) 41.

[25] See also Maja Brkan, ‘Data Protection and Conflict-of-laws: A Challenging Relationship’ (2016) 2 EDPL 324–341, 334; Thorn (n 15) Article 9 Rom I-VO para 47a. As regards public law provisions Michael Müller, ‘Amazon and Data Protection Law – The end of the Private/Public Divide in EU conflict of laws? (2016) EuCML 215–218, 218. Regarding the GDPR: Christian Kohler, ‘Conflict of Law Issues in the 2016 Data Protection Regulation of the European Union’ (2016) 52 RDIPP 653–675, 661 ff; Manuel Klar in Jürgen Kühling and Benedikt Buchner (eds), Datenschutz-Grundverordnung (2nd edn, C.H. Beck 2018) Article 3 para 105; Lüttringhaus (n 22) 74.

[26] Similarly, Adrian Hemler, Die Methodik der “Eingriffsnorm” im modernen Kollisionsrecht (Mohr Siebeck 2019, Tübingen) 171.

[27] Affirmative (pro EU law inclusion) Martiny (n 13) Article 9 Rom I-VO para 27; Ulrich Magnus in Staudinger, BGB XIII (Sellier 2016, München) Article 9 Rom I-VO para 164; (rather) affirmative Bonomi (n 14) Article 9 Rome I para 51; rejecting Wulf-Henning Roth, ‘Handelsvertretervertrag und Rom I-Verordnung – Eine Skizze’ in Jörn Bernreuther and others (eds), Festschrift für Ulrich Spellenberg (Sellier 2010, München, 309–328) 320.

[28] Staudinger (n 13) Article 9 para 12 f; Ulrich Magnus in Staudinger, BGB XIII (Sellier 2016, München) Article 9 Rom I-VO para 35 und Einl IPR para 9.

[29] Bonomi (n 14) Article 9 Rome I para 51.

[30] Case C-381/98 Ingmar ECLI:EU:C:2000:605; Case C-184/12 Unamar, para 40.

[31] Case C-184/12 Unamar, para 41.

[32] Interestingly, the Advocate General did refer to the content of said rules anyway, see Case C-381/98 Ingmar ECLI:EU:C:2000:230, Opinion of AG Léger, para 88.

[33] Case C-381/98 Ingmar, para 25.

[34] See Hemler (n 26) 221 (also for further references).

[35] Similarly, Jan Lüttringhaus, ‘Eingriffsnormen im internationalen Unionsprivat- und Prozessrecht: Von Ingmar zu Unamar’ [2014] IPRax 146–152, 147.

[36] See also Hemler (n 26) 162, who states that overriding mandatory rules are based on a lex specialis principle.

[37] Peter Mankowski in Ulrich Magnus, Peter Mankowski (eds), European Commentaries on Private International Law – ECPIL II (Otto Schmidt 2017, Köln) Article 23 Rome I para 13. See also Felix Maultzsch in beck-online.Grosskommentar (1.8.2018) Article 9 Rom I para 203.

[38] Council Regulation No 3921/91 of December 1991 laying down conditions under which non-resident carriers may transport goods or passengers by inland waterway within a Member State, OJ 1991 L 373, 1.

[39] Regulation (EU) 2017/1001 of the European Parliament and the Council of 14 June 2017 on the European Union trade mark, OJ 2017 L 154, 1.

[40] Jan Kropholler, Internationales Einheitsrecht (Mohr Siebeck 1975, Tübingen) 190 ff [Abgrenzungsnormen (delimitation rules)]; Ulrich Drobnig, ‘Anwendungsnormen in Übereinkommen zur Vereinheitlichung des Privatrechts’ in Walter Stoffel and Paul Volken (eds), Mélanges en l’honneur d’Alfred von Overbeck (Editions universitaires Fribourg Suisse 1990, Fribourg, 15–30) 15 [Anwendungsnormen (application rules)]. Generally, Jan von Hein in Münchener Kommentar zum BGB (6th edn, C.H. Beck 2015, München) Einl IPR para 97. Another commonly used notion for such rules is statutistic conflict-of-law rules (statutistische Kollisionsnormen), see Karl Kreuzer, ‘Zu Stand und Perspektiven des Europäischen Internationalen Privatrechts’ (2006) 70 RabelsZ 1–88, 46; Rolf Wagner, ‘Normenkonflikte zwischen den EG-Verordnungen Brüssel I, Rom I und Rom II und transportrechtlichen Rechtsinstrumenten’ [2009] TranspR 103–109, 107.

[41] Affirmative Peter Mankowski, ‘Rechtswahlklauseln in Luftbeförderungs-AGB auf dem Prüfstand’ [2014] RRa 118–123, 123; Mankowski (n 37) Article 23 Rome I para 11 (both with reference to the Passengers Rights Regulation). See also Erik Jayme and Carl Nordmeier, ‘Multimodaler Transport: Zur Anknüpfung an den hypothetischen Teilstreckenvertrag im Internationalen Transportrecht – Ist § 452a HGB Kollisions- oder Sachnorm?’ [2008] IPRax 503–507, 507; von Hein (n 4) 375; Florian Eichel in Rainer Hüßtege, Heinz-Peter Mansel, Bürgerliches Gesetzbuch Rom-Verordnungen – HUP – EuErbVO VI (3rd edn, Nomos 2015, Baden-Baden) Article 27 Rom II-VO para 3; Maultzsch (n 37) Article 9 Rom I para 205; Schulze in beck-online.Grosskommentar (1.8.2018) Art 23 Rom I-VO Rz 18 und Art 27 Rz 13. Rejecting, Johannes Schilling, Das Internationale Privatrecht der Transportverträge (Mohr Siebeck 2016, Tübingen) 87 ff; see also Wagner (n 40) 107 (regarding Article 25 Rome I-VO); Leible in Rainer Hüßtege, Heinz-Peter Mansel, Bürgerliches Gesetzbuch Rom-Verordnungen – HUP – EuErbVO VI (3rd edn, Nomos 2015, Baden-Baden) Article 23 Rom I-VO para 8; Eva-Maria Kieninger in Franco Ferrari and others, Internationales Vertragsrecht (2nd edn, Beck 2011, München) Article 23 para 3 (regarding the Passengers Rights Regulation); Magnus in Staudinger, BGB XIII (Sellier 2016, München) Article 23 Rom I-VO para 17. Inconclusive Dieter Martiny in Münchener Kommentar zum BGB (7th edn, C.H. Beck 2018, München) Article 23 Rom I-VO para 8.

[42] See in detail Kropholler (n 40) 190.

[43] Wagner (n 40) 107 ff; Schilling (n 5) 778 ff.

[44] Wagner (n 40) 107 ff.

[45] Similarly Jayme and Nordmeier (n 41) 507 (regarding Article 25 Rom I-VO).

[46] Leible (n 41) Article 23 Rome I para 9. Regarding the possibly (exclusive) external competence of the EU also for uniform substantive law due to such a broad understanding of conflict of laws (also) in the context of Article 25 Rome I and Article 28 Rome II see Wagner (n 40) 108 (footnote 61); Schilling (n 5) 779. See also Dieter Martiny in Münchener Kommentar zum BGB (7th edn, C.H. Beck 2018, München) Article 25 Rom I-VO para 4.

[47] Roth (n 27) 323.

[48] Similarly, as regards overriding mandatory provisions, Hemler (n 26) 180.

[49] Case C-149/18 Agostinho da Silva Martins, para 38; Joined Cases C-359/14 and C-475/14 ERGO Insurance ECLI:EU:C:2016:40, para 39.

[50] Kreuzer (n 40) 46; Jan Kropholler, Internationales Privatrecht (6th edn, Mohr Siebeck 2006, Tübingen) 97; Maultzsch (n 37) Article 9 Rom I para 205.

[51] Extensively, Kropholler (n 40) 189 ff. See also Kieninger (n 41) Article 23 para 3; von Hein (n 40) Einl IPR paras 96–98.

[52] Thorn in Thomas Rauscher (ed), Europäisches Zivilprozess- und Kollisionsrecht III (4th edn, Otto Schmidt 2016, Köln) Article 23 Rom I-VO para 6. Similarly Stefan Leible and Matthias Lehmann, ‘Die Verordnung über das auf vertragliche Schuldverhältnisse anzuwendende Recht („Rom I“)’ [2008] RIW 528–543, 531.

[53] Case C-149/18 Agostinho da Silva Martins.

[54] Roth (n 27) 314.

[55] See also Case C-149/18 Agostinho da Silva Martins, para 36 ff; Joined Cases C-359/14 and C-475/14 ERGO Insurance, para 38 ff; Leible (n 41) Article 23 Rome I para 5; Eichel (n 41) Article 27 Rome II para 2.

[56] Roth (n 27) 315.

[57] Directive 2009/103/EC of the European Parliament and of the Council of 16 September 2009 relating to insurance against civil liability in respect of the use of motor vehicles, and the enforcement of the obligation to insure against such liability [2009] OJ L263/11.

[58] Joined Cases C-359/14 and C-475/14 ERGO Insurance, paras 38–42; generally Case C-149/18 Agostinho da Silva Martins, paras 36–42.

[59] Case C-191/15 Verein für Konsumenteninformation ECLI:EU:C:2016:612, paras 72–81. In detail Müller (n 25) 215 passim.

[60] OGH 2 Ob 155/16g, JBl 2018, 464 = jusIT 2018, 54 (Thiele; Mader) = ÖBl 2018, 249 (Handig).

[61] See Martina Melcher, ‘Es lebe das Territorialitätsprinzip?’ in Susanne Gössl and others (eds), Politik und Internationales Privatrecht (?) 129–147, 137 ff. See also Maultzsch (n 37) Article 9 Rom I para 271. Dissenting Brkan (n 25) 341 (differentiating between Article 4 Data Protection Directive and Article 3 GDPR).

[62] Similarly, Roth (n 27) 314.

[63] Eidenmüller and others (n 7) 21; von Hein (n 4) 375, 389.

[64] See Jan von Hein in Thomas Rauscher (ed), Europäisches Zivilprozess- und Kollisionsrecht III (4th edn, Otto Schmidt 2016, Köln) Article 25 para 8a.

[65] See also Mankowski (n 37) Article 23 Rome I para 15. In this regard, also the case C-381/98 Ingmar (and C-184/12 Unamar) and its broad understanding of an overriding mandatory provisions should be challenged.