Thursday 30 January 2014

Pirates of the Indian Ocean: Legal Base and Democratic Debate

Steve Peers

Despite their central role in the world of EU law, legal base disputes often confound those outside the fairy-tale duchy of Luxembourg and the Brussels beltway, in particular when everyone agrees as to the substance of the issue concerned. And indeed, everyone agrees that the pirates in the Indian Ocean are a bad thing (except, presumably, the pirates themselves), and that the EU should establish (as it has done) a military action to combat them.

But what happens if the EU force catches the pirates? No-one wants the pirates coming to Europe to be tried, and it wouldn’t do to send them to Guantanamo. It would certainly be ironic if they could be forced to walk the plank, but that would violate their right to life. So they must be handed over to nearby countries in East Africa, for prosecution in those states, and everyone agrees that the EU must negotiate agreements to this end with the countries concerned.

One such treaty is between the EU and Mauritius, and the European Parliament (EP) has challenged the Council’s decision to conclude it on two separate grounds: the wrong legal base, and a failure to inform the EP sufficiently (Case C-658/11). The Advocate-General’s opinion, delivered today, is worthy of detailed analysis.

The legal base issue: foreign policy, or development and judicial cooperation? 

The Council believes that the treaty with Mauritius concerns the EU’s Common Foreign and Security Policy (CFSP) alone, while the EP believes that the treaty concerns also judicial cooperation and development. In this case, the choice of legal base has far greater consequences than usual. Either way, the treaty had to be agreed unanimously by the Council, since both parties agree that it concerns foreign policy at least in part. But if the Council is right, and the treaty only concerns foreign policy, then: the EP did not even have to be consulted; the treaty had to be negotiated by the EU foreign policy High Representative; and the CJEU has no jurisdiction (except the jurisdiction to rule on whether the Council used the right Treaty base, as in this case: see Article 275 TFEU). If the EP is right, then: the EP had the power of consent over the treaty; the treaty had to be negotiated by the Commission; and the CJEU has its full usual jurisdiction.

The Advocate-General first of all examines the EP’s arguments based on the precise wording of Article 218 TFEU, which specifies that the EP must consent to or be consulted about all treaties to which the EU will become a party, unless those treaties ‘relate exclusively’ to the CFSP. In his view, this rule simply echoes the legal distinction between the ‘legal bases’ of the CFSP and other EU policies, and so does not create a separate rule relating to the conclusion of external treaties.

Then the Advocate-General turns to the heart of the issue: which legal base applies? In his view, taking account of the overall legal context, including Security Council Resolutions addressing the threat to international security posed by the pirates and the EU’s military action to combat them, the treaty is a CFSP measure. In particular, the CFSP measure providing for the military action contains rules on the possible transfer of the pirates to third States, including human rights protection. That EU measure would not be effective without treaties with third States regulating the transfer of those pirates.

Also, the treaty falls within the scope of the CFSP due to its objectives, which include (from the EU’s general external relations objectives) the requirements that the EU act in order to: ‘safeguard its values, fundamental interests, security, independence and integrity’; ‘consolidate and support democracy, the rule of law, human rights and the principles of international law’; ‘preserve peace, prevent conflicts and strengthen international security’; and ‘promote an international system based on stronger multilateral cooperation and good global governance’ (Article 21(2) TEU). In the Advocate-General’s view, these ‘are among those [objectives] that are traditionally assigned to the CFSP’ and ‘essentially correspond’ to the CFSP objectives as they were set out in the Treaties before the Treaty of Lisbon. The activity of transferring pirates also falls within the scope of the defence policy provisions of Articles 42 and 43 TEU, which refer to the use of ‘civilian and military’ assets.

The Advocate-General also rejected the use of the EU’s powers concerning criminal judicial cooperation. In his view, the external use of the EU’s justice and home affairs powers must ‘have a close link with freedom, security and justice within the Union’, namely ‘a direct link between the aim of the internal security of the Union and the judicial and/or police cooperation which is developed outside the Union’. This was distinct from a CFSP measure which had the objective of, ‘first and foremost, peace, stability and democratic development in a region outside the Union’. In this case, transferring pirates to East African states was too far removed from the development of the EU’s justice and home affairs policies. Finally, the Advocate-General rejected the use of the EU’s development policy powers, since the assistance which the EU gives to Mauritius is linked only to the application of the rules on the transfer of pirates, which constitute (in his view) a CFSP measure.

Is this first part of the opinion convincing? Some parts are more convincing than others. Certainly, the treaty should not have a legal base relating to development policy, since the assistance being provided is purely ancillary to the transfer of pirates. But this begs the question of the legal base which should apply to the transfer of pirates.

It makes sense to apply the same legal base rules to the conclusion of international treaties as apply to the adoption of internal legislation, since the Treaty drafters have forged a strong link between those two facets of EU decision-making. On the other hand, while it is true to say that a treaty containing rules on the transfer of pirates is necessary to ensure the effectiveness of the military operation which catches them, it does not necessarily follow that it has the same legal base. For example, for the EU’s patent legislation to be effective, there need to be rules on patent translation and the creation of a patent court. But the patent translation rules were adopted pursuant to a different decision-making rule, and the patent court will be established pursuant to a treaty between Member States. The legal base of the treaty with Mauritius should depend only on the content of the specific rules in the treaty with Mauritius. 

Here, the arguments are finely balanced. The Advocate-General makes a persuasive case that EU military operations can use civilian assets, and that the EU’s justice and home affairs powers can be used externally only where there is a sufficient link to the EU’s internal rules in this area. Incidentally, this line of argument strangles at birth the idea (floated, as it were, by Italy) that an EU foreign policy measure could establish a military action in the Mediterranean to control immigration towards the EU. The link between such an action and the EU’s immigration, asylum and border control powers is blindingly obvious.

On the other hand, with great respect, the Advocate-General’s analysis of the EU’s general external relations objectives is not fully convincing. True, the first and third of the four objectives he refers to (safeguarding values, et al, and preserving peace, et al) previously appeared in Article 11 TEU, which set out the CFSP’s objectives prior to the Lisbon Treaty. But the reference to the promotion of an international system based on cooperation and good governance is new, as is the general reference to the principles of international law; and these objectives are obviously applicable to any form of external action by the EU, whether it concerns the CFSP, judicial cooperation, development or anything else. As for human rights, democracy and the rule of law, they were indeed previously referred to in Article 11 TEU. But they were (and are) also a foundational value of the entire EU legal order (see now Article 2 TEU, and previously Article 6(1)), and are in part specifically referred to in the Treaty rules governing justice and home affairs and development cooperation. So this line of argument is ultimately not very persuasive.

Nor is the Advocate-General’s reference to the urgency of measures on this issue. The complications which the EU institutions face in their external action which result from the external relations rules in the Treaties simply can’t change the analysis regarding the legal base of those treaties (see Opinion 1/94, as regards the WTO). Anyway, the Council has the power, according to Article 218 TFEU, to decide to apply a treaty provisionally as soon as it signs it on behalf of the EU. Therefore the involvement of the EP in concluding the treaty would not compromise the urgency of achieving the treaty’s ends in any event.

So which legal base should apply? In my view, this should be determined on the basis of a teleological argument which the opinion does not consider. In the SEGI case, decided in 2007 (Case C-354/04), the Court of Justice ruled that an EU measure which imposed sanctions upon individuals could not be adopted in the form of a Common Position, a third pillar legal act which resembled a CFSP measure with a similar name. This approach ensured a minimum degree of democratic participation and judicial review at EU level of acts which directly imposed sanctions upon individuals. The same logic should apply by analogy here.

In fact, the Court should go further still. The elevation of the EU’s Charter of Rights to the ‘same legal value’ as the Treaties suggests that there should be a new approach to the resolution of legal basis and other institutional conflicts. Where relevant, if there is any ambiguity about the choice between possible legal bases or decision-making processes, the Court should ensure that EU measures concerning human rights should be decided by means of whichever process ensures the maximum possible parliamentary input and judicial control. It has already followed this approach in a case involving the powers of Frontex and national authorities to intercept vessels (C-355/10 EP v Council), and should confirm it as a fully-fledged new norm of interpretation. It is hardly necessary to point out how that rule should be applied in this case, as regards a treaty designed to ensure that criminal suspects who are in the hands of the EU receive a fair trial and basic human rights protection against torture and the death penalty when handed over to a third country.

Failure to inform the European Parliament

The Advocate-General’s conclusion on the first issue is reasonably argued, but his opinion on the second issue, with great respect, is deeply objectionable. The starting point as regards this issue is Article 218 of the TFEU, which specifies that the EP ‘shall be immediately and fully informed at all stages of the procedure’ relating to the EU’s international treaties. Remember those words in italics! The Advocate-General, sadly, did not.

First of all, the Advocate-General rejects the Council’s argument that the CJEU has no jurisdiction to examine the application of this rule as regards CFSP treaties. In his view, even where the substance of a treaty concerns the CFSP, the CJEU can examine the validity of the procedure used to adopt it, despite the Court’s lack of jurisdiction over the treaty as such (besides legal base arguments).

This is a fine line, but his argument has merit. If the CJEU had no jurisdiction, the EP’s procedural rights would be unenforceable as regards CFSP treaties – and those are the only rights it has as regards such treaties. And the CJEU can rule on those procedural rights without entering into any interpretation of the substance of those treaties, thus respecting the jurisdictional limit set by the Treaty drafters. It should follow by analogy that the CJEU would have jurisdiction under another provision of Article 218 to rule in advance on the compatibility of planned CFSP treaties with EU law.

Next, the Advocate-General argues that while the obligation to inform the EP applies to CFSP treaties, the EP should get more information, more quickly, where a treaty does not concern the CFSP, in light of the EP’s greater role regarding the conclusion of such treaties. Conversely it can receive less information, more slowly, as regards CFSP treaties.

How much was the EP informed as regards this treaty? When the Council decided to open negotiations, it informed the EP on the same day. That was certainly immediate. But the next time the EP heard from the Council was three months after the treaty was signed. This was one month after the publication of the decision to sign it in the Official Journal!

Shockingly, for the Advocate-General, this is sufficiently ‘immediate’. One is tempted to ask how many months he would be willing to wait to get served in a restaurant, or to use a toilet. With great respect, this is not, using any conceivable canon of interpretation, a plausible interpretation of that word. Put simply, no-one would consider a person who kept us waiting three months to be acting ‘immediately’.

As for the full information of the EP, the Opinion argues that because this was a CFSP treaty, the EP did not have to be informed of the progress of negotiations.

Let’s go back to the wording of the rule. First, a textual interpretation. Unlike the rules regarding the negotiation and conclusion of treaties by the EP, it makes no distinction between CFSP and other treaties. So prima facie, the two types of treaties must be placed on the same footing as regards information for the EP. 

Secondly, a contextual interpretation. The Advocate-General’s view of this rule is that it is designed to supplement the EP’s subsequent role as regards concluding the treaty concerned. But the Treaty makes no such link expressly. So the difference in wording between this rule and the rules on the EP’s role in concluding treaties suggests that it has a different purpose: to facilitate democratic debate as regards any planned treaty.

Of course, the amount of information which can be disclosed in public concerning a planned treaty in the midst of negotiations might be limited by confidentiality concerns, but these are addressed by agreements between the institutions concerned. Disclosing information to the EP as regards a planned treaty allows the EP to express its opinion, either privately to the Council or following a public debate, about whether a particular planned treaty is a good idea in principle. For instance, it should surely be a matter for public debate whether a particular country which the EU plans to transfer pirates to has a good record as regards fair trials, treatment of prisoners and the use of the death penalty. Of course, the EP’s influence may be limited as regards draft CFSP treaties because it will not get to vote on them. But why add insult to injury, and prevent it from holding an informed debate and expressing an informed opinion until (three months) after the treaty has been signed?

If anything, the context of CFSP treaties suggests that the EP should have more information, not less, than as regards other treaties. After all, the EP has a formal role as regards the conclusion of other treaties, usually the power of consent. So if the EP only finds out at a late stage that a draft treaty contains something which it finds objectionable, it can veto that treaty. Whereas, as regards a CFSP treaty, the EP’s only chance to influence its content will be before its signature and conclusion.

Finally, what does it mean to require ‘full’ information ‘at all stages’ of negotiations? This obviously applies to decisions (including agreements in principle) to open negotiations, initial treaties, sign treaties, provisionally apply them, and conclude them. Given the wording and purpose of the Treaty rule, it also should apply to proposals to negotiate treaties, the progress of treaty negotiations, the denunciation of treaties and to EU actions within bodies established by treaties.

[Update: the CJEU gave its ruling in June 2014. See discussion here.]

Barnard & Peers: chapter 24

Tuesday 28 January 2014

Want to be an EU citizen? Show me the money!

Steve Peers

Give me your rich, your bored,
Your pampered mistresses shopping duty-free,
The randy playboys through our open doors.
Send these, the chinless, party-goers to me,
Come flash your cash inside our finest stores!

We can only imagine whether Emma Lazarus, the author of the famous poem inscribed on the Statute of Liberty, would indeed have adapted her poem thus, in light of the sale of Maltese nationality (and hence, citizenship of the European Union) circa 2014. However, putting parody aside, are there any constraints deriving from EU law on Member States’ rules on the acquisition of EU citizenship, and if so, what are they?

The starting point is the judgment in Micheletti, in which the Court of Justice of the European Union (CJEU) stated that ‘'[u]nder international law, it is for each Member State, having due regard to Community law, to lay down the conditions for the acquisition and loss of nationality’ (emphasis added).

What limits are set by the phrase ‘having due regard to Community law’? In Kaur, the CJEU ruled that the United Kingdom was free to refuse to grant full British nationality to one of its quasi-citizens holding a special form of ex-colonial legal status, mainly because these rules were among the ‘conditions of accession’ of the UK to the Communities in 1973. ‘Furthermore’, the UK’s declaration to this effect did not deprive anyone ‘of rights to which that person might be entitled under Community law. The consequence was rather that such rights never arose in the first place for such a person’.

However, in Rottmann, the Court ruled that there were some constraints upon Member States’ rules upon the loss of their nationality, deriving from the existence of EU citizenship. The Court began by quoting the Council of Europe’s Convention on Nationality, Article 3:

‘1. Each State shall determine under its own law who are its nationals. 2. This law shall be accepted by other States in so far as it is consistent with applicable international conventions, customary international law and the principles of law generally recognised with regard to nationality.’

So international law (although Malta has signed, but not ratified, this Convention) gives States the power to determine who their citizens are, and requires other States to accept that decision in principle, with caveats (as regards the latter rule) about international law. If we could end our analysis here, the answer to the question posed above would be simple: Malta could sell its nationality.

But the answer is not simple, because, as a matter of EU law, the rules of public international law on nationality are not fully applicable. First of all, in Micheletti and subsequently (see Chen and Zhu), the CJEU stated that EU law establishes a rule of pure mutual recognition of Member States’ nationalities. In other words, the international law exceptions set out in Article 3(2) of the Council of Europe Convention do not apply. It must follow that, since Member States are obliged to recognise each other’s nationalities, they have at least some interest in what other Member States’ rules on nationality are – limiting the application of Article 3(1) of that Convention.

Let’s return to Rottmann. In its judgment, the CJEU then mentioned declarations and decisions of Member States to the effect that the acquisition and loss of nationality was purely a matter for national law. But the Court then stated baldly that: ‘Nevertheless, the fact that a matter falls within the competence of the Member States does not alter the fact that, in situations covered by European Union law, the national rules concerned must have due regard to the latter’. EU law applied to a national decision on the loss of Member State nationality, because it caused the person concerned to lose his status as an EU citizen.

The Court then explained the application of EU law to Member States’ nationality law as follows: 'The proviso that due regard must be had to European Union law does not compromise the principle of international law previously recognised by the Court…that the Member States have the power to lay down the conditions for the acquisition and loss of nationality, but rather enshrines the principle that, in respect of citizens of the Union, the exercise of that power, in so far as it affects the rights conferred and protected by the legal order of the Union, as is in particular the case of a decision withdrawing naturalisation such as that at issue in the main proceedings, is amenable to judicial review carried out in the light of European Union law (emphasis added).'

The Court then distinguished Kaur, as it concerned a person who had never had EU citizenship, rather than someone (like Mr. Rottmann) who had it and lost it. It then indicated that EU law could accept the idea that nationality of a Member State could be lost due to deception, subject to the principle of proportionality. Finally, it hurled this hot potato back to the national courts, which were left to apply these principles to the facts of this case. 

What can we derive from Rottmann? While the case concerned the loss of EU citizenship, we cannot conclude necessarily that EU law has nothing to do with its acquisition. The crucial words ‘in particular’ indicate that EU law might impact upon that topic too. While Kaur did concern the acquisition of nationality, it can be distinguished from the Maltese case because it concerned the refusal of nationality, not the grant of it. While refusal of nationality does not impact directly upon other Member States, the grant of that nationality (and therefore EU citizenship) does.

Having said that, Rottmann does indicate that prima facie EU law does not set very tight constraints on Member States’ nationality laws. But the outright sale of citizenship arguably bursts through even this loose corset. While the Court of Justice might be willing to accept extensive variants of the jus soli and jus sanguinis principles fro acquiring the nationality of a Member State, would it be willing to accept the principle of jus argentum? We cannot be certain until the Court of Justice is seized of the question; but we can hardly be sure that it would.

 *For an alternative view, see Dimitry Kochenov, ‘Citizenship for Real: Its Hypocrisy, Its Randomness, Its Price’, at

Barnard & Peers: chapter 13

Monday 27 January 2014

Holocaust denial and hate crime: Can the EU and its Member States do more?

Steve Peers

The European Commission has chosen today, the International Holocaust Remembrance Day, to release its first report on Member States' implementation of the EU Framework Decision on combating racism and xenophobia by means of criminal law.

Implementing the EU legislation

Member States had to implement this law, a relic of the pre-Lisbon 'third pillar' of EU law (setting out special rules on policing and criminal law measures) by December 2010, and the Council had to assess the Member States' implementation of the rules, on the basis of the Commission report, by November 2013. So the Commission report is a little late, but most such reports arrive even later (due to Member States' tardy reporting on their implementation). All Member States must apply this law, although the UK is planning to opt out as of December 2014, and does not intend to apply to opt back in.

The first crucial legal question is whether Member States have properly implemented their obligations under the Framework Decision. They are obliged to criminalise: public incitement to violence or hatred based on race, et al; dissemination of tracts to the same end; the denial, trivilisation et al of war crimes, genocide and crimes against humanity as defined in the Rome Statute establishing the International Criminal Court; and denial or trivialisation of the Holocauat.

The Commission identifies some apparent breaches of the Framework Decision here; for instance, two Member States insist on some further conditions being met before criminal liability attaches to the first category of actions. Some Member States do not specify that the crimes can apply to an individual or to a group. Two Member States refer to 'nationality' instead of 'national origin'. As regards the Rome Statute crimes and Holocaust denial, some Member States' laws do not refer to all types of actions referred to as regards the first type of crime, or do not fully reflect the obligations as regards Holocaust denial. In particular, two Member States only criminalise Holocaust denial in relation to their own nation or citizens. Some Member States have no specific provisions on these issues, although arguably a general law on incitement to violence that fully covers all of the relevant actions would be sufficient.

According to the EU law, Member States must either provide for racist and xenophobic motives to be considered an aggravated circumstance, or provide that courts may take those motives into account. Some Member States restrict this obligation to certain violent crimes only. Furthermore, some Member States attach conditions to the rules regarding liability for legal persons, or do not fully apply the rules on jurisdiction, in particular as regards offences committed over the Internet.

As to the future, the Commission makes a number of recommendations to Member States, as regards (for instance) special hate crimes units, the exchange of information, cross-border cooperation, data collection, the rights of victims and comments by opinion leaders. The Commission intends to discuss the correct implementation of the Framework Decision with Member States up to 1 December 2014 - the date when it can begin infringement proceedings as regards pre-Lisbon third pillar legislation. It does not make any mention of any amendment of the legislation, or of the specific issues which the Council is required to review (the issue of judicial cooperation as regards the relevant crimes).


The Commission cannot be criticised for holding off on bringing infringement proceedings, since it cannot do so until the end of this year. After that point, this legislation will be another EU measure which the Commission ought to enforce vigorously by means of infringement proceedings if it is, as it claims, committed to ensuring the full implementation of the EU Charter of Fundamental Rights in practice.

It would be possible to clarify the interpretation of the Framework Decision if it were amended, and more importantly, its provisions could be improved. For instance, the recommendations which the Commission makes to Member States in its report could be incorporated into the legislation (except for the point concerning the rights of victims, which will become binding anyway once the EU's crime victims directive is implemented in 2015). The Commission does not consider the issue of possible amendments at all.

More significantly, the scope of the Framework Decision (or rather, the future Directive) could be enlarged, to cover other forms of hate crime. The Commission perhaps avoids mentioning the issue of amendment because of the lack of a specific legal base dealing with this issue in the current Treaties. True, racism and xenophobia are not listed among the crimes the EU can combat in Article 83(1) TFEU. But they surely fall within the scope of Article 83(2) TFEU, which gives the EU power to adopt criminal law measures when necessary in relation to a matter which the EU has harmonised. Since the EU has banned discrimination on grounds of race as regards all goods and services (inter alia), and it surely would interfere with equal access to transport, shopping and recreation (for instance) if crimes of racial hatred were committed, it could be argued that further EU measures could be adopted on this basis. The same would apply to bias crimes against women, given the scope of EU harmonisation already on the issue of gender equality.

However, it would be harder to argue that a legal power exists for the adoption of EU measures banning hate crime on grounds of sexual orientation, disability, religion or age, since the EU has only harmonised the law as regards equality in employment as regards those issues. When or if the Commission's proposed Directive extending equal treatment as regards these four grounds of discrimination is adopted, then a further measure relating to hate crime on the same grounds could be proposed.

Barnard & Peers: chapter 25, chapter 20

The EU’s Financial Supervisory Authorities: Mind the Accountability Gap

Dr Marios Costa, Lecturer in Law, City Law School

In 2010 we witnessed the establishment of three European Supervisory Authorities: the European Banking Authority; the European Insurance and Occupational Pensions Authority; and the European Securities and Markets Authority (ESMA). They were set up by the Union as a response to the current, unprecedented financial crisis. The Court of Justice of the European Union (CJEU) gave on 22 January 2014 a significant judgment in relation to more recent legislation empowering ESMA to adopt legally binding measures upon financial institutions of the Member States in the event of a threat to the proper functioning of the financial market or to the stability of the financial system of the EU (Case C-270/12, United Kingdom v Council & Parliament). The legal action concerned the annulment of Article 28 of Regulation 236/2012 in relation to ESMA’s power to ban ‘short selling’, a practice which permits the sale of shares not owned by the vendor at the time of sale with the view of benefiting from a fall in the share price.

There are broader constitutional implications which this judgment highlights. The judgment, which does not come as a surprise, clarifies issues in relation to the powers that can be lawfully exercised by EU independent financial regulatory agencies. This commentary examines, with all due respect, whether the recent ruling will remedy the lack of accountability of EU agencies.

ESMA can draft highly detailed technical and implementing standards which are later on adopted by the Commission under Article 290 and 291 TFEU (which concern, respectively, the adoption of delegated and implementing acts). In relation to Article 290 TFEU, the Commission sets out the conditions and specifies the criteria under which the agency can adopt further regulatory measures of a technical nature, but the drafting of the technical measure always comes from the agency. A very important issue here is whether the Commission has the sources, technical knowledge and scientific expertise required to control the appropriateness of the measures drafted by the agency. If the Commission decides not to adopt the measures drafted by ESMA then it is required to send it back to the Agency and explain why it has decided to not to endorse it (see Article 10 and 15 of Regulation 1095/2010). Interestingly enough, there are extreme limitations imposed upon the Commission. According to the preamble of Regulation 1095/2010, the Commission can only depart form the draft measures prepared by the agency only if they are incompatible with EU law, violate the principle of proportionality or contradict the EU’s financial services legislation.

Facts of the case

The UK government challenged the legality of article 28 of Regulation 236/2012 on the power of the ESMA to ban short selling practices. The Regulation was adopted on the basis of Article 114 TFEU which allows for the enactment of harmonisation measures necessary for the establishment and the functioning of the internal market. The rationale for the adoption of the Regulation and in particular Article 28 is for the ESMA to interfere and issue legally binding measures against the financial institutions of the Member States to prohibit short selling in the event of a threat to the proper functioning and integrity of financial markets or to the stability of the whole or part of the EU’s financial system. The ESMA has wide discretionary power to issue such bans, and it is the only adjudicator of whether such a threat exists.

The UK raised four arguments. First of all, it argued that ESMA is given political powers which entail policy choices to adopt legally binding measures vis-a-vis the financial institutions of the Member States. These powers do not fit well with the old Meroni line of case law, which states that delegation to autonomous bodies is considered to be acceptable as long as Commission retains control powers to monitor how the agency is carrying out its tasks. According to the Meroni line of reasoning, the conferment of broad discretionary power, reconciling competing public interests, to an EU agency cannot be justified on the basis of scientific expertise. In any case, the CJEU has several times emphasised that ‘[s]cientific legitimacy is not a sufficient basis for the exercise of public authority’ (Pfizer). However, in this judgment the Court of Justice ruled that the parent EU legislation, and the delegated and implementing acts adopted pursuant to that legislation by the Commission, sufficiently circumscribed ESMA’s powers.

Secondly, the UK argued that the power for ESMA to ban short-selling breached the principle in Romano that the EU legislature could not delegate the power to adopt ‘quasi-legislative measures of general application’. However, the Court ruled that Romano did not add anything to Meroni, noting in particular that the Treaty provides for agencies to adopt measures of general application.

Thirdly, the UK argued that Articles 290 and 291 TFEU (the provisions on the adoption of delegated and implementing acts) were in effect exclusive, ruling out a contrario the delegation of powers like the short-selling ban to EU agencies. In the Court’s view, the Treaty (in particular, the rules on judicial review) presupposed that agencies could adopt binding acts, and the provision allowing ESMA to ban short selling had to be seen in its overall legal context.

Finally, the UK argued that Article 114 TFEU cannot constitute a correct legal basis for the adoption of the rules laid down in Article 28 of the Regulation. Earlier in 2013 the Opinion of Advocate General Jääskinen concluded in favour of the annulment due to concerns in relation to the appropriateness of the legal basis of Article 114 TFEU. According to his view, the adoption of legally binding measures by the ESMA addressed to the financial institutions of the Member States cannot be considered as EU harmonising measures or uniform practices which could be justified under Article 114 TFEU. The Court, however, decided not to follow the non-binding view of the Advocate General and ruled that Article 114 TFEU constitutes an appropriate legal basis for the adoption of Article 28 of the Regulation since it aims (a) to approximate national law and (b) to improve the conditions for the establishment and functioning of the internal market in the financial field. On the first point, the Court brought together its prior case law which had specified that Article 114 could be a legal base for the creation of EU agencies (Case C-217/04 UK v Council and EP), and for the conferral of power upon the EU institutions to adopt legally binding acts (Case C-359/92 Germany v Council).


The Court’s judgment has significantly clarified the law relating to the conferral of powers to EU agencies. First of all, the Meroni doctrine, while still in force, does not prevent the conferral of such power when the relevant legislative framework is sufficiently detailed. Secondly, the Romano ruling adds nothing to Meroni. Thirdly, Articles 290 and 291 TFEU do not prevent the conferral of powers upon agencies, at least where such conferral of power takes place in the context of an overall legislative framework. Finally, at least the internal market powers of the EU (and arguably, by analogy, other legal bases) do not prevent the delegation of powers to agencies to adopt legally binding measures.

The Court’s ruling gives significant emphasis to the fact that the measures adopted by the EU financial agencies are subject to judicial review under Article 263 (4) TFEU. However, regulatory and implementing technical standards drafted by the EU financial agencies are subject to the Commission’s endorsement and although they constitute the basis for the adoption of the final act by the Commission they are technically and legally preparatory documents and as such they are excluded, in principle, from judicial scrutiny. Additionally, non-privileged applicants, such as financial institutions negatively affected by any ban adopted by ESMA, might not be in a position to satisfy the EU’s locus standi requirements. They may be excluded from direct actions under Article 263 (4) TFEU on the basis that the ban adopted still entails separate implementing measures within the meaning of the Telefonica judgment.

With great respect to the ESMA judgment, the fact that the founding Regulation of the ESMA [Regulation 1095/2010, Article 10(1) and 15 (1)] limits the power of the Commission to proceed with the drafting of technical standards or to unilaterally amend them empowers EU financial agencies with wide political decisions which entail policy choices. The wider implications of the judgment and also of the current framework establishing the agencies seem to suggest what has already pointed out in the literature that ‘in any event, it is clear that EU independent agencies are independent in the sense of being relatively free of control by any other organs of the [Union]’ (Shapiro, 1997). Another point is whether the Commission or the agency will be held responsible in cases where adverse consequences occur if the assessment by the agency proves to be wrong and has further financial repercussions? Moreover, according to the ESMA judgment additional powers could be conferred on agencies to adopt acts of general application outside the scope of articles 290 and 291 TFEU. The delegated and implementing acts of the Commission which detailed ESMA’s powers to adopt the short-selling bans were themselves drafted by…ESMA. This raises important questions in relation to the accountability of ESMA which the Court seemed, with all due respect, to ignore.


The Court’s ruling is important given that it is the first case which deals with the powers of the newly created financial supervisory authorities. It empowers, however, the financial authorities with further powers in order to be able to foresee and secure financial stability in the European Union. Therefore, there are certain constitutional questions that need to be answered: who are these highly independent autonomous bodies answerable to? Scientific legitimacy and complex decision-making in the area of EU’s financial regulation cannot be a legitimate justification for increasing the powers of the EU’s financial agencies, something which can only be accepted if there are control powers vested in the main EU institutions for securing the accountability of these EU agencies.

Barnard & Peers: chapter 8

Thursday 23 January 2014

Is readmission linked to development?

Steve Peers

In recent years, the EU has been negotiating Partnership and Cooperation Agreements (PCAs) with a number of Asian countries. These agreements replace the previous development policy agreements which the EU had with the countries concerned. Compared to the previous agreements, the PCAs include go into greater detail about the EU's cooperation with the countries concerned, including the addition of further topics for cooperation.

Given the addition of these new topics, and the expansion of detail on other topics, the issue arises whether (like the previous generation of treaties) the legal base for the treaties concerned should be trade and development alone, or whether other legal bases must be included. Unsurprisingly, the Commission takes the former view, and the Council takes the latter. In fact, the same issue has arisen in respect of some recent association agreements, although they are not at issue in this case and are subject to a different legal base.

Who is correct about the legal base argument, the Commission or the Council? In today's opinion by Advocate-General in Case C-377/12 Commission v Council, Advocate-General Mengozzi argues that the Commission is correct. This particular case concerns the Council's decision to sign the PCA with the Philippines on behalf of the EU, and the Commission is challenging the Council's decision to add legal bases relating to the environment, transport and immigration policy to the trade and development legal bases which the Commission had proposed.

The starting point of the A-G's analysis is the CJEU's prior judgment of 1996 in Portugal v Council, in which the Court ruled that the EU's development policy legal base could apply to the conclusion of a treaty with India which included provisions on issues like energy and culture, along with a clause on suspension of the agreement on human rights grounds. The core of the Court's reasoning was that a development policy treaty could include general provisions on issues like energy and culture, as long as cooperation on these issues contributed to the development of the countries concerned, but if there were any detailed commitments on such issues in the treaty, other legal bases would have to be added. Furthermore, it was acceptable for such treaties to include a clause allowing for their suspension on human rights grounds.

Applying that prior case law to the PCA with the Philippines, while the provisions in that treaty on transport and the environment are more detailed than those in prior development policy agreements, the A-G argues that such clauses still set out only general obligations, and still contribute to the development of that country. The Council also raised arguments about the impact of the clause allowing the PCA to be suspended in the event of concerns about human rights, democracy and the rule of law, but the A-G does not believe that should affect the legal base for signing the agreement.

The provision in the PCA on immigration raises more difficult issues. The A-G draws a distinction between the more general paragraphs on immigration management, and the more specific paragraphs dealing with readmission. The former clauses do not set out precise legal obligations, while the latter do: the parties have agreed to take back their own nationals who stay on the territory of the other parties without authorisation, and have agreed to negotiate a readmission treaty. In the A-G's view, these clauses don't contribute to the development of the Philippines either. In fact, they only serve the EU's interest.

But having said that, he still concludes that the Council should not have added a separate legal base for the readmission clauses, since (following other prior case-law) this issue is ancillary to the main thrust of the agreement. In particular, he reaches this conclusion because the EU only obtained the Philippines' consent to the readmission clauses in return for the Union's commitments as regards development in the rest of the PCA.


Why did the Commission bring this action? It is not contesting the nature of the PCA as a 'mixed agreement', ie concluded by the Member States as well as the EU. Also, the inclusion of the extra legal bases did not change the voting rules in the Council, as it will still vote by a qualified majority regardless. Anyway, as long as the Member States are party to the agreement, there is de facto unanimous voting, since they act by common accord.

The Commission may have an agenda to reduce the use of the transport legal bases in agreements which also relate to trade, given the 'transport services' exception in the Treaty rules relating to the common commercial (external trade) policy, which sullies the purity of a key EU external competence which would otherwise be purely exclusive. This makes sense, but the EU's free trade agreements usually include detailed commitments on trade in transport services, so the transport services exception is surely applicable. On this point, the better strategy for the Commission would be to propose internal EU legislation which more fully harmonises the regulation of transport services to and from third countries. If and when such legislation is adopted, the EU would gain exclusive external competence over this issue via the normal (ERTA) route. Of course, this is easier said than done.

Leaving aside the specific issue of transport services, perhaps the Commission wants to reduce the use of additional legal bases besides the common commercial policy in the EU's free trade agreements more generally (although the PCA with the Philippines is not itself a free trade agreement), given that such clauses are seemingly usually added in order to trigger mixed competence between the EU and the Member States (and therefore give Member States a veto). The most egregious example of this is the bizarre cultural protocol attached to the EU/Korea free trade agreement. If the Court follows the A-G's opinion in this case, the Council's ability to do this will be restrained a little, although the Member States can still get their way on this issue if they insist on attaching a protocol with sufficiently detailed obligations in the extraneous field.

This brings us to the key issue of external migration, in particular readmission. The effect of including legal bases on this issue is that the UK, Ireland and Denmark have an opt-out from legal rules concerned. In practice, the UK tends to opt in to EU readmission agreements anyway, so this is more a point of principle for that Member State.

Is the Opinion convincing? Certainly the PCA does not contain detailed rules on transport or the environment (compare it to the EU's aviation liberalisation treaties with third States, or international environmental treaties), and cooperation on such issues will surely aid the Philippines' development. Of course, increasing transport to and from the country and protecting the environment there may well be contradictory objectives, but the list of EU external objectives in Article 21 TEU is rife with such contradictions.

It could be added that including environmental provisions in the PCA is consistent with the obligation to integrate environmental protection into all EU polices. The A-G is also surely right to say that improving transport security and safety will aid the Philippines' development: no-one will want to fly there if their plane is liable to crash or be blown up.

As for the immigration clauses, the migration management rules are not only not very precise; it could be added that they are closely related to the development of the Philippines. However, the A-G's arguments relating to the readmission clauses are, with great respect, not convincing. It is true to say that they set out precise obligations and do not aid the Philippines' development, but then the A-G does not draw the obvious conclusion that a specific legal base concerning immigration is therefore necessary. While is certainly true that the readmission clauses and the development policy commitments are linked politically, it does not therefore follow that the legal base of the former clauses should change.

One final thought: the judgment in this case, whichever party is successful, could be an opportunity for the Commission to suggest a relaunch or review of the EU's policy on external migration management and/or readmission. For instance, why not adopt EU legislation setting out specific rules on coordination of Member States' external migration treaties? (Such legislation exists in a number of other EU policy fields). This can be justified as a necessary feature of the development of a common EU immigration policy, and would be an opportunity in particular to ensure that such treaties are consistent with the human rights obligations of the EU and its Member States.

Update: the CJEU gave its ruling in this case in September 2014. See comments on the ruling here.

Barnard & Peers: chapter 24, chapter 26

Monday 20 January 2014

Do potential asylum - seekers have the right to a Schengen visa?

By Steve Peers

In its recent judgment in Koushkaki, the Court of Justice clarified some key questions regarding the process of applying for a Schengen visa, in effect ruling that applicants have a right to a visa if they meet the necessary conditions. In doing so, it implicitly raised further questions about the grounds to refuse visa applications made by potential asylum-seekers.

The judgment

Mr. Koushkaki, an Iranian citizen living in Iran, applied to the German authorities for a Schengen visa to vist Germany. The German authorities refused, on the grounds that he did not have proof of subsistence for his stay or return. These are among the conditions for entry set out in the Schengen Borders Code, and also apply to the issue of Schengen visas under the EU visa code.

So Mr. Koushkaki applied again, this time explaining that he wished to visit his brother, who resided in Germany after obtaining asylum there. This second application was again rejected, this time on the grounds that Mr. Koushkaki had not shown an intention to return. Mr. Koushkaki challenged this decision, and the national court, believing that the authorities had refused Mr. Koushkaki on grounds not referred to in the visa code, asked the Court of Justice (inter alia) whether Mr. Koushkaki had the right to a visa if the conditions in the code were satisfied and none of the grounds for refusal listed in the code were applicable.

The Advocate-General took the firm view that there was no right to a Schengen visa, but the Court took a different approach. It reworded the national court's main question, answering instead the question whether the list of grounds for refusal of a visa set out in the Code was exhuastive. In the Court's view, the list was exhaustive, taking account of the wording, context and objectives of the Code.

First of all, the wording of the Code was ambiguous. Secondly, the context included: a reference to only those grounds for refusal listed in the code; a standard form for giving reasons for a rejection of a visa application which provided for no other reasons for refusal besides those listed in the Code; the wording of the legislation establishing the Visa Information System; and the link between the grounds for refusal of a visa application and the grounds for revocation or annulment of a visa which has already been issued. Since a Member State can annul or revoke a visa issued by another Member State, this presupposed that the 'conditions for the issue of uniform visas are harmonised', so there could be no differences between Member States as to the grounds to refuse a visa.

As for the objectives of the visa code, the preamble to the code states that it aims to 'establish the conditions for the issue of uniform visas'. Also, the preamble refers to 'the facilitation of legitimate travel', which would be jeopardised if Member States could add their own grounds for refusal of a visa application. So would the objective of 'preventing different treatment of visa applicants', also referred to in the preamble. Such variations in national practice would also encourage 'visa shopping', ie applications being to the Member State with the most favourable rules (leading to what a labour or company lawyer would call a 'race to the bottom').

Having said that, the CJEU did emphasise that Member States' authorities had a lot of discretion in applying the standard criteria. They have to consider the 'personality of the applicant', his or her integration in the country of origin, the 'political, social and economic situation of that country' and the possible threat to public policy, internal security, public health or international relations of the Member States. Some of these factors are not, in fact, expressly referred to in the visa code.

The Court of Justice then addressed the national court's other questions. It clarified that there need only be 'reasonable doubt', not certainty, that an applicant had the intention of leaving the country when the visa expired, and spelled out the criteria which had to be taken into account. Finally, it required the national court to interpret national law consistently with the judgment, even though that national law provided for residual discretion to refuse visa applications.


The Court's judgment is surely relevant by analogy to applications for most of the various forms of short-stay visa referred to in the visa code: airport transit visas, multiple-entry visas and visas with limited territorial validity. It is doubtful whether it applies to the issue of visas as the border, since the code says that such visas 'may' be issued if the relevant conditions are met. More obviously, the judgment is relevant by analogy to the Schengen Borders Code, given the similar wording and context of the rules on refusal of entry in the latter code, along with the use of a similar standard form to give reasons to refused applicants.

It is also arguable that the judgment applies by analogy to some or all of the EU's legal migration legislation: indeed the same German court has sent questions to the Court of Justice on exactly this point, as regards the students' Directive (Ben Alaya).

On the key question of the 'right' to a Schengen visa, the Court's ruling is welcome and its reasoning is convincing, in light of the structure and objectives of the visa code. It is notable that the Court sets out its ruling in technical language, referring to the exhaustive nature of the list of grounds for refusal, rather than a 'right to a visa'. But this is a distinction without a difference: national authorities must nonetheless issue the visa if the conditions are satisfied. As Mary Poppins might say, the different wording is simply a 'spoonful of sugar' to help the national authorities accept the Court's ruling.

Anyway, as the CJEU makes clear, those authorities have a lot of discretion left when they apply those criteria. As noted above, some of the factors referred to by the Court are not referred to in the Code. In particular, assessment of the applicant's 'personality' surely requires a psychiatrist, but it is doubtful that many consulates employ any of them.  

The position of potential asylum-seekers

According to the United Nations (Geneva) Convention on refugee status, a person cannot be considered a refugee until he or she is outside the country of origin. But of course a person fearing persecution on Convention grounds (or needing subsidiary protection) might well apply for a visa with the intention of leaving the country of origin in order to apply for asylum in the country which issues the visa.

On the facts of the Koushkaki case, it is strking that when Mr. Koushkaki mentioned that his brother had obtained asylum in Germany, the German authorities refused his application for a visa on the grounds that he had not proved his intention to return. Since Mr. Koushkaki might be assumed to have had life experiences comparable to his brother's, the national authorities might well have feared that he intended to apply for asylum in Germany. (Incidentally, the Dublin rules would have allocated Germany responsibility for his claim, if the German authorities had given him a visa).

Can potential asylum-seekers rely on this ruling to insist on the right to a Schengen visa? The key problem is that they do not have an intention to leave the country which might issue the visa to them. They would need to offer enough evidence that they do have such an intention to dispel any reasonable doubts of the national authorities. While such evidence might not be offered honestly, Article 31 of the Geneva Convention implicitly makes clear that the need to flee persecution justifies breaches of immigration law.

While Article 21(1) of the visa code refers to a need to decide whether an applicant presents a 'risk of illegal immigration', this clause should not be used against potential asylum-seekers, since they have the right to stay in the territory (subject to the conditions in the asylum procedures Directive), and the case-law of the CJEU on the Returns Directive confirms that they cannot be considered to be irregular migrants.

Potential asylum-seekers would also have to meet the other relevant conditions, such as holding a valid travel document, having subsistence, and not being listed on the Schengen Information System. For many, these conditions will be hard to fulfil.

However, as suggested above, following Koushkaki it could be argued that Member States are obliged to issue a visa with limited territorial validity (ie valid in one Member State only: this is good enough for an asylum-seeker to flee persecution, provided that the Member State concerned complies with its obligations towards asylum-seekers).  Such visas 'shall' be issued if the Member State 'considers it necessary on humanitarian grounds, for reasons of national interest or because of international obligations', to derogate from the rules in the Schengen Borders Code (among other things). Arguably the binding nature of the relevant international obligations, along with the EU Charter of Fundamental Rights and the use of the word 'shall', override the discretion suggested by the words 'consider it necessary'.

If this argument is correct, then the Koushkaki judgment has opened a significant crack in the wall of 'Fortress Europe' for would-be asylum-seekers.

Barnard & Peers: chapter 26

Sunday 19 January 2014

When is the family member of an EU citizen ‘dependent' on that citizen?

By Chiara Berneri, Lecturer at BPP Law School

On the 16th of January 2014 the Court of Justice of the European Union (CJEU) further clarified the circumstances under which a family member of an EU citizen who moves to another Member State can be considered a dependant within the meaning of the Directive on the free movement of EU citizens (Directive 2004/38, or the ‘citizens’ Directive’).

The meaning of the word ‘dependant’ in the citizens’ Directive is important, because that Directive allows EU citizens who move to another Member State to bring with them or to be joined by their spouse, registered partner (under certain conditions), descending direct relatives of the citizen or the citizen’s spouse/partner who are under 21 or ‘dependants’, and the ascending directive relatives of the citizen or spouse/partner who are ‘dependants’. The Directive defines these people as the EU citizen’s ‘family members’.

In Flora May Reyes v. Migrationsverket the applicant, Ms Reyes, a Philippines citizen who was born and had lived in the Philippines for all her life, applied for a residence permit in Sweden as a dependant family member of her mother, a dual citizen of the Philippines and Germany, and her mother’s Norwegian cohabiting partner (who subsequently married her mother). Ms. Reyes was then 23 years old and had qualified as a nursing assistant in the Philippines, but had not been able to find work there.

Her application was refused by the Migrationsverket (the Swedish immigration authority) since Ms Reyes could not prove that the money that she was constantly receiving from her mother and her mother’s partner when she was still in the Philippines was used to supply her basic needs in terms of board and lodging and access to healthcare (she had always been dependent on her grandmother during her childhood and adolescence). Moreover, she could not show how her country’s social insurance and security system could have provided for a person in her conditions.

On appeal against this decision, the Gothenburg Administrative Court stated that, although it was evident that the basic needs of the applicant were met by her mother and step-father, her situation could not be regarded being such as she could not support herself in her country without their help. When the case reached the Stockholm Immigration Court of Appeal the latter decided to refer to the CJEU questions to clarify the definition of a ‘dependant’ under the citizens’ Directive. and in particular (a) whether the person concerned had to show that he or she had applicant has to tried to obtain employment or support in the home country but that it was not possible, and (b) whether a person who had a good chance of finding a job in the host country could still be considered a ‘dependant’ of the EU citizen who was living there.

 The judgment

On the first question the Court began by pointing out that, in order to determine whether a family member was ‘dependant’, the host Member State had to assess whether the applicant was not in a position to support herself either in the state of origin or in the state in which she is applying for residency. Once this is established, there is no need to determine the reasons for that dependence.

This is because, reiterating the prior case-law of the Court, free movement rules, like those included in Directive 2004/38, must be interpreted broadly. For these reasons, the fact that a Union citizen pays regularly a sum of money to the applicant is enough to show that the latter is in a situation of dependence and, therefore, the descendant does not have to show that she tried unsuccessfully to find work or support from the authorities in the country of origin.

On this last point, the Court concluded that the potential requirement of proving the attempts to find a job or social assistance would make it excessively difficult for the descendent to obtain the right of residence. On the second question, the Court began by stating that dependence must exist in the country from which the family member comes. For this reason, the prospect of finding a job in the host Member State does not have to affect the interpretation of the condition of dependency. It concluded that the opposite solution would imply that the applicant, in order to be able to enjoy residency in the host Member State, should not apply for any job. However, this would expressly contradict Article 23 of the citizens’ Directive 2004/38, which authorises the family members of EU citizens to take up employment or self-employment in the host Member State.


The judgment of the Court sheds some light on the requisites necessary in order to be considered a dependant. Owing to this judgment an applicant simply has to prove that he or she has been receiving financial support from the EU citizen family member that he or she wishes to join. Simply showing the regularity of the remittances seems to be enough to fulfil the requisite of dependency (see the Advocate-General’s Opinion, para. 59).

It should be recalled that in practice, remittances to family members in developing countries from those family members who have established themselves in developed countries (like EU Member States) is a central facet of immigration. Furthermore, the fact that the applicant, for age, personal conditions or any other reasons, will be likely able to find a job in the host Member State does not count against fulfilling the dependency requirements.

This judgment finds its grounds in the earlier case-law of the Court, namely the Lebon and Jia cases. Lebon introduced the principle that there is no need to investigate the reasons underpinning the situation of dependency. Jia underlined that the situation of dependency is established by referring to the necessity of material support in the State of origin of the applicant.

The application of these two principles to the circumstances in Reyes further clarified the definition of dependence. The Court expressly excluded any requirement to prove that the family member had to have unsuccessfully tried to find a job or to have tried to apply for social assistance. Moreover, since the necessity of material support has to be established in the state of origin, it means that the applicant can become economically independent in the host Member State. This is particularly relevant to people like Ms. Reyes – young people, at the start of their working life, who have undertaken further or higher education (and so are now over 21) but who have not obtained a job in their country of origin.

The indications of the CJEU seem to be quite sharp and clear. Therefore, it is likely that national courts will easily apply them.

The same interpretation of ‘dependants’ should logically be applied to the category of ascending direct relatives, because they are equally included in the definition of the core 'family members' in the citizens' Directive. A different treatment of this group of family members would amount to unjustified discrimination.

A different consideration should apply for dependants referred to in Art. 3(2)(a) of the citizens’ Directive, namely those relatives of the EU citizen who are not included in the definition of ‘family member’ in the Directive, but whose right of entry ‘shall’ be be facilitated by the host Member State (on this category of persons, see further the Court’s judgment in Rahman). From the Advocate-General’s opinion in this case it is possible to understand why national courts cannot be forced to apply the same interpretation of ‘dependants’ also to this group of family members. As he clearly pointed out, the provision concerning these family members in the citizens' Directive does not grant them the right to reside with their EU sponsor but simply obliges the host Member State to 'facilitate' their entry and residence. This means that, potentially, Member States are entitled to undertake an extensive examination of the personal life of the applicant. Perhaps this would mean that the relatives belonging to this category could be asked to prove that they have tried to find a job or seek social assistance in their home state.


With Reyes, the Court has begun to give more concrete guidance to national courts on how to apply the concept of dependence. This improvement was long awaited due to the lack of indications in any sense from the legislation. Given the new development introduced by this case it is likely that in the future other challenging questions on this issue will arise.

Barnard & Peers: chapter 13

Thursday 16 January 2014

Enforcing Previous Judgments of the Court of Justice

By Steve Peers

The Court of Justice of the EU (CJEU) has recently delivered an interesting judgment concerning the so-called 'infringement procedure', which is the system for enforcing EU law directly in the EU courts. This system runs in parallel to the process of enforcing EU law via the national courts (see the previous post). The details are set out in Article 258 of the Treaty on the Functioning of the European Union (TFEU), which provides that the Commission can bring actions to the Court of Justice against Member States which have allegedly breached EU law.

But what happens if Member States do not comply with those rulings? There might be follow-up in the national courts. Also, there is a process, set out in Article 260 TFEU, for the Commission to bring a further action to the Court of Justice against a Member State which has not implemented an infringement ruling against it. If the Court of Justice agrees with the Commission, it may impose a lump sum and/or a penalty payment against that Member State. The penalty payment is charged at regular intervals, for as long as the Member State has not complied with the original judgment. For more on this, see my 2012 article in European Public Law.

Once the Member State has indisputably complied with the original judgment, or is indisputably not complying with it, its liability to pay the penalty payment or not will be perfectly clear. (There might be issues if the Member State refuses to pay the sums due, or if the Commission waives that Member State's obligation to pay, but let's leave them aside). But what happens if there is a dispute about compliance with the original judgment? The Member State in question is arguing that it has now fully complied with the judgment, but the Commission is arguing that it has not.

This sort of problem is not unique to EU law: some may remember one twist in the 'banana saga', when the EU claimed that it had complied with rulings of the World Trade Organisation (WTO) against it, while the USA claimed that it had not. In the absence of clear rules in the WTO's Dispute Settlement Understanding on this point, the WTO bodies had to craft an ad hoc solution.

Equally there are no clear rules governing this scenario in EU law, so the EU General Court had to craft them back in 2011, when Portugal argued that the Commission had wrongly sent it a demand for a penalty payment, whereas it had already complied with the original judgment against it, which concerned its breach of an EU Directive on public procurement (Case T-33/09 Portugal v Commission). The General Court agreed with Portugal, ruling that in this scenario the Commission could not simply demand more penalty payments, but had to begin a wholly new infringement procedure from scratch.

The Commission appealed this judgment to the Court of Justice (Case C-292/11 P Commission v Portugal), but in its judgment that Court upheld the General Court's judgment. In the Court's view, the obligation for Portugal arising from its case-law was to repeal a particular law, and Portugal had done so. There was a replacement national law, which the Commission claimed still violated EU obligations, but the Commission had to bring new infringement proceedings from scratch to establish that this was a breach of EU law.

The first reason for the Court's ruling was that it had exclusive jurisdiction to rule on whether a Member State had breached EU law. Neither the Commission nor the General Court could encroach on this jurisdiction. It would also be problematic because the Court of Justice would be limited to reviewing the legal conclusions of the General Court, being unable to examine its findings of fact.

Also, giving the Commission the power to decide whether Member States had fully complied with Article 258 judgments or not would impinge upon the procedural rights which Article 258 sets out for Member States, namely the possibility to dispute the Commission's allegations at pre-litigation stage.


The case-law of the CJEU is often criticised, sometimes justifiably, for prioritising the effectiveness of EU law above all else. But with great respect, this judgment can be criticised for exactly the opposite reasons. It simply makes it too easy for a Member State to replace a national law which breaches EU law with a new law which breaches EU law nearly as much as the old one (I am assuming that the judgment would not apply if the new law had exactly the same content as the old one, for that would be truly absurd).

If a Member State does this, the Commission can no longer collect penalty payments, unless it can establish the breach all over again by bringing proceedings again under Article 258 and then Article 260. And when this second round of proceedings concludes years later, what is to stop the Member State making another small change to its national law, compelling the Commission to start Article 258 proceedings a third time if it wishes to (re-)establish the continuing breach?

The Court never examines the issue of effectiveness, although the Commission had raised it. As for the reasons which the Court does give, it can't really be said that the Commission or General Court would be encroaching upon the Court of Justice's exclusive jurisdiction as regards infringement proceedings, since such a case would only arise in the first place after a first round of judgments under Articles 258 and 260.

Nor is it convincing to argue that Member States' procedural rights would be impinged: they could exercise those rights in full when the original Articles 258 and 260 proceedings were brought, and they would have the right to challenge a Commission decision declaring that they were still in breach of the original judgment of the Court of Justice.

Finally, as for the role of the Court of Justice on appeal, it would have been just as plausible to rule that the question of whether a Member State had complied with a judgment of the Court of Justice was an issue of law, not fact, and so can be fully reviewed by the Court of Justice on appeal.

Much of the Court's concern is jurisdictional, and so would be addressed if jurisdiction over infringement proceedings were transferred to the General Court. There would be no need to amend the Treaties in order to do this; an amendment to the Statute of the Court of Justice, which can be adopted by means of the ordinary legislative procedure, would suffice.

Of course, such a change in the EU's judicial architecture is undesirable as long as the General Court is overworked. Yet its workload could be alleviated if more judges were appointed to it. The Court of Justice has suggested this, but its proposal is blocked in the Council, due to squabbles about which Member States would get the extra judges. So perhaps there is a 'hidden agenda' to this judgment?

Barnard & Peers: chapter 10

Wednesday 15 January 2014

When does the EU Charter of Rights apply to private parties?

Professor Steve Peers, University of Essex

The Court of Justice of the European Union (CJEU) has today further clarified the circumstances in which the EU's Charter of Rights can be invoked against private parties. In Association de Mediation Sociale (AMS), trade unions challenged a private employer's refusal to establish worker consultation pursuant to an EU Directive. The employer had acted consistently with French law, which had implemented the EU Directive but allowed exclusions for apprentices and other forms of special employment contract.

The judgment

First of all, the CJEU ruled that the French law breaches the Directive, which does not provide for the exclusion of such categories of employee from its scope. But how can this breach of EU law be remedied, on the facts of this case? The Court states that while the relevant rule in the Directive is precise enough to have direct effect, Directives (following well-known and long-established case law) cannot be invoked against private parties.

Also, the CJEU confirms prior case law which states that the principle of indirect effect (as it is usually called), ie the requirement for national courts to interpret national law to be consistent with EU Directives, meets its limits where (as in this case) there is an unambiguous contradiction between the national law and the relevant EU Directive.

Towards the end of the judgment, the Court mentions the long-established prospect of suing a Member State for damages for its failure to implement a Directive correctly. However, on the facts of this case, it's hard to see how that remedy could work well. Perhaps the union or an employee could argue that if worker consultation had been established, it could have persuaded the employer not to make redundancies (for instance). But really what the unions seek to achieve here is to have worker consultation established in the first place.

So this brings us to the heart of the matter: does the EU Charter of Fundamental Rights affect the issue? Article 27 of the Charter states that:

'Workers or their representatives must, at the appropriate levels, be guaranteed information and consultation in good time in the cases and under the conditions provided for by Union law and national laws and practices.’

Previously, in the judgment in Kucukdeveci, building on its prior judgment in Mangold, the CJEU had stated that employees could invoke the principle of non-discrimination on grounds of age against private employers, and this principle is set out in the Charter (Article 21(1)). Can Article 27 of the Charter also been invoked against private employers?

No it can't. According to the Court, it's 'clear from the wording of Article 27 of the Charter that, for this article to be fully effective, it must be given more specific expression in European Union or national law'. More precisely, 'It is not possible to infer from that Article or the explanations to it' that the rule in the Directive is a 'directly applicable' rule which addresses a prohibition to Member States from excluding categories of employees when implementing the relevant Directive.

The CJEU distinguishes its prior judgment in Kucukdeveci, on the grounds that the principle of non-discrimination on grounds of age laid down in the Charter 'is sufficient in itself to confer on individuals an individual right which they may invoke as such.' Finally, the Court states that the Charter Article and the Directive cannot be combined to require that the national law breaching the Directive has to be set aside, because 'that [Charter] article by itself does not suffice to confer on individuals a right which they may invoke as such', so 'it could not be otherwise if it is considered in conjunction with that directive'.


The Court's judgment has clarified some key points. The old argument that the Charter can never apply to private parties at all, since Article 51 of the Charter (which sets out its scope) states that it is addressed to EU institutions and other EU bodies, plus the Member States only when they implement EU law - and so implicitly not to private parties - has surely been rejected by the Court here. While the Court does not reject this argument expressly, its judgment obviously assumes that the Charter can apply to private parties in some cases, otherwise why distinguish between Articles 21(1) and 27 of the Charter? The key point is that the underlying challenge here (as in Kucukdeveci) is to a law adopted by a Member State which breaches an EU obligation, so the Charter still applies to the dispute even on a narrow interpretation of Article 51.

The Court has also expressly rejected the argument - which was a reasonable extrapolation of its judgments in Mangold and Kucukdeveci - that somehow the legal effect of a Directive could be 'supercharged' when combined with a relevant provision of the Charter. Rather, the Court makes clear that the Charter Article must be judged separately.

Most importantly, we now have a test for determining whether Charter provisions can apply against private parties - whether the Charter provision requires 'more specific expression' in national or EU law on the one hand, or on the other hand whether the Charter Article is 'sufficient in itself to confer on individuals an individual right which they may invoke as such'.

The Court elaborates upon the first test, but not the second. It seems that Article 27 requires 'more specific expression' because neither that Article nor the explanations concerning it indicate that the key clause in the worker consultation Directive is a directly applicable prohibition against excluding categories of employees from that Directive.

With great respect, the Court's explanation cannot be taken literally. After all, the Charter and the explanations to it make few references to EU legislation, and none of those references indicate that the key rules in the legislation concerned are directly applicable. Most importantly, neither Article 21(1) of the Charter nor the explanations to it make any reference to the framework employment equality Directive (Directive 2000/78); and yet the Court confirms that Article 21(1) can be invoked in litigation brought against private parties.

Perhaps the key point is that Article 27 refers to the 'conditions' of national or EU law. Several other Charter provisions refer to national or EU law, but without using the word 'conditions' (for instance, Articles 28 and 30, concerning trade union rights and unfair dismissal). Some key provisions do not refer to national or EU law at all: for instance, Article 31, on the right to 'fair and just working conditions', including regulation of working time.

Also, there's a dog that didn't bark in this judgment. Article 52(5) of the Charter sets out a distinction between the legal effect of 'rights' on the one hand, and 'principles' on the other. This distinction is the subject of much academic literature, and was extensively discussed in the Advocate-General's opinion, but the Court doesn't mention it. Neither - for obvious reasons - does the Court mention the Protocol limiting the legal effect of the Charter in the UK and Poland. But of course it could arguably be relevant in litigation in those countries which seeks to apply the Charter against employers.


The Court has begun to answer some key questions about the horizontal effect of the Charter. This is an improvement from the most recent judgment on this issue (Case C-282/10 Dominguez), which avoided answering any of them. But the answers in the ASM judgment raise many further questions in turn, which the Court will surely be called upon to resolve in future.

Barnard & Peers: chapter 9, chapter 20