Opinion 3/15 on the Marrakesh Treaty: ECJ reaffirms narrow ‘minimum harmonisation’ exception to ERTA principle
By Thomas Verellen
On Valentine’s Day 2017, the Grand Chamber of the ECJ issued its opinion on the competence of the EU to conclude the ‘Marrakesh Treaty to Facilitate Access to Published Works for Persons Who Are Blind, Visually Impaired, or Otherwise Print Disabled.’ As happens increasingly often, the Commission, on the one hand, and several Member States and the Council on the other, disagreed on the nature of the competence of the EU to conclude the agreement. The Commission considered the agreement to be covered entirely by the EU’s exclusive competences, whereas the Member States, and to a lesser extent the Council, argued that at least part of the agreement fell outside of the scope of those competences, and instead fell within the scope of the EU’s shared competences.
The distinction between exclusive and shared competences matters. Unless an agreement is covered entirely by the EU’s exclusive competences, it will most likely be concluded in the form of a mixed agreement, i.e. an agreement to which not only the EU, but also the Member States are parties. This typically is the case even when the agreement falls within the scope of the EU’s shared competences, as the Council considers that when the Commission proposes to negotiate and conclude an international agreement parts of which are covered by shared competences, the Council can opt not to exercise those competences with regard to part of that agreement, however small this part may be. In such an event, the Member States must fill the gap by exercising their own competences, rendering the agreement a mixed agreement.
This dynamic was at play also in the context of the Marrakesh Treaty, where the discussion revolves around the question of whether the EU is endowed with an exclusive or a shared competence to conclude the agreement. If the EU’s competence is exclusive, the agreement would be concluded by the EU alone; if it is shared, the Member States would opt for a mixed agreement.
The ECJ came down in favour of EU exclusivity. The Court rejected the Commission’s argument that the agreement was covered by the EU’s exclusive competence to conduct a common commercial policy, but did follow the Commission in its argument that the agreement is covered by an ERTA effect.
Two pleas were made in Opinion 3/15: one based on Article 207 TFEU, which confers on the EU an exclusive power to conduct a common commercial policy, and one based on Article 3(2) TFEU, which codified the ERTA principle. This post focusses on the latter plea. It explains and contextualises the ECJ’s treatment of the ‘minimum harmonisation’ exception to the ERTA principle first introduced in Opinion 2/91, invoked more recently in Neighbouring Rights (C-114/12), and today again in this Opinion. By rejecting the Member States’ argument that exceptions to rights established in Directive 2001/29 ought to be understood as minimum standards, the Court in Opinion 3/15 reaffirms its original understanding of the ‘minimum harmonisation’ exception. The refusal to endorse the Member States’ argument fits well within the ECJ’s broad reading of the ERTA principle in the post-Lisbon era, whereby the ECJ engages in a wholesale analysis aimed at determining whether the relevant ‘area’ is ‘largely covered.’
The Grand Chamber on ERTA
The ‘ERTA doctrine’ was introduced in the 1971 ERTA case. Today, the doctrine is given textual expression in Article 3(2) TFEU, which provides that ‘[t]he Union shall … have exclusive competence for the conclusion of an international agreement … in so far as its conclusion may affect common rules or alter their scope.’
In 1993, the ECJ introduced an exception to the ‘ERTA principle.’ In Opinion 2/91, the ECJ rejected the Commission’s call for ERTA-type exclusivity as the ILO convention was not liable to affect common EU rules at issue there. This was the case, the ECJ submitted, because both the internal common EU rules only provided for minimum standards of health and safety protection at work. The proposed ILO convention aimed to lift the bar higher (rather than lower). Doing so would not undermine the full effectiveness of the pre-existing internal EU rules, the ECJ held, also because the ILO Constitution itself allows its Members to aim for a higher level of protection than the level provided for in norms adopted in the framework of the ILO.
In Opinion 3/15, several Member States had invoked the ‘minimum harmonisation’ exception in response to the Commission’s argument that the Marrakesh Treaty is covered by an ERTA effect. The common rules, which in the Commission’s view justify an ERTA effect, are contained in Directive 2001/29/EC on the harmonisation of certain aspects of copyright and related rights in the information society. The Member States considered that this directive does not harmonise the exceptions and limitations to those rights, and therefore Member States had to provide for such exceptions and limitations. Intervening Member States thus advocated that, as in Opinion 2/91, since the common rules themselves authorise the Member States to act, no risk of ‘affectation’ in the meaning of the ERTA doctrine was present, and consequently the Member States had ‘retained’ a competence to act.
The ECJ was not persuaded by these arguments. As suggested by the Commission, the Grand Chamber distinguished the ‘minimum harmonisation’ exception introduced in Opinion 2/91 from the exceptions and limitations provided for in Directive 2001/29. The Court held that the latter were authorisations provided for by EU law. This means that the Member States may provide exceptions to the rights established in the Directive, but in doing so must respect EU law, in particular the objective of the Directive and the conditions the Directive attaches to the adoption of limitation and exceptions.
This arrangement differs from the arrangement at issue in Opinion 2/91, the Court continued. In that opinion, the EU wished to conclude an agreement in an area within which the EU itself did not possess a competence to harmonise legislation fully; its competence was limited to setting out minimum requirements (see Article 118a EEC). By contrast, when enacting Directive 2001/29, the EU legislator had enacted a ‘harmonised legal framework’ (para. 119). In doing so, the EU legislator had authorised the Member States to provide certain limitations and exceptions.
In short, in Opinion 2/91 the Member States acted on the basis of their own competence, whereas in Opinion 3/15 they acted as ‘trustees’ of the EU, charged with the responsibility of implementing EU law. It followed that the ‘minimum harmonisation’ exception did not apply. Since instead the Marrakesh Treaty fell within an ‘area that is already covered to a large extent by common EU rules’, the EU alone has the necessary competence to conclude the treaty.
The ‘minimum harmonisation’ exception in the post-Lisbon era
The ECJ had introduced the ‘minimum harmonisation’ exception in Opinion 2/91, issued in 1993. In a fashion reminiscent of its ruling in Keck & Mithouard, issued in the same year, where the ECJ excluded ‘certain selling arrangements’ from the scope of application of the freedom of movement principles, the Court in Opinion 2/91 excluded from the scope of the ERTA principle those Member State international commitments which provided for a higher level of protection than that provided for in pre-existing common EU rules.
The rationale of the exception was sensible: if the EU legislator itself authorises Member States to ‘lift the bar’ further, and if the proposed Member State international agreement in turn does not prevent the EU from lifting the bar further still, the Member State international agreement would most likely not ‘affect common rules or alter their scope’ in the meaning of the ERTA principle.
The ‘minimum harmonisation’ exception was reaffirmed in Opinion 1/03 on the Lugano Convention, where a full court summarised the exception’s rationale by stating that ‘the Court did not find [in Opinion 2/91] that the Community had exclusive competence where, because both the Community provisions and those of an international convention laid down minimum standards, there was nothing to prevent the full application of Community law by the Member States’ (para. 123).
In the 2014 Neighbouring Rights case, AG Sharpston returned to the ‘minimum harmonisation’ exception. At issue in that case was the nature of the competence of the EU to negotiate and conclude a Council of Europe Convention on the protection of the rights of broadcasting organisations. The Commission considered the agreement covered by an ERTA effect because the proposed international agreement would fall within the scope of the acquis in the area of broadcasting rights (para. 45).
The Council and a number of Member States disagreed with the Commission. They contested the Commission’s ‘wholesale’ approach under which it would suffice for an area to be ‘largely covered’ for an ERTA effect to be triggered. In the Member States and the Council’s view, ‘a conclusion [in favour of an ERTA effect] may be reached only after a precise and specific analysis of the nature and content of the EU rules concerned and of the relationship between those rules and the envisaged agreement which shows that that agreement is capable of affecting those rules or of altering their scope’ (para. 50).
Against the Commission’s wholesale approach, the Member States and the Council thus advanced a ‘retail’ approach aimed at determining whether or not specific provisions and rules are covered by an ERTA effect.
Following up on a suggestion in this sense by Poland and the United Kingdom, Advocate General Sharpston advanced the argument that with regard to the retransmission of broadcasting signals the EU legislator had only provided for minimum standards. She argued in particular that the legislator had provided for a right of retransmission only with regard to retransmission by wired means; retransmission by wireless means was not covered.
It followed, the AG considered, that the Member States had ‘retained’ a competence to provide for a higher level of protection by extending the scope of the right to retransmission to retransmission by wireless means. Consequently, the inclusion of such a right in the proposed international agreement was not liable to affect the common EU rules on the right to retransmission.
The AG’s understanding of the ‘minimum harmonisation’ exception fitted well with her broader approach to applying the ERTA principle, which aligned more closely to the retail approach advocated by the Council and Member States. By engaging in an analysis at the level of the specific rules most likely to be included in the agreement, the AG put a high value on protecting the limits of the EU’s competences.
The ECJ rejected this conception of the ERTA doctrine. Instead of engaging in the type of ‘precise and specific’ analysis advocated for by the Council and the Member States and endorsed by AG Sharpston, the ECJ opted for the wholesale approach supported by the Commission. As many elements of the proposed agreement were already covered by common EU rules, the ECJ concluded that the proposed convention was covered by an ERTA effect (para. 85).
Having opted for the wholesale approach advocated by the Commission, the ECJ did not follow its AG on the issue of minimum harmonisation. The Court distinguished Neighbouring Rights from Opinion 2/91 by repeating that in that opinion the EU did not have exclusive competence because both the provisions of EU law and those of the international convention in question laid down minimum requirements. This was not the case in Neighbouring Rights, as the right to retransmission provided for in common EU rules had a precise material scope (para. 93).
The AG had advised the Court to opt for a retail approach; the Court instead opted for a wholesale approach to its ERTA analysis. Nonetheless, the Court in Neighbouring Rights did not go as far as to overturn Opinion 2/91; the ‘minimum harmonisation’ exception remained good law. In particular, when both the common EU rules and the proposed international agreement set out minimum standards, the presumption remains that the proposed agreement will not affect common EU rules. In such an event, no ERTA effect will be triggered, and the EU does not acquire an exclusive competence to conclude the proposed agreement.
Despite the ECJ rejection of the retail approach to ERTA analysis, and the rejection of the conception of the ‘minimum harmonisation’ exception proposed by AG Sharpston in Neighbouring Rights, in Opinion 3/15 a number of Member States (but, interestingly, not the Council) advanced an argument similar in some respects to the one advanced in Neighbouring Rights.
In Opinion 3/15, as in Neighbouring Rights, the argument was made that the provision of exceptions to rights established in Directive 2001/29 ought to be regarded as minimum standards, as these exceptions left a degree of discretion to the Member States. The argument is similar to that by AG Sharpston in Neighbouring Rights, in the sense that in both instances the EU legislator excluded certain aspects from the scope of the rights provided by the common EU rules.
The Member States in Opinion 3/15 went further than the AG had gone in Neighbouring Rights, however, as they here argued that a permission to lower the level of protection provided for in common EU rules ought to be regarded as a minimum standard.
The Member States here pushed the ‘minimum harmonisation’ exception beyond the limits of its own rationale. As mentioned, the objective of the exception as conceived in Opinion 2/91 was to allow for a ‘lifting of the bar’; the level of protection already provided for by common EU rules would not be affected in the meaning of the ERTA principle if a proposed international agreement provides for a higher level of protection while allowing the EU legislator to ‘lift the bar’ further if it wishes to do so.
This rationale does not apply to the Member States’ argument in Opinion 3/15. Directive 2001/29 does not set forth minimum standards. Rather, the permission to provide for exceptions to the rights established in the Directive authorises the Member States to lower the level of protection. If they make use of this possibility, a risk that a proposed international agreement affects the rights set out in Directive 2001/29 arises, and an ERTA effect is warranted.
Unsurprisingly, the ECJ was not persuaded by the Member States’ reasoning. The Court again reaffirmed its initial understanding of the ‘minimum harmonisation’ exception by emphasising that both the provisions of EU law and those of the international convention must lay down minimum requirements for the exception to apply (para. 120). A permission to establish a lower level of protection does not fall within the scope of the exception (para. 121).
Opinion 3/15 fits within a broader trend in the ECJ’s ERTA case law in the post-Lisbon era. After a period in which the ECJ interpreted the doctrine narrower – Opinion 1/94 on the nature of the EU’s competence to conclude the WTO agreements being the clearest example –in a string of recent cases the ECJ has returned to a broader reading of the doctrine (see in particular Neighbouring Rights, Opinion 1/13 and Green Network).
Having revived the ‘area largely covered’ test, the ECJ adopts a wholesale approach to its ERTA analysis in which it seeks to determine whether the scope of a proposed international agreement overlaps with the scope of the pre-existing acquis. In doing so, the Court does not descend to the retail level; it searches neither for concrete obstacles to the full effectiveness of particular EU rules, nor for actual contradictions between those rules and the terms of a proposed international agreement.
The Court’s rejection of both AG Sharpston’s conception of the ‘minimum harmonisation’ exception articulated in Neighbouring Rights and of the Member States’ second attempt at advancing this conception in Opinion 3/15 fit within this broader tendency to read broadly the scope of the ERTA principle. By opting for this broad reading of the ERTA principle, the ECJ places itself on a collision course with several Member States. For, as two members of the Council legal service recently argued: ‘It appears doubtful that the Council and its Member States will change their position about mixity, a practice that, reinforced with provisional application, is founded on the fundamental principle of conferral and which in their view has proven to be very useful.’
 In this sense, see e.g. Ricardo Gosalbo-Bono and Frederik Naert, ‘The Reluctant (Lisbon) Treaty and its Implementation in the Practice of the Council’ in Piet Eeckhout and Manual Lopez-Escudero (eds), The European Union’s External Action in Times of Crisis (Hart 2016) 23.: ‘… any competence not yet exercised by the Union, even for specific aspects in a field largely regulated by the Union, remains with the Member States and … this is an entirely discretionary and political choice whether in any given case the Union chooses to exercise a shared competence, or not to do so.’
 Article 19(8) of the ILO constitution provides that ‘[i]n no case shall the adoption of any Convention or Recommendation by the Conference, or the ratification of any Convention by any Member, be deemed to affect any law, award, custom or agreement which ensures more favourable conditions to the workers concerned than those provided for in the Convention or Recommendation.’
 In this sense, see also Geert De Baere, ‘EU External Action’ in Catherine Barnard and Steve Peers (eds), European Union Law (Oxford University Press 2014) 721.
 Gosalbo-Bono and Naert (n 1) 26.
The ECJ’s ‘largely covered’ approach in its ERTA line of cases always makes me think of the opposite test it uses in cases that involve ‘functional succession’. So in cases when the EU might have become bound by an international agreement to which all Member States are parties but not the EU, the ECJ argues that this can only happen if the fields covered by the agreement have been taken over by the EU in their ‘entirety’. This makes me wonder whether the ‘largely covered’ test has any sound legal basis and reasoning or it is just a tool to expand EU competences externally? I tend to think that it is the latter. Thus, when the EU can gain implied external competences, it is enough that the areas covered by the international agreement be covered ‘largely’ by internal EU rules. However, if the EU risks being subjected to international obligations via functional succession, then the ‘entirely covered’ test is good to shield the EU from international obligations.
In Item 128 of the ECJ opinion it states, in part, that following the EU Ratification of the Marrakesh Treaty, Member States : ” … would henceforth be required to provide for such an exception or limitation under Article 4 of that treaty.”