A guide to EU procedures for the conclusion of international trade agreements (EP Briefing)


by  Laura Puccio (European Parliament Members’ Research Service)


The European Union (EU) was the world’s biggest exporter and importer of goods and services in 2015, representing 32.51 % of global trade in goods and services. The USA and China, meanwhile, accounted for 12.01 % and 10.68 % respectively. The EU has been negotiating trade agreements since the 1970s, then as the European Communities. Over time it has diversified its trading partners, and is now negotiating trade agreements with partners from every continent. The content of trade agreements has also evolved as EU trade competences have developed. The EU is currently in the process of amending and modernising some of its older trade agreements and is working on some of the most ambitious trade agreements since its inception (such as the Comprehensive Economic and Trade Agreement (CETA) with Canada and the Transatlantic Trade and Investment Partnership (TTIP) with the USA).The Lisbon Treaty modified both the EU’s competences in trade and the procedure for concluding trade agreements, giving a stronger role to the European Parliament. This briefing looks at how trade negotiations are conducted and concluded in the EU, and discusses some of the key issues in the current EU trade policy debate.


In this briefing:

3.EU competences, mixed agreements and the legal basis for Council decisions regarding trade agreements
4. Signature and provisional application
5. Conclusion of trade agreements


In 2015, the EU-28 was the largest global exporter and importer of goods and services, representing 32.51 % of total world trade in goods and services (Source: World Bank data), while the US and China represented 12.01 % and 10.68 % respectively.
The EU has been negotiating trade agreements since the 1970s, back then as the European Communities. Originally focused on European, African, Caribbean and Pacific (ACP) and Mediterranean trade partners, the EU now negotiates with partners on every continent. From the point of view of substance, the content of trade agreements has evolved, from mainly agreements on trade in goods, instituting free trade areas, to agreements including WTO+ commitments1 in a wide range of areas (such as services, intellectual property rights, investment and regulatory cooperation). The EU has started a great number of negotiations in order either to modernise older agreements (such as those with Mediterranean countries and with Mexico and Chile), or to negotiate new bilateral agreements with Asian, Oceanic and North American partners, as well as to advance the multilateral trading system (through the Trade in Services Agreement (TiSA) or the Environmental Goods Agreement (EGA) for instance).
The evolution in the content of trade agreements reflects that of EU competences in trade, but has raised several questions as to whether the more recent agreements fall entirely within EU competence and, consequently, whether ratification at national level is required. Moreover, growing criticism and political debate at national level have raised some procedural issues, such as whether an individual EU Member State can stop EU negotiations, and what happens if one Member State does not ratify a trade agreement.
The procedures for concluding international agreements are mainly set out in Article 218 of the Treaty on the Functioning of the European Union (TFEU). In the case of trade agreements, rules are also to be found in the specific provisions of Article 207 (TFEU) dealing with common commercial policy and any other article mentioned as a legal basis in the Council decisions to sign and to conclude a given trade agreement.

Figure 1: State of play of EU trade relations


Source: European Commission, DG Trade, 2016.

EPRS      EU procedures for conclusion of international trade agreements



 TTIP negotiations: can a single Member State stop the negotiations?

The negotiating directive establishing the Commission negotiating mandate for TTIP was adopted by the Council on 17 June 2013, launching the negotiations. The EU and the USA concluded the 15th round of negotiations in October 2016.

At the end of August 2016, the French minister for foreign trade expressed his government’s wish to request a halt in the TTIP negotiations at the informal Council meeting of 22-23 September. Member States are divided on the issue and 12 Member States clearly expressed their opposition to the French proposal. After the meeting on 23 September, the Slovak Prime Minister declared that the TTIP negotiations would continue but that it was unrealistic to finalise an agreement before the end of US President Barack Obama’s term in office.

The Council can withdraw or suspend the negotiating mandate for a trade negotiation but only on the basis of Article 218 TFEU, which requires a qualified majority (see below for details of the Council’s voting procedure). In general, the Council always tries to take decisions by consensus (i.e. with the agreement of all parties) if the decision concerns shared competences.

Before negotiations begin, the Commission first holds a public consultation and conducts what is known as a scoping exercise. The scoping exercise is a series of informal dialogues with the other country (or countries, if the agreement is inter-regional) on what could be the broad lines of the content of the negotiations between the parties.

If after the scoping exercise the Commission considers it appropriate to open negotiations on a trade agreement with the country/countries, it then makes recommendations to that end to the Council on the basis of Article 207(3) TFEU. The Council must give a green light to the start of the negotiations by adopting a decision on the basis of Articles 207(3) and 218(2) TFEU. The Council also issues a negotiating mandate detailing the area in which the Commission is authorised to negotiate.

Under Article 207(3) TFEU, the Commission is in charge of conducting negotiations, reporting to the Council’s Trade Policy Committee (TPC). The negotiating team is led by a chief negotiator and includes experts covering all the topics of the negotiation. While the Commission’s DG Trade takes the lead, experts may come from other DGs within the Commission. Negotiations are conducted in rounds, but meetings and contacts between lead negotiators and experts continue outside these. In its guide to trade negotiation procedures, the Commission considers the duration of negotiations to be two to three years on average. The negotiations are conducted on the basis of multiple and specific negotiating directives that the Council issues on the basis of Articles 207 and 218 TFEU. These frame the position that the EU must hold during the negotiations.

With TTIP, the Commission began publishing the EU’s text proposals online; these are the EU’s proposals for the drafting of concrete provisions within the various chapters of the agreement. The text proposals that the Commission drafts must be agreed with the Council before they can be tabled for discussions with the other party (parties) to the negotiations.

Often forgotten but fundamental, Article 207(3) makes the Council and Commission jointly responsible for ensuring that the agreement negotiated is compatible with internal EU policies and rules.

The European Parliament’s role in negotiations

While the European Parliament has no formal role in starting and conducting trade negotiations, the TFEU imposes a duty of information: the European Parliament must be informed immediately and fully at all stages of the procedures. Moreover, the fact that the European Parliament has to give its consent at the end of the negotiations, has made it necessary to discuss some EU positions in the negotiations with the European Parliament first, in order for the Commission to verify the existence of political support. The Commission therefore reports regularly to both the European Parliament and the TPC. While it has no legal obligation to do so, Parliament will often signal its political position by issuing a resolution. In the past, the European Parliament has adopted resolutions on the opening of negotiations, either prior to or after the issuing of the negotiating mandate These resolutions give an initial sense of Parliament’s political stance on the negotiations, and set out the main concerns that Parliament wants the Commission to include or exclude from the scope of the negotiations (e.g. the resolution adopted on TTIP in 2013). EP resolutions can be issued during the negotiations in order for Parliament to give the Commission recommendations on the future development of the negotiations (e.g. the resolution adopted on TTIP in 2015). The Commission is not legally bound to follow the EP’s recommendations but given that EP consent is needed to adopt the agreement, it does take them into account when devising the EU positions and discussing them with the Council or the other party.

Ex ante sustainability impact assessments during the negotiations have become the norm for all major multilateral and bilateral EU trade negotiations since 1999, when the EU began incorporating the concept of sustainable development into the definition and planning of its trade policy. Sustainability impact assessments (SIAs) comprise a consultation process and analysis by independent organisations (think-tanks or universities) to assess the potential economic, social, human rights and environmental impacts that a trade agreement could have. SIAs are carried out after the scoping exercise, as the latter defines the scope of the negotiations and will indirectly define the SIA’s coverage.

When negotiations reach the final stage, i.e. parties have agreed in principle on a single text, the European Parliament and Council are informed and legal scrubbing starts to ensure that the text is legally coherent. Some minor changes may still occur at this stage. Once legal scrubbing is complete, the text is initialled, i.e. the chief negotiators from each party place their initials on every page of the agreement to signify that this is the agreed text. Initialling does not amount to the text being legally binding. In order to enter into force an agreement must be signed and ratified. To start such a procedure, the EU needs to define the legal basis for the trade agreement, which determine who has competence in the EU to ratify the treaty.

EU competences, mixed agreements and the legal basis for Council decisions regarding trade agreements

The various types of EU competence and their implications for trade agreements

The EU is based on the principle of conferral; in other words the EU acts within the limits of the competences conferred upon it by the Treaties. There are different types of competence that can influence the way in which procedures for concluding an agreement unfold. These are: exclusive competences,2 shared competences3 and concurrent competences.4 Whenever an international agreement includes shared competences or concurrent competences or Member States’ competences, then the agreement is said to be ‘mixed’. Whenever a trade agreement also contains provisions belonging to shared competences, it is concluded as a mixed agreement. While for agreements falling under exclusive EU competence the EU ratification procedure (explained below) is sufficient to ensure the entry into force of the agreement, mixed agreements must be ratified by EU Member States in accordance with their domestic ratification procedures. Domestic procedures vary from Member State to Member State. In federal Member States, ratification procedures also involve approval by the chamber of the national parliament representing the regions (such as the Bundesrat in Germany) or the approval of the regional and community parliaments (as in the case of Belgium), whenever competences of sub-federal entities are concerned by the agreement. While mixed agreements concluded by the EU and only some EU Member States (called partial or incomplete mixity) do exist, trade agreements concluded as mixed agreements (such as association agreements) require the participation of all Member States.

The evolution of trade competences and recent EU trade agreements

The common commercial policy (CCP), which defines EU trade policy, has always been an exclusive competence of the EU, however the content of the CCP has evolved over time. While services and intellectual property rights were originally considered shared competences, the Lisbon Treaty includes all services and commercial aspects of intellectual property rights within the CCP’s scope (Article 207(1) TFEU).

Article 207(1) TFEU also introduces foreign direct investment to the list of CCP competences. This evolution in the scope of CCP has led the Commission to consider whether it can conclude that some trade agreements focusing on purely commercial matters (including investment provisions) fall under exclusive EU competence. The argument of the Commission is contested by most Member States who consider that those agreements must be concluded as mixed. The main controversy between the Council and the Commission concerns whether investment now falls under exclusive EU competence. On the one hand, Member States consider that CCP covers only foreign direct investment (FDI) and not portfolio investments. However the Commission derives an implicit exclusive competence on portfolio investments from third countries from a rule in the internal market prohibiting the introduction of barriers at Member State level to capital flows and payments from third countries. For that reason the European Commission asked for an opinion of the Court of Justice of the EU (CJEU) to decide on the nature of the EU competence to conclude the EU-Singapore agreement. A hearing has been held and the opinion is expected for either late 2016 or early 2017. The CJEU opinion will formally only affect the EU-Singapore agreement, but could influence the choice of the competence (exclusive or mixed) for other agreements (such as the EU-Vietnam agreement).

The concept of mixity in EU agreements and the choice of legal basis

The decision with respect to the mixed character of an agreement depends on the legal basis given to that agreement, which also defines the main competences involved. When the Commission proposes a Council decision to sign, conclude or provisionally apply a Treaty, it must also propose a legal basis, which will define the nature of the agreement (exclusive or mixed). The legal basis is usually discussed with Member States and it is normally in the Commission’s interests to agree with the Member States on this point.

The Council can always modify the Commission proposal in accordance with Article 293(1) TFEU, which requires unanimity. A proposal can remain blocked if the Council decides not to act. This can happen in situations where the Council is divided on the issue of legal basis (i.e. no unanimity is reached to modify the Commission proposal), and it cannot reach a qualified majority (or unanimity, depending on the procedure required by the legal basis in the original proposal) to pass the act as is. In that case, the Commission can modify its proposal at any time in order to unblock the situation (Article 293(2) TFEU).

 Understanding the political and legal implications of the legal basis: the case of CETA

There were divergent opinions on whether the Comprehensive Economic and Trade Agreement negotiated between the EU and Canada should be concluded as a mixed agreement. The Commission considered that CETA fell under exclusive EU competence as in the case of the EU-Singapore agreement. The Commission was reported therefore to favour the idea of submitting a Council decision on CETA as an EU-only agreement with CCP as the sole legal basis. However, after discussing the matter informally with the Member States, the Commission ultimately decided to submit it as a mixed agreement. (See also: W. Schöllmann, Is CETA a mixed agreement?, EPRS ‘at a glance’ note, July 2016)

Signature and provisional application

Once the Council has adopted the decision to sign the treaty, a date for its signature can be chosen. In practice, for mixed agreements, the EU and the Member States sign the treaty simultaneously. Signature signals the intention to conclude, it does not conclude the agreement as such.

The possibility for an international treaty to apply provisionally under EU law is set out under Article 218(5) TFEU, which provides for the Council decision on provisional application to be taken simultaneously with the Council decision to sign the treaty.

In theory, under Article 218(5) TFEU, the decision on provisional application can take place even before the treaty is concluded at EU level, i.e. before the EP gives its consent and the Council adopts the decision to conclude the treaty in accordance with Article 218(6). However, in practice (since the South Korea FTA),5 provisional application is enforced only after hearing the European Parliament’s position on the agreement or even only after the European Parliament has given its consent to conclusion. Consequently, the Commission normally submits the draft decisions to the Council simultaneously: the draft decision to sign, that to provisionally apply the treaty and one for the conclusion of the treaty.

CETA: provisional application

Provisional application of CETA will be effective from the first day of the month after the parties have notified each other that they have completed the domestic procedures necessary for provisional application. The procedure in the EU is contained in Article 218(5) TFEU. In line with EU practice, the decision on the provisional application of CETA, if adopted by the Council, will be applied only after the EP has taken a position on the agreement.

The main discussions in Council concerning the provisional application of CETA focus on its scope. The draft Council decision on provisional application of CETA, submitted by the Commission, does not refer to any specific provisions, thus provisional application would refer to the whole treaty.

In order to modify the scope of this decision, the Council needs to act by unanimity (pursuant to Article 293(1) TFEU) or has to agree with the Commission that it submit a new proposal for a Council decision. Furthermore, partial provisional application requires the agreement of Canada. Indeed under Article 30(7)(3)(b) CETA, Canada can object to partial provisional application of the treaty and either decide not to allow provisional application of the treaty or to propose unilaterally to exclude equivalent provisions from provisional application.

While a decision on the matter is yet to be reached, a proposal was circulated in the Council on 5 October 2016. This would exclude, inter alia, part of the investment and financial services chapters. It also requires provisional application of sustainable development chapters, as only provisions falling under EU competences, as these do, can be provisionally applied.

The provisional application of mixed agreements negotiated by the EU takes place, however, before the completion of ratification procedures at the Member State level. This makes sense as the entire rationale of provisional application is to allow for application while waiting for the completion of the ratification procedure. However provisional application under Article 218(5) TFEU can only be granted for provisions relating to EU competence and cannot include Member State competences unless all the Member States have agreed to it separately. Decisions on the provisional application of a mixed agreement in its entirety usually include a statement clarifying that Member States have given their agreement with respect to their competences.

Conclusion of trade agreements

For trade agreements, the special procedure under Article 218(6) TFEU is applied. This procedure requires the European Parliament’s consent.6 Once Parliament has given its consent, the Council can then adopt a decision to conclude the agreement following the procedure and voting rules set out in Article 218(6) and Article 218(8) TFEU respectively. As mentioned above, mixed agreements also require the agreement to be ratified at national level by Member States. In the case of mixed agreements, the treaty enters into force only when the non-EU trade partner, the EU and all Member States have exchanged ratification instruments.

Voting procedure in the Council

Voting procedure in Council under Article 218(8) TFEU

EU-Ukraine Association Agreement: provisional application and ratification procedure

The EU-Ukraine Association Agreement, negotiated between 2007 and 2012, has been partly provisionally applied since 2014, while the provisional application of the commercial part of the Association Agreement began on 1 January 2016. The provisional application currently applies only to EU competences. In order to enter fully into force, the EU-Ukraine Association Agreement, being a mixed agreement, requires the ratification procedure to be complete at EU and also at Member State level. At EU level, the EP has given its consent for the Council to conclude the agreement in two different resolutions (one covering treatment of third-country nationals and one covering the other provisions). The Council is waiting for the Member States to finalise the ratification process in order to formally adopt its decision on the conclusion of the agreement. All Member States, with the exception of the Netherlands, have ratified the Treaty.

The Netherlands held an advisory referendum on the EU-Ukraine Association Agreement on 6 April 2016, which yielded a negative result (over 61 % of the voters rejected the ratification of the Association Agreement (AA) between the EU and Ukraine, though turnout was low, at only 32 %). The referendum was an advisory referendum and as such has not put an end to the ratification procedure in the Netherlands. However, should the Netherlands notify its intention not to ratify the agreement, this would then signify that the EU-Ukraine Association Agreement could not enter into force in its present form. Under Article 25 of the Vienna Convention on the Law of Treaties, provisional application can only remain pending the entry into force of a treaty. If ratification fails and entry into force of the treaty becomes impossible, provisional application would also have to be lifted. Suspension of the provisional application would have to be carried out in accordance with Article 218(9) TFEU and the notification procedure under Article 486(7) of the EU-Ukraine Association Agreement. As has been done in the past, the Commission could argue on the basis of the duty of cooperation that the provisional application, which concerns EU competences only, should be maintained in order to allow renegotiation and to find a mutually acceptable solution.

The voting rule for Council decisions is contained in Article 218(8) TFEU. It refers to the Council voting procedure throughout the entire process of negotiating and concluding international agreements under Article 218 TFEU. These voting rules therefore apply equally to Council decisions taken pursuant to Article 218(5) TFEU in order to sign and provisionally apply a treaty, and to Council decisions taken pursuant to Article 218(6) TFEU to conclude an agreement. Although in practice the Council tries to take all decisions regarding shared competences on the basis of the ‘common accord’ of all Member States (i.e. by consensus), the voting procedure under the TFEU does not depend on the nature of the agreement but on the competences and legal basis upon which the agreement is adopted.

Article 218(8) TFEU states that qualified majority must be used throughout international agreement negotiation and conclusion procedures. There are some exceptions to this qualified majority rule; these include the following situations or agreements.

  • Fields for which unanimity is required for the adoption of a Union act: Any measure which requires unanimity for the purpose of EU internal legislation, will also require unanimity for any decisions taken under Article 218 TFEU. In practice, this rule provides for parallel decision-making on external policy and EU internal legislative procedures. This rule on parallelism ensures that the EU internal legislation procedure requiring   unanimity   for   certain   measures   is   not circumvented by the conclusion of similar measures within an international agreement under Article 218 TFEU.
  • Association agreements
  • Agreements referred to in Article 212 TFEU (i.e. economic, financial and technical cooperation arrangements) with states that are candidates for accession
  • Accession of the Union to the European Convention for the Protection of Human Rights and Fundamental Freedoms.7

The parallel internal and external decision-making provided for under Article 218(8) TFEU requires analysis of all fields mentioned in the legal basis chosen for the decisions to sign and conclude the agreement in order to understand which voting rule applies.

The trade legal basis under Article 207 TFEU and its voting procedure

The relevant rules for common commercial policy agreements and measures always include Article 207(4) TFEU (in addition to other legal bases, such as the transport legal basis in CETA). Article 207(4) TFEU normally requires qualified majority but it specifies that the Council must take its decision by unanimity for the following measures.

  • In the field of trade in services and the commercial aspects of intellectual property, as well as foreign direct investment, where such agreements include provisions for which unanimity is required for the adoption of internal rules: this is another formulation of the parallelism between internal and external decision-making procedures in order to avoid circumvention of unanimity in the internal rules via external relations agreements. However, the impact of this provision may be rather limited. There are few internal legal bases requiring unanimity in these fields. Article 118 TFEU requires unanimity only for regulations establishing language arrangements for European intellectual property rights, whereas qualified majority remains the rule under the same Article 118 TFEU for measures establishing the creation of measures for the ‘uniform protection of intellectual property rights throughout the Union and for the setting up of centralised Union-wide authorisation, coordination and supervision arrangements’.8 Other internal market provisions requiring unanimity include: Article 113 TFEU on tax harmonisation, Article 115 TFEU on approximation of laws and Article 64(2) TFEU on the introduction of restrictions to capital movements.9 If the agreement in question incorporates provisions covering one of these types of measure then unanimity will be required.
  • Unanimity is also required for EU actions in the field of trade in cultural and audiovisual services where there is a risk of prejudicing the Union’s cultural diversity. Again this provision has a rather limited impact on the conclusion of trade agreements. Unanimity only applies if there are measures related to trade in cultural and audiovisual services and where there is risk of prejudicing the Union’s cultural The EU has sometimes omitted the field of cultural and audiovisual services from negotiations in its entirety (exception culturelle) as is the case for TTIP (where the audiovisual sector is excluded from the negotiating mandate). Other agreements include a protocol on cultural cooperation aiming for instance to implement, in the context of those agreements, the Unesco Convention on the Protection and Promotion of the Diversity of Cultural Expressions. Still, even in those agreements where provision is made for cultural cooperation, cultural and audiovisual services are carved out from the services commitments or are included in specific reservations so that there can be no risk of prejudicing cultural diversity.
  • Unanimity is also required for EU actions in the field of trade in social, education and health services where there is a serious risk of disturbing the organisation of these services at national level. These social, education and health services are subject to several reservations in the Treaties so as to prevent trade agreement commitments from having any negative impact on them.

Main references

Devuyst, Youri, ‘European Union Law and Practice in the Negotiation and Conclusion of International Trade Agreements’, Journal of International Business and Law, 12: 259 (2013)
Eeckhout, Piet, EU External Relations Law, Oxford University Press, 2011
Hillion, Christophe, Koutrakos, Panos, Mixed Agreements Revisited, Hart Publishing, 2010
Koutrakos, Panos, EU International Relations Law, Hart Publishing, 2015Endnotes
1 WTO+ commitments are commitments in trade agreements that go beyond those made at the WTO. In some literature a distinction is made between commitments in trade agreements that extend liberalisation commitments already existing at WTO level (WTO+) and commitments that deal with issues not covered by WTO law (WTO extra).
See: H. Horn, P. C. Mavroidis and A. Sapir, ‘EU and US Preferential Trade Agreements’, in Preferential Trade Agreements: A Law and Economics Analysis, Kyle W. Bagwell, Petros C. Mavroidis (eds), Cambridge University Press, 2011.
2 Exclusive competences of the EU signify a complete transfer of competences from the Member States to the EU; the existence of an EU exclusive competence means that Member States cannot act on their own unless an EU regulation allows for Member State actions.
3 Shared competences are competences that fall within the remit of both the EU and the Member States. EU action in these competences pre-empts any action on the part of the Member States, in other words, Member States cannot act unilaterally if action is being undertaken at EU level.
4 Concurrent competences are EU competences to support, coordinate or supplement Member States’ action; these EU competences co-exist with Member States’ competences.
5 Youri Devuyst, ‘The European Parliament and International Trade Agreements: Practice after the Lisbon Treaty’, in The European Union in the World: Essays in honour of Marc Maresceau, I. Govaere, E. Lannon, P. Elseweghe and S. Adam (eds), Martinus Nijhoff Publisher, 2014.
6 Parliament gives its consent in accordance with Rules 108 and 99 of its Rules of Procedure.
7 The Court of Justice of the European Union issued an opinion in 2014 on the draft agreement for EU accession to the ECHR, considering it not in line with EU law.
8 On the relevance or not of this provision to the decision-making procedure under Article 218(8) TFEU, see C Pitschas, ‘Economic Partnership Agreement and EU trade policy: objectives, competences, and implementation’, in J. Drexl, EU bilateral trade agreements and intellectual property for the better or worse, H. Grosse, R. Khan, S. Nadde-Phlix (eds)
Springer, 2013, p. 222.
9 See also: Panos Koutrakos, EU International Relations Law, Hart Publishing, 2015, p. 135.

Disclaimer and Copyright The content of this document is the sole responsibility of the author and any opinions expressed therein do not necessarily represent the official position of the European Parliament. It is addressed to the Members and staff of the EP for their parliamentary work. Reproduction and translation for non-commercial purposes are authorised, provided the source is acknowledged and the European Parliament is given prior notice and sent a copy.
© European Union, 2016. Photo credits: © gustavofrazao / Fotolia.
http://www.eprs.ep.parl.union.eu (intranet) http://www.europarl.europa.eu/thinktank (internet) http://epthinktank.eu (blog)

The Bratislava Declaration on migration: European irresponsibility instead of solidarity

ORIGINAL PUBLISHED ON OMNIA (Odysseus Network) SITE (27 Sep 2016)

By Phillippe De Bruycker (ULB/EUI) Evangelia (Lilian) Tsourdi (Max Weber Fellow, EUI)

The Bratislava Declaration refers on two occasions to “the principles of responsibility and solidarity”. The basic idea is to “broaden EU consensus” by devising a “long term migration policy” on the basis of the two principles.

At first look, this seems logical and even advisable. Since 2015, the EU has been unable to respond effectively to the ‘refugee crisis’. It is only the fragile ‘deal’ with Turkey that brought the illusion of a solution by externalising asylum provision to a third country. The EU remains profoundly divided about possible internal solutions. A European East-West divide has appeared, in addition to the well-known North-South division about the principles evoked in the Bratislava Declaration. Member States in the South have been complaining for years about the lack of solidarity measures, while many Member States in the Northwest have castigated them about their inability to implement their responsibilities. More recently, Member States in the Central/Eastern part of the EU (more precisely the Visegrad group consisting of Hungary, the Czech Republic, Slovakia and Poland) are refusing, ostensibly in the name of responsibility, to engage in the type of solidarity requested by no longer only the Member States in the South, but also those in the Northwest.

The objective to heal the wounds and reunify EU Member States around the same principles of solidarity and responsibility appears reasonable and even attractive in this setting. If all Member States (including those in the South) are fully responsible, the others (in particular those in the East) will demonstrate greater solidarity, so that the problem may be solved in a balanced way. This presentation based on an opposition between responsibility and solidarity is, however, simplistic and even incorrect from a legal point of view. If there is indeed in EU law a precise legal provision that can be considered to embody responsibility, applicable in the same manner throughout EU law, the same does not hold true for solidarity (1). Moreover, effective solidarity and fair sharing are a prerequisite to responsibility in EU migration and asylum policies, rather than the other way round  (2).

1. More responsibility than solidarity in EU law in general

When searching in the EU treaties for the word “responsibility”, Article 165(1) TFEU provides an excellent example of the kind of answer that appears: following this provision, “The Union shall contribute to the development of quality education by encouraging cooperation between Member States and, if necessary, by supporting and supplementing their action, while fully respecting theresponsibility of the Member States for the content of teaching and the organisation of education systems and their cultural and linguistic diversity”.Responsibility refers in this sense simply to competence.

Responsibility understood as competence can be envisaged as a power as well as a duty. It is not so surprising that this notion has been linked in the case law of the Court of Justice with the principle of loyalty, now referred to as the principle of sincere co-operation under Article 4(3) TEU. The principle embodies, respectively, a positive obligation (taking measures to ensure fulfilment of obligations), and a negative obligation (abstaining from measures that could jeopardize this fulfilment). It is this first part that is often evoked by Member State governments; with ‘responsibility’ they refer to Member States’ duty to fulfil their obligations and honour their commitments under EU law.

Loyalty has been made explicit under Article 4(3) of the TEU. The principles of loyalty and solidarity are sometimes used interchangeably in legal scholarship, with loyalty considered a facet of solidarity. Under this understanding, the responsibility of Member States to implement their obligations under EU law is a sign of solidarity to each other. This is, however, a narrow understanding of solidarity, which is a notion different from responsibility.

When searching in EU treaties for the word ‘solidarity’, one finds, in particular since the Lisbon Treaty, more results than a similar search for ‘responsibility’. In some instances, solidarity fulfils an aspirational role, providing political orientation, rather than forming the basis of legally binding duties.  For example, following article 3(5) TEU, “In its relations with the wider world, the Union shall (…) contribute to peace, security, the sustainable development of the Earth, solidarity and mutual respect among peoples…”

However, in other areas solidarity forms the basis of concrete actions and legally binding duties as in article 222(1) TFEU, following which “The Union and its Member States shall act jointly in a spirit of solidarity if a Member State is the object of a terrorist attack or the victim of a natural or man-made disaster. The Union shall mobilise all the instruments at its disposal, including the military resources made available by the Member States, to:

(a)       – prevent the terrorist threat in the territory of the Member States;

– protect democratic institutions and the civilian population from any terrorist attack;

– assist a Member State in its territory, at the request of its political authorities, in the event of a terrorist attack;

(b)       assist a Member State in its territory, at the request of its political authorities, in the event of a natural or man-made disaster”.

These latter provision shows that solidarity is not linked with the fulfilment of responsibilities but rather with providing assistance to other Member States in order to allow them to implement their obligations.

Interestingly, solidarity understood in this sense does not have the same status as responsibility understood as loyalty. There is indeed no legal provision of solidarity applicable throughout different policies that would create a general duty to support, but rather different and more or less strong expressions of solidarity. As a consequence, one has to examine each particular policy and the provisions in the EU treaties pertaining to it in order to ascertain whether there are concrete solidarity duties and what the extent of these may be. This leads us to the meaning of solidarity in policies on border checks, asylum and immigration as governed by Articles 77 to 80 TFEU.

2. More solidarity than responsibility in EU migration and asylum policies

When searching for the word ‘responsibility’ or ‘responsible’ in those provisions, there are four hits. Firstly, Article 72 states that the EU competences regarding border checks, asylum and immigration do not affect the “responsibilitiesincumbent upon Member States with regard to the maintenance of law and order and the safeguard of internal security” and, secondly, in Article 73, following which “it shall be open to Member States to organise between themselves and under their responsibility forms of cooperation and coordination as they deem appropriate between the competent departments of their administrationsresponsible for safeguarding national security”. Responsibility in those provisions refers to the notion of competence, i.e. that the Member States remain competent for the maintenance of law and order and internal security, and even exclusively competent for national security.

Another ‘hit’ is found in Article 78(2), requesting the European Union to adopt measures for a common European asylum system comprising, under point (e), “criteria and mechanisms for determining which Member State is responsible for considering an application for asylum or subsidiary protection”. This is the legal basis of the famous “Dublin System”, based on Regulation 604/2013, determining the responsible Member State for examining an application lodged in the EU. As the flaws of this system have already been analysed in numerous publications,including in this blog, it is not necessary to explain them once more.

Let us just remind ourselves that the origin of this regulation goes back to aConvention signed in Dublin on 15 June 1990 (this explains why specialists of EU asylum continue to speak about ‘Dublin’ in relation to this system). The aim of this system is to indicate which Member State is competent when an asylum application is introduced in the EU on the basis of a certain number of criteria. In practice, the responsible Member State will more often than not be the one of the legal or illegal first entry of the third-country national to the EU.

Responsibility in this regulation refers to the idea of competence regarding the examination of asylum applications, so that all Member States have to deal with the asylum applications for which they are responsible. The problem is that the Dublin system was not devised on the basis of solidarity. On the contrary, apart from exceptions based on the right to family unity, or the rights of the child, it is premised on the idea that each Member State should deal with the applications of asylum seekers whose presence is attributable to actions of that Member State. This could be either because it let them enter the EU voluntarily by issuing a visa or residence permit, or involuntarily by not controlling its external borders effectively. It is not a coincidence that the Dublin system was conceived by the North-Western Member States who drafted the Schengen Convention (France, Germany, Belgium, the Netherlands and Luxembourg) which is at the origin of the Dublin Convention. Solidarity was not an issue at that time in such a small and coherent space. Moreover, Dublin was devised in a purely intergovernmental framework, a decade before the beginning of the implementation of the supranational method with regard to asylum policy, as introduced by the Treaty of Amsterdam, without any actor such as the European Commission looking out for the general interest rather than the national interest of each State. It is an excellent example of the kind of measure that Northern governments managed to impose on other Member States of the European Union, who can try to amend it subsequently, although only with the support of those governments, which explains why this has not been possible regarding the core of the system with the regulations Dublin II in 2003 and Dublin III in 2013.

This is crucial as this policy is, like the area of external borders, characterised by asymmetric burdens between the Member States due to the fate of geography. Following this logic, Greece should have examined all the asylum applications that could have been introduced by the hundreds of thousands of third-country nationals who entered the EU through its borders during the year 2015. It should also have intercepted the persons trying to enter the EU through the Greek borders without the requested documents (a passport with very often at least a short-term visa), as well as taken their fingerprints in order to store them inEurodac, a database helping to determine in practice the responsible Member State. In this particular case, it would mean that Greek authorities should have assumed responsibility of one million third-country nationals just because they entered the EU through the Greek territory.

Does it mean that the Southern Member States are legally wrong when they ask for solidarity from the other Member States, and that they should instead, or at least firstly, fulfil their responsibilities deriving from EU law? The answer is actually much more complicated due to Article 80 TFEU, which reads as follows:The policies of the Union set out in this Chapter and their implementation shall be governed by the principle of solidarity and fair sharing of responsibility, including its financial implications, between the Member States. Whenever necessary, the Union acts adopted pursuant to this Chapter shall contain appropriate measures to give effect to this principle”.

This provision is one of those detailing the idea of solidarity in the policies for border checks, asylum and immigration. A quick reading may give the impression that this provision is precisely about two principles that have to be balanced, much like in the Bratislava Declaration. Under this reading, Member States should first fulfil their responsibilities by applying the Dublin Regulation and assuming responsibility for the asylum seekers arriving on their territory before they can expect solidarity. In the event of a failure to take up their responsibilities, they should not expect solidarity, or rather they should be found ‘undeserving’ of it.

However, this provision is about one and not two principles and, more importantly, about the principle of “solidarity and fair sharing of responsibility”.It is interesting to note that the words “fair sharing of” have simply been omitted from the Bratislava Declaration, while they completely change the meaning and content of what is at stake. Instead of an opposition between responsibility and solidarity that should be balanced against each other, the idea of fair sharing of responsibility actually reinforces that of solidarity. The policies of the Union on border checks, asylum and immigration are governed by the principle of solidarity, and responsibilities between the Member States in these areas must be shared in a fair way. If one will agree that fairness leaves some margin of discretion to the European Union, this notion refers to the ideas of equity and justice and thus provides an indication about how the EU policy on borders, immigration and asylum must be conceived and implemented.

It therefore appears that the legal obligation of the EU is not to balance the two principles of solidarity and responsibility, but rather to realise solidarity through a fair sharing of responsibilities. This means also that the concerned Member States should not be expected to implement Dublin as pre-condition for solidarity, but should instead benefit from a system aiming at a fair sharing of responsibility between all EU Member States. Some will say that Dublin is as such not contrary to EU law and that the system could be accompanied by “appropriate measures to give effect to the principle of solidarity and fair sharing of responsibility”, following the wording of Article 80 TFEU. The problem is that Dublin is the source of the asymmetric burdens between Member States, so that it seems difficult to amend or revise it without reversing the basic principle on which it is based.

Conclusion: responsibility or irresponsibility?

Nothing about this constitutional requirement is mentioned in the Bratislava Declaration. On the contrary, the issue of the relocation of asylum seekers, as a concrete solidarity measure at the core of the debate since 2015, has simply disappeared from the agenda, despite the call of the first summit of the Mediterranean countries of the EU organized in Athens on 9 September. This is the case despite the fact that the relocation measures were based on mandatory EU rules, which most Member States do not apply, while some of them openly challenge them, for instance Hungary through the organisation of a referendumcalling the population to vote against them.

What remains is a kind of “flexible solidarity”, following the words of the joint statement of the Heads of Governments of the V4 Countries (the Visegrad group) defined as a concept that “should enable Member State to decide on specific forms of contribution talking into account their experience and potential”, knowing that “any distribution mechanism should be voluntary”. Some observers have already tried to imagine what this could entail. This will become clearer when the Council of Ministers takes a position on the Commission proposal reforming the Dublin system (Dublin IV), which contains a relocation mechanism that appears ambitious but that would in fact be dysfunctional, as underlined by Francesco Maiani in his report for the European Parliament. The European legislator should keep in mind that, despite the discretion left by this provision, Article 80 TFEU requires a strong solidarity mechanism aiming at “fair sharing of responsibility” between the Member States.

The retreat of the EU regarding the issue of solidarity had actually been announced by the President of the Commission himself in his State of the Union speech, where he stated that “Solidarity must be given voluntarily. It must come from the heart. It cannot be forced”. This clearly contradicts the mandatory character of the relocation decision, which was imposed on 22 September 2015 by a qualified majority in the Council against the opposition of Hungary, Slovakia, Romania and the Czech Republic.

The Bratislava Declaration announces a double evolution. First, a so-called principle of responsibility is prioritised over the principle of solidarity and fair sharing, the latter reduced to a “commitment by a number of Member States to offer immediate assistance to strengthen the protection of Bulgaria’s border with Turkey and continue support to other frontline States”. Secondly, “the objective to ensure full control of external borders” is prioritised over the asylum policy, which is not even mentioned in the text.

The so-called “responsibility to ensure full border controls” is nothing else than a rhetoric contrary to the Treaties, ignoring that the Schengen Borders Code is without prejudice to the rights of asylum seekers (see in particular Articles 3 and 4 of Regulation 2016/399 codifying the Schengen Borders Code). Trying to convince public opinion that asylum seekers can simply be rejected at the border without any further examination of their claim is not only illegal but also populistic. This has proven to be impossible, even in the case of a safe third-country, for example Turkey on the basis of the EU/Turkey agreement of 18 March 2016 (see in this blog Henri Labayle’sThe EU-Turkey Agreement on migration and asylum: False pretences or a fool’s bargain?).

The President of the European Council, Donald Tusk wrote in his letter of invitation to the Bratislava Summit that “Europeans all too often heard politically correct statements that Europe cannot become a fortress and that it must remain open”. This is indeed not the case of the Bratislava Declaration where the Heads of State and government want to improve the communication with citizens through the “use of clear and honest language (…) with strong courage to challenge simplistic solutions of extreme or populist political forces”. The problem is that they do exactly this by pretending to build a Fortress Europe, that is de jure impossible. They probably want to prove that this is possible de facto. This is nothing less than European irresponsibility instead of solidarity.

‘Inside’ the European Parliament’s Closed Reading Rooms: Transparency in the EU


by Dr Vigjilenca Abazi (Assistant Professor of European Law Maastricht University)

What do documents about negotiations of the Transatlantic Trade and Investment Partnership (TTIP), oversight of the EU’s Food Safety Authority or Tax-Justice have in common? In order to access these documents, (selected) Members of the European Parliament are requested to attend closed reading rooms. This blog post discusses how an exception to open parliamentary oversight is increasingly becoming a regular institutional practice and questions its spillover effect on requests for public access to documents.


As the wording suggests, ‘closed reading rooms’ are meetings that take place behind closed doors with the purpose of reading certain sensitive documents, particularly EU official secrets. Documents are distributed at the beginning of the meeting and collected again at the end; documents may not be copied by any means, such as photocopying or photographing; no notes may be taken; and the minutes of the meeting cannot make any mention of the discussion of the item containing official secrets (Art. 6, Interinstitutional Agreement between the European Parliament and the Council of 12 March 2014).

Closed reading rooms are an exception to generally open meetings and discussions of the European Parliament. This practice emerged with the introduction of rules on EU official secrets and specifically the Interinstitutional Agreement of 2002 between the European Parliament and the Council concerning European Parliament’s access to sensitive information in the field of security and defence policy (see Art. 3 and Annex, second paragraph). The rationale of what this Agreement called ‘secured room’ was to make sensitive documents available for purposes of parliamentary oversight without ‘risks’ of public disclosure or possible leaks, i.e. unauthorised disclosure of documents.

Initially, this practice was mostly confined to the area of security and defence for documents classified as official secrets. Yet, with the expansion of rules on EU official secrets to areas well beyond security and defence to ‘activities in all areas that require handling classified information’ via a Council Decision on official secrets in 2013, the use of closed reading rooms by MEPs to access sensitive documents became an increasing practice.

Closed Oversight

At first glance, closed reading rooms, or more generally ‘closed oversight’ (as I have elaborated in-depth in this recent article), might seem an inevitable institutional practice when dealing with official secrets and certainly this is not an issue confined to the EU, but a much wider world practice of oversight (e.g. see here for a recent report). Yet, the following salient questions arise:

Is it possible to keep account of closed oversight?

Accountability does not stop with executive institutions. It is equally important that oversight actors, such as the European Parliament, have appropriate institutionalised processes of keeping track of documents that have been reviewed, that meeting minutes reflect at least in some broad sense what has been discussed when official secrets are involved, or any other means that leave a traceable mark of institutional oversight having taken place. As the current procedure of getting access to official secrets stands (see above section on ‘background), it seems that keeping (some sort of public) track of the oversight process is deeply challenging.

To what extent intra/inter institutional rules alter primary law oversight architecture?

Another disconcerting aspect to closed oversight is the way it has been developed, i.e. mostly through rules of procedure and inter-institutional agreements. Indeed, EU institutions in line with primary law have clear prerogatives to make arrangements for their cooperation and to set out their rules of procedure (see respectively Art. 295 TFEU, Art. 240(3) TFEU). However, it remains to be more critically discussed whether this route of designing how oversight will take place in practice follows the constitutional principle of openness in the EU in full spirit and to what extent it alters the process of oversight in EU.

Does recent case law offer insights on closed oversight? 

In a series of recent cases, the CJEU has clarified the relevance, scope and procedural aspects of institutional access to information by the European Parliament in the context of international negotiations (see previous EU Law Analysis blogs here and here). However, case law does not address the manner in which these documents should be read and importantly, primary law only refers that accessibility to information is ‘immediately and fully’ (see Art. 218(10) TFEU) with no further details as to how access ought to be organised.

What about public deliberation?

A crucial role for the European Parliament as the direct representative of citizens (Art. 10 TEU) is to provide a link between what takes place in Brussels and what citizens know. But actively creating space for public deliberation and prompting public debate on issues that are overseen behind closed doors remain yet to be delivered by the European Parliament.

Spillover Effect Even to Public Access to Information?

Recently four MEPs filed a public access request to the European Food Safety Authority to gain access to unpublished studies determining the carcinogenicity of glyphosate on basis of which EFSA made its assessments. EFSA was not immediately open to provide public access to these studies. Remarkably, in its response, EFSA offers a ‘physical reading room’ for the MEPs to read these studies and reasons that the owners of these studies seem open to sharing the studies in this manner.

In other words, the EFSA is offering the MEPs a closed room to read the studies as a response to a public access request that should result in making the documents public, not only for these four MEPs but also for the general public. It should be stressed that the EU public access to documents regime does not foresee ‘physical reading rooms’ and indeed that would be contrary even to its rationale of granting the widest possible public access to documents. It seems that in the eyes of EFSA, a closed reading room offers a ‘solution’ to the potential unwillingness of the authors of these studies to disclose the documents. Yet, this possibility is also completely outside the legal contours of public access to information. Legally, authors of these studies do not have a veto on whether the studies would be public and certainly do not have prerogatives to decide how public access to documents should be organised in practice.

The EFSA response is ongoing and the four MEPs have still not received access to all requested documents. Yet, beyond this case, is the practice of closed reading rooms expanding not only toinstitutional access but also to public access to documents? This is a issue that we should continue to examine more closely.

Leaving the EU: UK Parliament’s Role in the Process

Published Thursday, June 30, 2016

Following the result of the referendum held on 23 June 2016, this House of Lords Library briefing examines what Parliament’s role would be in the process of withdrawing from the European Union in several key areas: invoking Article 50; overseeing the negotiation process; ratifying agreements; repealing and reviewing domestic legislation.

Jump to full report >>

Following a vote in the referendum on 23 June 2016 in favour of the UK leaving the European Union, the Prime Minister said that this decision “must be accepted”, adding that “Parliament will clearly have a role in making sure that we find the best way forward”. Drawing on parliamentary material and recent legal and constitutional comment, this Library briefing examines what Parliament’s role would be in the process of withdrawing from the European Union in several key areas:

Invoking Article 50—The Prime Minister has said it would be for his successor and his or her Cabinet to decide whether the House of Commons should have a vote on the decision to trigger Article 50, the formal process set out in the Treaty on European Union for member states to follow should they decide to leave the EU. Some legal commentators agree that prerogative powers would enable a Prime Minister to take this decision; some have suggested that Parliament could have a role, and others have gone further, arguing that prior parliamentary approval would be required before Article 50 could be invoked.

Overseeing the Negotiation Process—Formal negotiations between the UK and the European Union would not begin until the UK made a notification under Article 50 of its decision to withdraw from the EU. Parliament’s involvement in overseeing or scrutinising such negotiations has not yet been set out in great detail. The chair of the House of Lords European Union Committee has called for Parliament to be “fully involved” in the process.

Ratifying Agreements—Parliament would have a statutory role in ratifying an eventual withdrawal agreement and any other international agreements arising from the negotiations if they were subject to the usual procedure for ratifying treaties. The House of Commons potentially has the power to block the ratification of a treaty indefinitely; the House of Lords does not. Under the terms of Article 50, the UK’s membership would cease two years after it gave formal notification of its intention to leave, if no withdrawal agreement had come into force by that point, although the two-year period could be extended on the unanimous agreement of all EU member states.

Repealing and Reviewing Domestic Legislation—As part of the process of leaving the EU, decisions would need to be made about how to deal with existing domestic legislation passed to enable EU law to have effect in the UK, a process which the House of Lords European Union Committee has described as “domestic disentanglement from EU law”. Parliament would have an important role to play in reviewing, repealing, amending and replacing legislation, a process which is predicted by many to be complex and time-consuming. Once the UK had formally triggered Article 50, its timescales would apply independently of Parliament approving domestic legislative changes associated with leaving the EU.

Jump to full report >>

Openness, Transparency and the Right of Access to Documents in the EU



Abstract . Upon request of the PETI Committee, the Policy Department on Citizens’ Rights and Constitutional Affairs commissioned the present analysis, which examines the situation in relation to openness, transparency, access to documents and information in the EU. Case law and developments in the jurisprudence of the CJEU are examined, notably for legislative documents, documents relating to administrative proceedings, to Court proceedings, infringement proceedings and EU Pilot cases, protection of privacy and international relations. Current and future challenges, as well as conclusions and policy recommendations are set out, in order to ensure compliance with the Treaties’ and Charter of Fundamental Rights’ requirements aimed at enhancing citizens’ participation in the EU decision-making process, and consequently stronger accountability and democracy in the EU.


The Treaty of Lisbon, in force since December 2009, includes a number of reforms emphasising open-decision making, citizen participation and the role of transparency and good administration in building up the democratic credentials of the European Union (EU).

As regards democratic decision-making and transparency in particular, a specific Title in the Treaty on the European Union (TEU) now includes a number of core provisions on democratic principles, applicable in all areas of Union action.

They underline the principle of representative democracy through the European Parliament, representing the citizens directly at Union level, and through the governments forming the European Council and the Council and that are democratically accountable either to their national parliaments, or to their citizens.1

Even participatory democracy enjoys a pivotal role in the new Treaty framework; in order to guarantee the right of ’every citizen’ to ’participate in the democratic life of the Union’, the Treaty establishes that ’[d]ecisions shall be taken as openly and as closely as possible to the citizen’ and that both citizens and representatives should be given opportunities to ’make known and publicly exchange their views in all areas of Union action’.2

These provisions have a linkage both with the new citizens’ initiative3 and with Article 15 TFEU, which places the legislature under an obligation to act publicly, and establishes that citizens have the right to access documents held by all Union institutions, bodies and agencies.

The right of access to documents, and its nature as a fundamental right, is further emphasised by Article 42 of the EU Charter of Fundamental Rights, which now enjoys ‘the same legal value as the Treaties’.4

In practice, open decision-making is to a large extent realised through the right of the general public to access documents. Regulation No 1049/2001 on public access to documents held by the EU institutions (Access Regulation),5 builds on the principle of ‘widest possible access’, and has together with case law been instrumental in operationalising the right of citizen access by establishing procedures and standards for the exercise of their democratic rights.

All documents held by the European Parliament, Council and Commission are public, as the main principle, but certain public and private interests are protected through specific exceptions under Article 4. But as exceptions derogate from the principle of the widest possible public access to documents, they must, according to established case-law, be interpreted and applied narrowly.6

Article 15(3) TFEU extends the public right of access to documents of all the Union institutions, bodies, offices and agencies. The Court of Justice, the European Central Bank and the European Investment Bank are subject to this provision only when exercising their administrative tasks.

The original 2001 Regulation only directly applies to the European Parliament, the Council, and the Commission. However, its application has been extended to the agencies by virtue of a specific provision in their respective founding acts. Furthermore, a number of institutions and bodies have adopted voluntary acts laying down rules on access to their documents which are identical or similar to Regulation No 1049/2001.

It has been 15 years since the adoption of Regulation No 1049/2001. In the same time frame the Commission and the Council set about adopting internal rules based on their rules of procedure on security and other classifications for documents. Such rules continue to exist in amended form today and exist alongside the legislative rules on access to documents.

Discussions on the reform of Regulation No 1049/2001 have been pending since 2008.7

While one would think that the tendency was – in line with the recent Treaty reforms – to strengthen the rights of citizens further, in fact the opposite seems to be the case, with discussions on reform mainly circulating around new ways to limit citizen access,8 many of them in rather fundamental ways that seem to be at odds with the letter of the Treaties.

These discussions bear witness to what seems to be a change of paradigm and priorities.

The tendency since the Treaty of Maastricht has been to strengthen the rights of citizens,9 now this objective seems lees squarely at the forefront of either the policy agenda or actual institutional practice. Staffan Dahllöf, a journalist specialising in freedom of information, describes the situation as follows: The voices asking for openness and citizen’s involvement are today weaker and fewer than they were when the present rules were decided in 2001 – at least amongst the Member State governments, and definitely in the Commission. It’s more like the Empire strikes back.10

Since there is a complete impasse in the legislative procedure (already for a very long time) on amending the 2001 Regulation, the role of the CJEU is very much centre-stage with litigants attempting to challenge a range of embedded secretive practices across a range of institutions and tasks.11

From a democratic point of view this can be considered problematic as it shifts responsibility from the EU legislator to the courts who cannot re-design the system in the required manner but deal with issues on a case by case basis, as and when they are brought before it. The same applies to the European Ombudsman, although her work is increasingly significant in bringing specific secretive practices to light and tackling them both on a case by case basis and more structurally through a growing number of own initiative enquiries.

Keeping in mind Dahllöf’s accurate observation quoted above, opening negotiations on the reform of Regulation No 1049/2001 naturally brings with it a risk of discussions leading to a further tightening of the EU transparency regime. The current Commission is not necessarily positively disposed to increasing transparency (as evidenced in legal observations before the CJEU in particular), and it has the backup of the majority of Member States in the Council.

Despite this, we think that there should be an open discussion about the possibilities of increasing openness. If this proves to be impossible, the Parliament can always block any reform that could result in negative outcomes or a levelling down.

In this note we discuss recent developments in jurisprudence and the challenges that currently exist in the application of the Regulation No 1049/2001 with a focus on public access by citizens. We conclude with a number of policy recommendations for consideration.


NOTES (to the section above)

1 Article 10(1) and (2) TEU.
2 Article 10(3) TEU, Article 11 TEU.
3 See Regulation No 211/2011 on the citizens’ initiative, OJ L [2011] 65/1.
4 Article 6(1) TEU.
5 Regulation (EC) No 1049/2001 of the European Parliament and of the Council of 30 May 2001 regarding public access to European Parliament, Council and Commission documents, OJ L 145/43.
6 See e.g. C-280/11 P Council v Access Info Europe para 30 and the case law quoted in the paragraph.
7. See e.g. Ian Harden, ‘The Revision of Regulation 1049/2001 on Public Access to Documents’, 15(2) European Public Law (2009) 239-256.
8 See the open letter by Beatrice Ask, Minister for Justice, Sweden and Anna-Maja Henriksson, Minister of Justice, Finland, published at http://www.wobbing.eu/sites/default/files/Open%20letter.pdf.
9 For one account of the EU’s transparency development so far, see Deirdre Curtin, ’Judging EU Secrecy’, Cahiers de Droit Européen, 2012 (2) 459 – 490.
10 Staffan Dahllöf, ‘Guide to the battle of transparency – UPDATED’, 09/06/2012, available at the EU wobbing website http://www.wobbing.eu/news/guide-battle-transparency-%E2%80%93-updated. On the varying positions of the Member States to the reform process, see M.Z. Hillebrandt, D.M. Curtin and A.J. Meijer, ‘Transparency in the EU Council of Ministers: An Institutional Analysis’, 20(1) European Law Journal, 2014, 1-20.
11 For a discussion, see Päivi Leino, “Transparency, Participation and EU Institutional Practice: An Inquiry into the Limits of the ‘Widest Possible’”, EUI Working Paper (LAW 3/2014).

Referendum Briefing 3: Does the EU have a ‘democratic deficit’?


by Professors Laurent Pech and Steve Peers

The EU is regularly accused from suffering a ‘democratic deficit’. In particular, it is often asserted that all EU decisions are made by the EU Commission – who are ‘unelected bureaucrats’.

As we demonstrate in this post, this criticism is clearly invalid. It fundamentally misunderstands (a) the powers that the Commission has – and more generally how decisions are made in the European Union; and (b) the way in which the European Commission gets into office. We will examine each of those issues in turn.

Who does what? Does the Commission adopt all EU laws?

The crucial thing about the EU system is that the Commission does not have anywhere near as much power as many people think it has. The standard EU’s decision-making process is: the Commission alone makes legislative proposals. Those proposals are thenconsidered by the Council (the intergovernmental body representing elected national governments), jointly with the elected Members of the European Parliament (representing EU citizens), whose powers have been gradually and significantly increased over the last three decades. Some people say that the Council and European Parliament simply rubber-stamp Commission proposals, but that is not true: they sometimes reject them and almost always amend them.

This graph illustrates the EU decision-making process: Continue reading “Referendum Briefing 3: Does the EU have a ‘democratic deficit’?”

EU Referendum Briefing 1: Can the UK control the EU’s future if it stays a member?


by Steve Peers

During the EU referendum campaign, a number of arguments have been made that staying in the EU is risky, because of possible future developments of the EU itself. While there will always be someone somewhere who says they would like to see an EU army, or some development related to the single currency, such an expression of opinion is meaningless by itself.  The fundamental issue is whether the UK could control such developments – either by vetoing them or opting out.

So what’s the worst that can happen? In this post, I’ll examine in turn the main alleged risks to staying in the EU. As we’ll see, in every single case the UK has control, either by an opt-out or a veto. In other words, none of these things can happen without the British government’s consent. Nearly all of them would also need our Parliament’s consent. And the large majority – all the fundamental possible changes to the EU that many are concerned about – would actually need the consent of the British public in another referendum. (Anyway, there’s nothing to stop the UK holding another referendum on EU membership in future, if it wanted to).

All of these safeguards for UK control of further developments of the EU exist in the current law of the EU – as I will show in detail. None of them are first created by the renegotiation of EU membership agreed this February.

I’ll look at seven issues where the UK has control over future EU developments:

a) defence;
b) transfers of power;
c) new Member States, including Turkey;
d) taxation;
e) non-EU immigration, asylum and criminal law;
f) the single currency; and
g) the EU budget, including the UK rebate.

There’s also an earlier blog post on the controversial issue of the planned EU/US trade deal (TTIP) and the NHS.

a) EU Defence and foreign policy Continue reading “EU Referendum Briefing 1: Can the UK control the EU’s future if it stays a member?”

Accord politique ou juridique : Quelle est la nature du “machin” conclu entre l’UE et la Turquie en matière d’asile?

ORIGINAL PUBLISHED HERE  on 10 Friday Jun 2016

Par Olivier Cortenet Marianne Dony, Professeurs ordinaires à l’Université libre de Bruxelles,

Alors que trois demandeurs demandeurs d’asile – apparemment deux Pakistanais et un Afghan dans les affaires T-192/16, 193/16 et 257/26 – ont demandé au Tribunal de l’Union Européenne l’annulation de l’accord conclu le 18 mars 2016 entre l’UE et la Turquie, il est permis de s’interroger sur la nature exacte de ce “machin” considéré par le service juridique du Parlement Européen comme un simple accord politique, sachant cependant que la recevabilité du recours sera tout d’abord au coeur des débats…

Les autorités européennes affirment à l’unisson que la « déclaration UE-Turquie », dont le contenu a été détaillé dans un communiqué de presse du Conseil européen du 18 mars dernier, n’est pas un accord international mais une simple déclaration d’intention. Qu’en est-il vraiment, au regard des règles du droit de l’Union européenne et du droit international public ?

Si l’on s’en réfère au droit de l’Union européenne, il est permis d’en douter au vu de la jurisprudence de la Cour de justice (de l’Union européenne).

Ainsi, dans un arrêt du 23 mars 2004, C-233/02, France c. Commission (points 42 à 45), la Cour s’est interrogée sur le point de savoir si des « lignes directrices » finalisées en février 2002 par communication entre les négociateurs des services de la Commission et leurs homologues américains et sur lesquelles aucune signature n’avait été apposée pouvaient être considérées comme un accord international. La Cour a indiqué que le critère décisif était de savoir si ces lignes directrices avaient, ou non, force obligatoire et qu’à cette fin, il fallait s’en référer à l’intention des parties. En l’espèce, la Cour a alors constaté qu’il résulte explicitement du texte de ces lignes directrices que les parties avaient l’intention de les appliquer sur une base volontaire et que de surcroît l’intention des parties de ne pas contracter d’engagements obligatoires avait été à maintes reprises expressément réitérée durant la phase de négociations des lignes directrices. C’est sur cette base que la Cour conclut que ces lignes directrices ne constituent pas un accord international et ne sont donc pas visées par l’article 300 CE, devenu article 218 TFUE.

Par ailleurs, dans l’arrêt du 26 novembre 2014, C-103/12 et C-165/12, Parlement et Commission c. Conseil (points 60 à 74) la Cour a analysé le contenu et le but d’une déclaration relative à l’attribution de possibilités de pêche dans les eaux de l’Union européenne à des navires de pêche battant pavillon » de la République du Venezuela pour considérer qu’elle devait s’analyser comme une offre adressée à cette dernière, subordonnée au respect de certaines conditions précises et que la République du Venezuela  en transmettant des demandes d’autorisation de pêche dans les eaux concernées avait consenti à cette offre. Elle a dès lors conclu qu’un accord avait bien été conclu entre ces dernières, en ajoutant que « le fait qu’un tel accord est formalisé dans un seul document commun ou dans deux ou plusieurs instruments écrits connexes est dépourvu de pertinence ».

Il résulte donc clairement de ces arrêts que la forme n’importe pas. Ce n’est pas parce que le choix d’une déclaration, ou d’un communiqué de presse, a été fait qu’il ne peut s’agir d’un accord international. Au contraire, il faut analyser le but et le contenu de cette déclaration pour déterminer si elle contient des engagements ayant force obligatoire.

Et, en passant à l’analyse de la déclaration UE-Turquie, une lecture même rapide permet de relever que l’UE et la Turquie « sont convenues » d’un certain nombre d’actions pour atteindre un objectif commun. Au nombre de ces actions:

  • il est d’abord prévu que « tous les nouveaux migrants en situation irrégulière qui partent de la Turquie pour gagner les îles grecques à partir du 20 mars 2016 seront renvoyés en Turquie », étant entendu que tous les « coûts des opérations de retour des migrants en situation irrégulière seront pris en charge par l’UE » ;
  • ensuite, « pour chaque Syrien renvoyé en Turquie au départ des îles grecques, un autre Syrien sera réinstallé de la Turquie vers l’UE », un mécanisme devant, à cette fin, être « mis en place, avec le soutien de la Commission, des agences de l’UE et d’autres États membres » ;
  • de plus, la Turquie « prendra toutes les mesures nécessaires pour éviter que de nouvelles routes de migration irrégulière, maritimes ou terrestres, ne s’ouvrent au départ de son territoire en direction de l’UE »  et l’UE « accélèrera (…) le versement du montant de trois milliards d’euros initialement alloué au titre de la facilité en faveur des réfugiés en Turquie ».
  • Finalement la déclaration affirme que tous « ces éléments progresseront en parallèle et feront l’objet d’un suivi mensuel mené conjointement ».
  • A titre accessoire, on mentionnera que le considérant 4 de la proposition de décision modifiant la décision du Conseil du 22 septembre 2015 instituant des mesures provisoires en matière de protection internationale au profit de l’Italie et de la Grèce, déposée par la Commission le 21 mars 2016, indique que « les chefs d’État ou de gouvernement sont convenus, le 7 mars, d’une série de principes devant constituer la base d’un accord avec la Turquie … », que la proposition entend partiellement mettre en œuvre. Nous noterons que c’est la Commission elle-même qui a souligné ces termes dans la présentation de sa proposition.

L’ensemble de ces éléments est de nature à indiquer que la déclaration ne contient pas simplement un certain nombre d’actions que les parties entendent appliquer sur une base volontaire mais bien des engagements à caractère obligatoire. On peut donc conclure que cette déclaration constitue bien en réalité un accord international.

Conséquences en droit européen

La conséquence en est que cette déclaration relève de l’article 218 TFUE. En effet, ainsi que l’a indiqué expressément la Cour dans son arrêt du 26 novembre 2014 (point 83), cet article « régit la négociation et la conclusion des accords entre l’Union et des pays tiers », étant entendu que le terme « accord » utilisé à cet article « doit être compris dans un sens général, pour désigner tout engagement pris par des sujets de droit international et ayant une force obligatoire, quelle qu’en soit la qualification formelle ».

En outre, comme l’a souligné la Cour de justice dans un arrêt du 16 juillet 2015C-425/13, Commission c. Conseil (point 62), cette disposition « constitue une norme autonome et générale de portée constitutionnelle, en ce qu’elle attribue aux institutions de l’Union des compétences déterminées », en « visant à établir un équilibre entre ces dernières ».

Que prévoit cette disposition? (voy. à ce sujet, M. Dony,   Droit de l’Union européenne, points 358 à 368). Il résulte de l’article 218, paragraphe 6, TFUE, que c’est le Conseil, sur proposition du négociateur (qui est normalement la Commission), qui adopte, en principe à la majorité qualifiée, une décision portant conclusion de l’accord, et ce après approbation du Parlement européen, lorsque cet accord couvre un domaine auxquels s’applique la procédure législative ordinaire, ce qui est le cas pour la politique d’asile (article 78, paragraphe 2, TFUE) et d’immigration (article 79, paragraphe 2, TFUE).

Il est patent qu’en l’espèce, cette procédure n’a absolument pas été respectée : en effet, il n’y a pas trace d’une quelconque proposition de la Commission ; la « déclaration UE-Turquie » a fait l’objet d’un communiqué du Conseil européen et non d’une décision du Conseil et enfin, last but not least, le Parlement européen, non seulement n’a pas approuvé l’accord, mais n’a tout simplement pas été impliqué du tout. Il est même permis de penser que le terme « déclaration » a été consciemment choisi pour notamment éluder l’application de la procédure de l’article 218 TFUE.

En vertu d’une jurisprudence constante, le non-respect de la procédure de l’article 218 TFUE est de nature à affecter la légalité de la décision de conclusion d’un accord. C’est ainsi que la Cour, dans son arrêt précité du 26 novembre 2014, a annulé la décision adoptant la déclaration en cause au motif que le Parlement avait été simplement consulté et n’avait pas donné son approbation à l’accord. De même, dans un arrêt du 9 août 1994, C-327/91, Conseil c. Commission, la Cour a annulé l’acte par lequel la Commission avait entendu conclure un accord avec les Etats-Unis concernant l’application de leur droit de la concurrence, après avoir jugé que l’accord aurait dû être conclu par le Conseil, qui seul détient la compétence de conclusion des accords internationaux.

Cela étant, deux difficultés se présentent dans le cadre du droit de l’Union européenne:

D’abord, la question se pose de savoir comment il est possible de contester la validité du communiqué de presse contenant la « déclaration UE-Turquie », qui doit être considérée comme « la décision de conclusion » de cet accord.

La voie à laquelle chacun pense spontanément est l’action en annulation prévue par l’article 263 TFUE. La forme qu’a prise cette « décision » n’est pas un obstacle, car il est de jurisprudence constante que « recours en annulation doit être ouvert à l’égard de toutes les dispositions prises par les institutions de l’Union, indépendamment de leur nature ou de leur forme, à condition qu’elles visent à produire des effets de droit », comme l’a rappelé la Cour de justice dans son arrêt précité du 16 juillet 2015 (point 26).

Ce recours est ouvert sans conditions aux Etats membres, au Parlement européen, au Conseil ou à la Commission. En revanche, les personnes physiques ou morales doivent quant à elle établir que l’acte dont elles demandent l’annulation les concerne « directement et individuellement » et il paraît très difficile voire téméraire de soutenir que tel pourrait être le cas d’un tel acte. Elles devraient donc s’en remettre à une action introduite par un des requérants dits privilégiés, ou plus précisément au Parlement européen ou un Etat membre.

Une voie plus indirecte, et plus incertaine, pourrait être envisagée, qui consisterait, en application de l’article 267 TFUE, à saisir une juridiction nationale et de lui demander de poser à la Cour de justice une question préjudicielle en appréciation de validité de cette « décision ».

Ensuite, il résulte d’une jurisprudence constante que la Cour de justice peut seulement annuler ou invalider la décision de conclusion d’un accord international mais non l’accord lui-même. Ainsi, dans l’arrêt précité du 9 août 1994, la Cour a interprété le recours en annulation de l’accord introduit par la France comme étant dirigé contre l’acte par lequel la Commission a entendu conclure cet accord. La question de la conséquence d’une telle annulation sur la validité de l’accord international doit être abordée au seul regard du droit international public, vers lequel il importe donc de se tourner.

Conséquences en droit international

A cet égard, en droit international public, la notion de traité est entendue assez largement, comme en témoigne cette définition codifiant le droit coutumier reprise dans la Convention de Vienne de 1986 sur le droit des traités entre Etats et organisations internationales ou entre organisations internationales : « […] l’expression ‘traité’ s’entend d’un accord international régi par le droit international et conclu par écrit […] que cet accord soit consigné dans un instrument unique ou dans deux ou plusieurs instruments connexes, et quelle que soit sa dénomination particulière » (voy. les commentaires de cet article et des autres dispositions pertinentes des conventions de Vienne sur le droit des traités dans O. Corten et P. Klein (dir.), Les Conventions de Vienne sur le droit des traités. Commentaire article par article).

La dénomination, ou plus généralement la forme de l’accord, n’importe donc pas. Ont ainsi été considérés comme des traités entre Etats (mais on peut sans aucun doute transposer ces enseignements aux traités entre Etats et organisations internationales) : un échange de lettres (C.I.J., Affaire du Différend territorial (Libye/Tchad), Recueil 1994, not., § 31) , un simple procès-verbal de réunion (C.I.J., Affaire de la Délimitation maritime et des questions territoriales entre le Qatar et Bahreïn, Recueil 1994, § 23.), un communiqué conjoint (C.I.J., Affaire du Plateau continental de la Mer Egée, Recueil 1978, §§ 95-98) ou encore … une déclaration commune (C.I.J., Affaire de la Frontière terrestre et maritime entre le Cameroun et le Nigéria, Recueil 2002, § 263). Le caractère informel de la « déclaration UE-Turquie » ne constitue donc aucunement un obstacle à sa qualification de traité. Le droit international et le droit européen se rejoignent sur ce point.

La seule condition potentiellement problématique en l’espèce est celle selon laquelle l’accord doit être « régi par le droit international », ce qui suppose une volonté des parties de produire des engagements juridiques relevant du droit international public. Ici aussi, les critères du droit international sont très similaires à ceux dégagés par la Cour de justice qui d’ailleurs s’y réfère explicitement.

En l’espèce, et comme indiqué ci-dessus (point 4), la « déclaration UE-Turquie » a pour objet la circulation des personnes et le statut de réfugié, un domaine qui relève indéniablement de l’ordre juridique international. La terminologie utilisée témoigne d’ailleurs d’une volonté de s’engager, comme en témoignent les extraits suivants : « l’UE et la Turquie ont décidé ce jour de … »,  « sontconvenues des points d’action complémentaires suivants… »… « La Turquieprendra toutes les mesures nécessaires pour éviter que… » ; … « un programme d’admission humanitaire volontaire sera activé. Les États membres de l’UE ycontribueront… ». Les termes ainsi soulignés ont une portée qui, loin d’être simplement exhortative ou indicative, est normative et prescriptive. De manière tout à fait explicite, les parties précisent encore que : « Cela se fera en totale conformité avec le droit de l’UE et le droit international » ou encore « conformément aux normes internationales applicables ».

Bref, au vu des termes mêmes de cette déclaration, il paraît difficile de lui dénier la qualité de « traité », au sens du droit international public. Cela étant, le non-respect des procédures internes de l’UE lors de la conclusion de ce traité pourrait avoir des conséquences non sur l’existence, mais sur la validité de ce dernier. Le droit coutumier en la matière est exprimé dans deux dispositions de la Convention précitée de Vienne de 1986 :

« Article 27 […]

2. Une organisation internationale partie à un traité ne peut invoquer les règles de l’organisation comme justifiant la non-exécution du traité.

3. Les règles énoncées dans les paragraphes précédents sont sans préjudice de l’article 46 ».

« Article 46 […]

2. Le fait que le consentement d’une organisation internationale à être liée par un traité a été exprimé en violation des règles de l’organisation concernant la compétence pour conclure des traités ne peut être invoqué par cette organisation comme viciant son consentement, à moins que cette violation n’ait été manifeste et ne concerne une règle d’importance fondamentale.

3. Une violation est manifeste si elle est objectivement évidente pour tout Etat ou toute organisation internationale se comportant en la matière conformément à la pratique habituelle des Etats et, le cas échéant, des organisations internationales et de bonne foi ».

Il en découle que, si, en principe, l’UE ne pourrait se prévaloir de la violation de ses propres règles pour justifier la non-exécution du traité, une exception est cependant réservée aux cas de violation « manifeste » concernant une règle d’une « importance fondamentale ».

Pour ce qui est de la deuxième condition,  l’accord a été conclu  en violation totale du prescrit de l’article 218 TFUE (point 7). Or cette disposition doit indéniablement être considérée comme une règle d’une importance fondamentale. La Cour internationale de justice a ainsi indiqué que  « les règles relatives au pouvoir de signer des traités au nom d’un Etat sont des règles constitutionnelles d’une importance fondamentale » (Affaire de la Frontière terrestre et maritime entre le Cameroun et le Nigéria, précitée, § 265) et cette affirmation peut assurément être transposée au droit d’une organisation internationale comme l’UE, d’autant que, comme déjà indiqué, la Cour de justice de l’Union européenne lui a expressément accordé ce statut. On peut d’ailleurs penser que c’est ce qui explique que le Conseil européen ait préféré dénier la qualité d’accord à la déclaration commune avec la Turquie : par le biais de cette qualification, sans doute espérait-il échapper à l’obligation de respecter l’article 218 TFUE…

Peut-on également considérer que cette violation était « manifeste » et donc « objectivement évidente » pour la Turquie ? Dans l’affaire que l’on vient de citer, la Cour internationale de justice est restée très prudente sur cette question, puisqu’elle estime que la connaissance par un Etat d’une règle constitutionnelle de droit interne d’un autre Etat suppose que ce dernier l’ait « rendu[e] publique de manière appropriée ».

Dans notre cas, on n’est cependant pas devant une règle de droit interne, mais devant des règles conventionnelles qui sont par définition publiques, les traités européens faisant l’objet de diverses publications, écrites et électroniques. Plus précisément, il est difficile d’imaginer que la Turquie, candidate à l’adhésion et engagée dans des négociations à cet effet depuis des années, n’ait pas connaissance des principes régissant la conclusion des accords internationaux de l’Union et en particulier des principes selon lesquels le Parlement européen doit être impliqué dans la procédure de conclusion de ces accords et que le Conseil européen n’est  en revanche pas habilité à y intervenir.

En conclusion

Dans ces conditions, et toujours si l’on s’en tient au droit international public, il apparaît que la « déclaration UE-Turquie » peut a priori être qualifiée de traité, mais un traité dont il est sérieusement permis de douter de la validité. Rien ne devrait dès lors sérieusement empêcher de remettre en cause les effets juridiques de cette déclaration, que ce soit en droit européen ou en droit international public.

Casse-tête jurisprudentiel autour de l’exequatur (A propos des arrêts Meroni et Avotins)


by Maxime Barba, EDIEC

Les arrêts Avotins et Meroni, rendus coup sur coup selon un timing remarquable, l’un par la Grande Chambre de la Cour EDH, l’autre par la CJUE en formation classique, mettent aux prises la discipline européenne de l’exequatur et les exigences du procès équitable. Si le relief individuel de ces décisions est déjà remarquable, leur comparaison n’en est que plus exceptionnelle. Certains enseignements fondamentaux peuvent être retirés d’un bref exercice de confrontation.

1. Dans sa configuration, l’affaire Avotins (CEDH, 23 mai 2016, req. n°17502/07) est relativement connue et l’on renverra sur ce point à d’autres contributions (sur ce blog : -S. BERGE, Avotins ou le calme qui couve la tempête; J.-S. BERGE, Une, deux et… trois lectures : de l’avis 2/2013 (CJUE) à l’affaire Avotins (CEDH); notre contribution ; dans la littérature spécialisée, v. not. F. MARCHADIER, « Présomption d’équivalence dans la protection des droits fondamentaux », RCDIP 2014.679 ; P. DEUMIER, « Le règlement Bruxelles I, l’exequatur et la CEDH », RDC 2014.428).

Quant aux aspects factuels de l’affaire Meroni (CJUE, 25 mai 2016, C-559/14), il faut brièvement en faire état. Synthétiquement, M. Meroni s’opposait, au for polonais, à l’exequatur d’une ordonnance de gel, décidée au for anglais, destinée à un autre que lui et qui produisait malgré tout des effets indésirables à son endroit (pts. 20 et ss). Pour ce faire, il se fondait techniquement sur l’article 34§1 du règlement n°44/2001, dont nul n’ignore qu’il met en place l’exception d’ordre public international qu’il est possible d’opposer à un jugement étranger dont l’exequatur est requis. Plus précisément, il était suggéré d’avoir recours à l’ordre public international pris dans sa composante procédurale et, au regard des fondements, de lire l’article 34§1 du règlement « Bruxelles I » en conjugaison avec l’article 47 de la Charte des droits fondamentaux de l’Union européenne (CDFUE).

2. Juridiquement, la question préjudicielle posée était de savoir si « la reconnaissance et l’exécution d’une ordonnance rendue par une juridiction d’un Etat membre, qui a été prononcée sans qu’un tiers dont les droits sont susceptibles d’être affectés par cette ordonnance ait été entendu, doivent être considérées comme manifestement contraires à l’ordre public de l’Etat membre requis et au droit à un procès équitable» au sens de l’article 34§1 du règlement « Bruxelles I » lu à la lumière de l’article 47 de la CDFUE. La réponse de la Cour de Justice passe par plusieurs temps, dont certains sont classiques en matière d’ordre public international, et d’autres moins.

3. Quelques rappels sont judicieusement articulés quant à cette notion d’ordre public international à la spécificité indéniable. Tout d’abord, celle-ci est d’interprétation stricte (pt. 38). Ensuite, si la détermination de son contenu relève de la liberté des Etats membres, elle ne relève pour autant pas de leur discrétion puisqu’un contrôle de la Cour de Justice est toujours envisageable au regard de la mise en œuvre de l’exception (pts. 39 et 40). Enfin, il n’est pas question de procéder à une révision au fond sous le couvert de l’exception d’ordre public international (pt. 41). En bref, « un recours à l’exception d’ordre public […] n’est concevable que dans l’hypothèse où la reconnaissance ou l’exécution de la décision rendue dans un autre Etat membre heurterait de manière inacceptable l’ordre juridique de l’Etat membre requis, en tant qu’elle porterait atteinte à un principe fondamental» (pt. 42), atteinte qui « devrait constituer une violation manifeste d’une règle de droit considérée comme essentielle dans l’ordre juridique de l’Etat membre requis ou d’un droit connu comme fondamental dans cet ordre juridique» (idem). Ce sont là de solennels rappels mais qui demeurent classiques. L’intérêt se situe ailleurs.

4. Le passage suivant de l’arrêt Meroni est décisif : « le règlement n°44/2001 repose sur l’idée fondamentale selon laquelle les justiciables sont tenus, en principe, d’utiliser toutes les voies de recours ouvertes par le droit de l’Etat membre d’origine. Sauf circonstances particulières rendant trop difficile ou impossible l’exercice des voies de recours dans l’Etat membre d’origine, les justiciables doivent faire usage dans cet Etat membre de toutes les voies de recours disponibles afin d’empêcher en amont une violation de l’ordre public» (pt. 48, cnqs). Le cœur de l’arrêt est là. La Cour de Justice met explicitement en place un principe général de mise en œuvre préalable des voies de recours au pays d’origine avant tous recours à la clause d’ordre public international au pays requis. Ce positionnement, déjà présent dans la célèbre jurisprudence Diageo Brands (CJUE, 16 juill. 2015, C-681/13; sur ce blog : NOURISSAT, De l’ « encadrement » à la « mise sous tutelle » de l’exception d’ordre public international en matière de reconnaissance des décisions civiles ou commerciales… Nouvelle étape sous la plume de la CJUE), appelle une double analyse : l’une externe, en contemplation du très récent arrêt Avotins de la CEDH ; l’autre interne, au regard de la jurisprudence de la CJUE en matière d’ordre public international (v. billet C. NOURISSAT à paraître). On se concentrera ici sur la première de ces analyses. Si la ressemblance entre les arrêts Avotins et Meroni est palpable et incite à évoquer une certaine convergence des solutions (I), un examen plus attentif permet de parler d’une divergence fondamentale, non seulement en termes de raisonnement, mais surtout en termes de posture des juges européens sur la question du recours préalable au for d’origine en matière d’exequatur (II).

I) La convergence des solutions

5. Les affaires Avotins et Meroni présentent une parenté presque intuitive, jusqu’à l’Etat concerné, la Lettonie. La chose n’avait d’ailleurs nullement échappé à l’avocat général Kokott, chargée de conclure dans l’affaire Meroni, qui s’était abondamment référée à l’arrêt de 2014 intervenu dans l’affaire Avotins (Conclusions présentées le 25 février 2016, pts. 39 et 40). Les ressemblances entre ces deux affaires sont sensibles. Dans leur configuration d’abord : dans un cas comme dans l’autre, la personne procéduralement lésée, défendeur dans Avotins, tierce dans Meroni, n’avait pu faire valoir ses droits lors de l’instance directe au for d’origine. Et, dans un cas comme dans l’autre, la personne procéduralement lésée n’avait pas mis en œuvre les recours disponibles au for d’origine. Et, dans les deux hypothèses, la solution finalement adoptée est incontestablement à la défaveur de la partie absente, dont on attendait donc qu’elle mette en œuvre préalablement les recours au for d’origine. Ainsi, il revenait bien au requérant, dans l’affaire Avotins, de s’employer au for chypriote à recourir contre la décision de première instance l’ayant condamné. À défaut de quoi, son inertie procédurale pouvait lui être reprochée au for letton, au regard du droit européen incarné par le règlement « Bruxelles I » pris en son article 34§2, comme – mais dans une mesure moindre à notre sens – au regard du droit de la Convention, lequel interdit au requérant de se plaindre devant la Cour d’une situation qu’il aurait lui-même contribué à créer « par son inaction et son manque de diligence» (CEDH, 23 mai 2016,Avotins, §124 ; v. égal. CEDH, 6 mai 2004, Hussin c/ Belgique, req. n°70807/01 et CEDH, 29 mai 2008, McDonald c/ France, req. n°18648/04). De même, il revenait au tiers lésé dans l’affaire Meroni de procéder d’abord au for anglais d’origine avant d’exciper une éventuelle violation de l’ordre public international de procédure au for polonais requis.

6. Dans les deux cas, le message est clair : ce n’est qu’après épuisement du contentieux au for d’origine, en amont, qu’il est possible de se plaindre d’une éventuelle iniquité procédurale au for requis, en aval. La différence de fondement, articles 34§1 ou 34§2 du règlement n°44/2001, n’apparaît pas décisive : les conclusions semblent convergentes, si ce n’est similaires. Le principe du recours au pays d’origine se présente comme invariable. Cette belle cohérence vole pourtant en éclats à l’analyse car si les deux arrêts se rejoignent en termes de résultat concret, ils diffèrent radicalement dans leur cheminement théorique, voire jusque dans les positionnements sous-jacents respectivement adoptés par les juges européens. Derrière cette convergence des solutions se loge en réalité une véritable divergence de posture.

II) La divergence de posture

7. La divergence de points de vue peut être mise en exergue au travers d’une démarche progressive, en repartant des fondements. Dans l’affaire Avotins, la clause d’ordre public international n’avait pas été mobilisée devant le juge letton (CEDH, 23 mai 2016, §108). C’est d’ailleurs regrettable car la solution aurait pu être substantiellement différente sur le fondement de l’article 34§1 du règlement n°44/2001. Au regard de la technique internationaliste, tout d’abord, il demeure en effet que l’article 34§1 n’exige textuellement pas la mise en œuvre préalable des recours au pays d’origine. Et, à une époque où la jurisprudence Diageo Brands n’avait pas encore émergé, la chose a son importance. Ensuite, au regard du droit de la Convention, la notion d’ordre public international induit des différences notables. Si la présomption Bosphorus a pu être mobilisée par la Cour EDH dans cette affaire, c’est en raison de l’absence de marge de manœuvre du juge letton dans la mise en œuvre du droit de l’Union (la chose se discute d’ailleurs, CEDH, 23 mai 2016, §§105 et ss). Or, la notion d’ordre public international ne se prête pas aux mêmes conclusions, les Etats membres conservant une certaine liberté, même encadrée, en la matière (supra, n°3). Ainsi, le point d’entrée « ordre public international », bien présent dans l’arrêt Meroni, ne se retrouve techniquement pas dans l’arrêt Avotins. Et de cette différence de fondements s’induisent d’importantes différences dans le raisonnement, lesquelles laissent transparaître une véritable divergence de points de vue.

8. Le principe de recours préalable au pays d’origine est le fruit d’une construction dans l’affaire Meroni alors qu’il est le fruit d’une déduction dans l’affaire Avotins. Ce constat est essentiel. La conclusion de l’affaire Meronin’est pas mécaniquement extraite de l’article 34§2 et de son libellé explicite, mais artificiellement construite sur le fondement de l’article 34§1 – d’une façon relativement poussive qui plus est. La différence dans la lettre des deux dispositions devait ainsi être méthodiquement dépassée, le juge de Luxembourg se retrouvant forcé d’élever le principe de recours préalable au for d’origine au rang d’« idée fondamentale» (CJUE, Meroni, préc., pt. 48), expression qui mériterait d’ailleurs l’analyse. Synthétiquement, la Cour de Justice fait du recours préalable au for d’origine son cheval de bataille.

9. Dans l’affaire Avotins, c’est bien l’application de la présomption de protection équivalente, combinée à quelques observations propres à l’espèce, qui mène à la conclusion du recours préalable au pays d’origine. La solution n’émane alors pas d’une confrontation directe de la difficulté au droit de la Convention. Et d’ailleurs, là où tout s’était fait de façon mécanique et automatique en 2014, la Cour EDH exprime en 2016 ses réserves, et de la belle manière. Un extrait doit être évoqué, tout à fait représentatif : « Lorsque les juridictions des Etats qui sont à la fois parties à la Convention et membres de l’Union européenne sont appelées à appliquer un mécanisme de reconnaissance mutuelle établi par le droit de l’Union, c’est en l’absence de toute insuffisance manifeste des droits protégés par la Convention qu’elles donnent à ce mécanisme son plein effet. En revanche, s’il leur est soumis un grief sérieux et étayé dans le cadre duquel il est allégué que l’on se trouve en présence d’une insuffisance manifeste de protection d’un droit garanti par la Convention et que le droit de l’Union européenne ne permet pas de remédier à cette insuffisance, elles ne peuvent renoncer à examiner ce grief au seul motif qu’elles appliquent le droit de l’Union» (CEDH, 23 mai 2016, §116). Que dire de plus ? On le voit : la CEDH ne fait nullement du principe de recours préalable au for d’origine sa religion, à l’inverse du juge de Luxembourg. C’est la présomption Bosphorus et la réunion de circonstances factuelles, supposées accablantes pour le requérant, qui mèneront finalement à la nécessité d’un recours préalable au for d’origine et au constat de non-violation de l’article 6§1er. Mais l’insuffisance manifeste, permettant de renverser la présomption, n’était pas loin (CEDH, 23 mai 2016, §121, faisant état d’une simple « défaillance regrettable »).

10. En conclusion, là où le juge de l’Union s’emploie en matière d’exequatur à construire et à développer un principe transversal de recours préalable au pays d’origine, le juge de la Convention paraît, quant à lui, s’ingénier à tempérer cette ardeur parfois excessive, et surtout susceptible de générer des insuffisances manifestes dans la protection des garanties fondamentales dont elle a la garde. Entre les deux postures, nul doute qu’il faudra inexorablement trancher.

Data retention and national law: whatever the CJEU rules, data retention may still survive!


by Matthew White, Ph.D candidate, Sheffield Hallam University

Should governments be able to retain data on everyone’s use of the Internet and their phones – because it might arguably aid the fight against terrorism and serious crime? This ‘data retention’ issue raises fundamental questions about the balance between privacy and security, at both national and EU level. Initially, in the electronic privacy (e-Privacy)Directive, EU legislation set out an option for Member States to adopt data retention rules, as a derogation from the normal rule of confidentiality of communications in that Directive. Subsequently, in 2006, at the urging of the UK government in particular, the EU went a step further. It adopted the Data Retention Directive (DRD), which requiredtelecom and Internet access providers to keep data on all use of the Internet and phones in case law enforcement authorities requested it.

However, on 8 April 2014, the Court of Justice of the European Union (CJEU) ruled that the latter Directive went too far. In its Digital Rights Ireland judgment (discussed here), that Court said that the EU’s Data Retention Directive (DRD) was invalid in light of a lack of compliance with the rights to privacy and data protection set out in Articles 7 and 8 of the EU Charter of Fundamental Rights (CFR) (para 69 and 73). This left open an important question: what happens to national data retention laws? Can they also be challenged for breach of the EU Charter rights, on the grounds that they are linked to EU law (the derogation in the e-Privacy Directive)? If so, do the standards in the Digital Rights Ireland judgment apply by analogy?

Instead of addressing this matter urgently, the United Kingdom government sat on its hands for a while and then unprecedentedly rushed through the Data Retention and Investigatory Powers Act 2014 (DRIPA 2014). DRIPA 2014 was intended to be a reactionto the Digital Rights Ireland ruling, giving the UK as a matter of national law the power to retain data that had been struck down by the CJEU as a matter of EU law.

In 2015, Tom Watson (now the deputy leader of the UK Labour Party), David Davis (a Conservative party backbencher) and others challenged s.1 of DRIPA 2014 arguing that the powers to obligate data retention on public telecommunication operators set out in that section of DRIPA did not sufficiently reflect what the CJEU ruled in Digital Rights Ireland. Although that CJEU ruling only applied to EU legislation, they argued that it also applied by analogy to national legislation on data retention, since such legislation fell within the scope of the option to retain communications data set out in the derogation in the e-Privacy Directive, and so was linked to EU law (and therefore covered by the Charter). Even though the e-Privacy Directive only related to publicly available electronic communications services (Article 3(1)), it is submitted that any extension of the definition of public telecommunications operator would fall within the Data Protection Directive, and thus the CFR would still apply. The High Court (HC) ruled in the claimants’ favour inDavis where an order was made for s.1 of DRIPA to be disapplied by the 31st of March 2016, insofar as it is incompatible with Digital Rights Ireland (para 122). This was in the hopes that it would give Parliament sufficient time to come up with a CFR compliant data retention law (para 121).

The government appealed to the Court of Appeal (CoA) which took a radically different approach maintaining that ‘the CJEU in Digital Rights Ireland was not laying down definitive mandatory requirements in relation to retained communications data’ (para 106). But for the sake of caution, the CoA made a preliminary reference to the CJEU asking:

(1) Did the CJEU in Digital Rights Ireland intend to lay down mandatory requirements of EU law with which the national legislation of Member States must comply?

(2) Did the CJEU in Digital Rights Ireland intend to expand the effect of Articles 7 and/or 8, EU Charter beyond the effect of Article 8 ECHR as established in the jurisprudence of the ECtHR?

The CoA was not the only national court to make a preliminary reference to the CJEU on matters regarding data retention and the reach of Digital Rights Ireland. On the 4th May 2015, the Force was with Kammarrätten i Stockholm when it asked the CJEU:

Is a general obligation to retain traffic data covering all persons, all means of electronic communication and all traffic data without any distinctions, limitations or exceptions for the purpose of combating crime (as described [below under points 1-6]) compatible with Article 15(1) of Directive 2002/58/EC [the electronic privacy Directive], 1 taking account of Articles 7, 8 and 15(1) of the Charter?

If the answer to question 1 is in the negative, may the retention nevertheless be permitted where:

access by the national authorities to the retained data is determined as [described below under paragraphs 7-24], and

security requirements are regulated as [described below under paragraphs 26-31],

and all relevant data are to be retained for six months, calculated as from the day the communication is ended, and subsequently deleted as [described below under paragraphs 25]?

The way in which the first question in Davis and Watson is asked doesn’t specify whether the general obligation applies to every service provider under the state’s jurisdiction or specific service providers to retain what they individually process. The assumption is the former as ‘all means of electronic communication and all traffic data without any distinctions’ implies a catch all to the relevant services. The Home Secretary (and indeed the government) may argue that if the CJEU rules in the negative (note that Article 15(1) of the e-Privacy Directive only applies to publically available electronic communications services, thus the justification for retaining data from other services would have to be found in the Data Protection Directive (DPD)) it would mostly have affected cl.78 of theInvestigatory Powers Bill (IPB) (currently before Parliament) which would grant the Secretary of State the power to issue retention notices on a telecommunications or any number of operators to retain for e.g. any or all data for 12 if the power in cl.1 of the draft Communications Data Bill (dCDB) had been replicated. The dCDB was a legislative measure introduced in 2012 to allow public authorities to keep up to date with the sophistication of e-Crime. Clause 1 maintained that:

1 Power to ensure or facilitate availability of data

(1) The Secretary of State may by order—

(a) ensure that communications data is available to be obtained from telecommunications operators by relevant public authorities in accordance with Part 2, or

(b) otherwise facilitate the availability of communications data to be so obtained from telecommunications operators.

(2) An order under this section may, in particular—

(a) provide for—

(i) the obtaining (whether by collection, generation or otherwise) by telecommunications operators of communications data,

(ii) the processing, retention or destruction by such operators of data so obtained or other data held by such operators.

This measure was, however abandoned because the Liberal Democrats (in the then Coalition Government) did not approve of the far reaching nature of the proposal. In regards to cl.1, it clearly was a general power, as no distinction was made on who the obligation to retain may fall upon, and thus it is submitted that this power is analogous to the power which is the subject of the question being asked of the CJEU. Clause 78(1) of the IPB on the other hand, makes the distinction that a data retention notice may require a telecommunications operator to retain relevant communications data. Though there are two possible conflicts, the first, based on the assumption that the CJEU rules in the negative (to the first question) is cl.78(2)(a) and (b). This gives the Secretary of State the discretion to issue retention notices on any description of operators to retain all or any description of data. This could be considered a general obligation because it could affect all telecommunications operators and then be classed as a general obligation.

Secondly, retention ‘without distinction’ or ‘exceptions’ may be important when it comes to traffic data pertaining to journalists, politicians, and the medical and legal professions. But because the reference doesn’t mention specific service providers it cannot be said with certainty how much this would affect cl.78(1) which doesn’t make distinctions or exceptions.

When it comes to limitations on data retention, there is at least one, which was first noted in s.1(5) of DRIPA 2014 which allowed for a 12 month maximum period of retention. This is replicated in cl.78(3) and takes on board the recommendation of the Advocate General’s opinion (AG) in Digital Rights Ireland (para 149).

The President of the CJEU felt it was desirable to combine both preliminary references. The questions of access by both the Swedish and UK courts do not directly affect the cl.78 issuing of retention notices (insofar that it at least doesn’t involve everytelecommunications operator) nor does answering whether Article 7 and 8 was intended to extend beyond Article 8 ECHR jurisprudence. The security arrangements are dealt with by cl.81 (whether they are adequate is a different matter) and thus not relevant to the issuing of retention notices.

This, however, proceeds on the assumption that the CJEU will rule in the negative to the Swedish preliminary reference regarding retention being lawful for the purposes ofaccess, because if it does not, cl.78(2)(a) and (b) would not be affected at all. Moreover, the HC in Davis felt that the CJEU believed that data retention genuinely satisfied an objective of general interest (para 44) and that it must be understood to have held that a general retention regime is unlawful unless it is accompanied by an access regime which has sufficiently stringent safeguards to protect citizens’ rights set out in Articles 7 and 8 of the CFR (para 70). The CoA was silent on this matter, and therefore for the mean time, it is understood that if the CJEU rules in the positive, cl.78 would not be affected as a matter of EU law.

On the matter of whether the HC or the CoA had interpreted Digital Rights Irelandcorrectly, it is important to highlight one of the justifications for the CoA conclusions. It maintained in relation to mandatory requirements, that in the opinion of the AG, he was at least, not looking for the Directive to provide detailed regulation (para 77). Yet the CoA failed to mention his conclusions, where it was stated that the DRD was invalid as a result of the absence of sufficient regulation of the guarantees governing access to (by limiting access, if not solely to judicial authorities, at least to independent authorities, or, failing that, by making any request for access subject to review by the judicial authorities or independent authorities and it should have required a case-by-case examination of requests for access in order to limit the data provided to what is strictly necessary (para 127)) the data collected/retained and that the DRD should be suspended until the EU legislature adopts measures necessary to remedy the invalidity, but such measures must be adopted within a reasonable period (para 157-158). So at least in this regard the AG actually supports the stance of the HC (even though no reference was made on this point) and may therefore have had implications for the IPB (which does not require judicial or independent authorisation/review) in relation to access to communications data without a word from the CJEU.

Many thanks to Steve Peers for helpful comments on an earlier draft.