Friday 27 June 2014

New EU rules on maritime surveillance: will they stop the deaths and push-backs in the Mediterranean?

Steve Peers


A new EU Regulation, published today in the EU’s Official Journal, sets out new rules on maritime surveillance and rescue operations coordinated by Frontex, the EU’s borders agency. What effect will these rules have on reducing the tragic death toll of migrants in the Mediterranean? And what will happen to the asylum claims of those rescued or intercepted in the high seas?

These new rules are a response to the continued argument that the EU must bear at least some of the blame for the deaths of migrants in the Mediterranean. Furthermore, Member States’ authorities and Frontex have often been blamed for violent behaviour or ‘push-backs’: the forced return of migrants’ vessels to unsafe countries, which were condemned by the European Court of Human Rights in its 2012 judgment in Hirsi v Italy.

The Regulation replaces prior rules adopted by the Council alone in 2010, in the form of a Decision implementing the Schengen Borders Code, which was annulled by the Court of Justice of the European Union (CJEU) after the European Parliament (EP) challenged it on procedural grounds. According to the Court, an EU act concerning human rights and coercive measures had to be adopted by means of the EU’s legislative process.

That meant that the European Commission had to propose a legislative measure, which it did in April 2013. At first, a hard-line group of Member States opposed most of the provisions in this proposal concerning search and rescue and disembarkation (ie the rules on the destination of migrants who were intercepted and rescued), even after the particularly tragic loss of 300 migrants’ lives in autumn 2013. However, these Member States relented, and the European Parliament also pressed to retain and improve upon the Commission’s proposal.

The new Regulation was subsequently agreed, and will come into force on 17 July. But does it mean that the EU will be doing enough to address the loss of life and push-backs in the Mediterranean?

This post addresses these issues in turn, and concludes with an assessment of the issue of the accountability of Frontex. It is an updated and amended version of a previous Statewatch analysis on the new rules.

Search and rescue

The previous Decision contained binding rules on interception, and non-binding rules on search and rescue and disembarkation. However, the CJEU said that even the latter category of rules was binding. As noted above, a group of Member States wanted to water down (as it were) most of these rules in the new Regulation, but was ultimately unsuccessful.

There are new rules on search and rescue, which retain (at the EP’s behest) the Commission’s detailed proposal on this issue, including particularly the definition of whether vessels can be considered in a state of alert, uncertainty or distress. Provided that sinking vessels are detected in time and that these rules are properly applied, the Regulation should therefore ensure that migrants are rescued from drowning wherever possible.

Protection and disembarkation rules

On the other hand, the situation is more complex as regards the rules on protection of those migrants who are potentially at risk of persecution, torture or other forms of ill-treatment in their country of origin (or another country).

The core of the new Regulation is Article 4 – the protection against non-refoulement (removal to an unsafe country) and protection of fundamental rights. Article 4(1) states that no-one can be ‘disembarked in, forced to enter, conducted to or otherwise handed over to’ an unsafe country as further defined in the Regulation. Compared to the 2010 Decision, the Commission proposal, and the Council position, the EP successfully insisted on adding the words ‘forced to enter’ and ‘conducted to’, which clearly covers push-backs.

What is an unsafe country? Article 4(1) goes on to define two situations: (a) a serious risk of subjection of the migrant to the death penalty, torture, persecution or other inhuman or degrading treatment; and (b) the migrant’s life or freedom would be threatened on the grounds set out in the Geneva Convention on Refugees (race, religion, nationality, political opinion or membership of a social group), as well as sexual orientation. Also, a ‘chain refoulement’ is banned: a migrant cannot be handed over to a country which is safe in itself, but which would hand the migrant over to an unsafe country.

Compared to the criteria in EU asylum law (the ‘Qualification Directive’), the first category includes two of the grounds concerning the grant of ‘subsidiary protection’ (ie protection for those who do not qualify as refugees under the Geneva Convention): the death penalty and torture or other inhuman or degrading treatment. It does not include the third category, concerning ill-treatment in the event of armed conflict; however, it does include ‘persecution’, without further definition. The second category is identical to Article 33(1) of the Geneva Convention, except that it does not include the exception in Article 33(2) of that Convention for persons posing security threats et al., and it adds the grounds of ‘sexual orientation’ to those referred to in the Convention. However, the CJEU has confirmed that homosexuals can form a ‘particular social group’ under the EU’s Qualification Directive.

The 2010 Decision referred simply to ‘non-refoulement’ without any further explanation in the main text, while the 2013 proposal (and the Council’s position) referred only to the first category of grounds, without the general reference to ‘persecution’. So the EP clearly succeeded in strengthening this provision.

Next, how must an unsafe country be determined? Article 4(2) states that when considering disembarking migrants in a third country, the host Member State (the Member State from which an operation takes place or from which it is launched: Article 2(3)) must ‘take into account the general situation in that third country’, and cannot disembark or otherwise force to enter, conduct to or hand over if the host Member State or other participating Member States ‘are aware or ought to be aware’ that such a State presents such a risk. The EP insisted on adding the references to forcing to enter, conducting to or handing over.

The EP also obtained an amendment further clarifying the sources of information to take into account – a ‘broad range’, including other Member States, EU bodies, agencies and offices and international bodies. The Member States ‘may’ take into account existing agreements and projects carried out using EU funds.

What are the migrants’ procedural rights? Article 4(3) of the Regulation specifies that (in accordance with the Hirsi judgment) before disembarking or otherwise conducting, etc the migrants to a third State, taking into account the general situation in that State, the Member States’ units shall ‘use all means’ to identify the migrants, assess their circumstances, inform them of their destination and give them an opportunity to object on grounds of the non-refoulement rule. These obligations are subject to an override in the interests of the safety of all the persons involved (see Article 3).

The operational plan must ‘where necessary’ provide for medical staff, interpreters, legal advisers and other relevant experts on shore. Also, the annual reports which Frontex must provide on the application of the Regulation must include ‘further details’ on cases of disembarkation in third States, as regards the application of the relevant criteria. These provisions were insisted upon by the EP, in particular the reference to ‘legal advisers’, but there is ‘wiggle room’ conferred by the words ‘where necessary’ and ‘use all means’.

There are also limits on the exchange of personal data with third countries, an obligation to respect human dignity, and rules on training of staff.

The protection rules cannot be separated from the rules on disembarkation of migrants. According to Article 10, there are three scenarios. First, if migrants are intercepted in the territorial sea or contiguous zone (the waters adjacent to the territorial sea, according to international law: see the definition in Article 2(13)) of a Member State, then they must be disembarked in the coastal Member State, ie the Member State in whose territorial waters or contiguous zone the operation takes place (definition in Article 2(14)). But this is subject to a crucial exception: it is possible under the Regulation that a vessel that has made it this close to a Member State could still be ordered to alter course towards another destination.

Secondly, if migrants are intercepted in the high seas, they may be disembarked in the country from which they are assumed to have departed, subject to the non-refoulement rules in the Regulation. If that is not possible, then disembarkation ‘shall’ take place in the host Member State.

Thirdly, in the event of a search and rescue, the migrants shall be disembarked in a place of safety. If that is not possible, then they shall be disembarked in the host Member State.

These provisions raise many important questions. First of all, it should have been clearly specified that the general non-refoulement rule takes priority over any possibility of disembarkation in a third State.

Secondly, as correctly noted in the preamble to the Regulation, the EU’s asylum legislation applies to anyone in the territorial waters of the EU. This means that, in accordance with that legislation, once an asylum application is made in the territorial waters, the asylum applicant cannot be removed to a third State before there is a decision on the asylum application in accordance with that legislation, save for some limited exceptions not relevant here. The obvious corollary of this is that asylum-seekers who make their application in the territorial waters must be disembarked on the territory of the Member State concerned, since it is unlikely that it will be practical to keep them on board a ship for the entire duration of a full asylum procedure. However, the main text of the Regulation does not reflect the wording of this legislation, since it provides for the possibility of persons intercepted or rescued in the territorial waters to be removed to third countries.

Thirdly, when migrants are disembarked on the territory of a Member State, an awkward question could arise: is that Member State safe? While the specific non-refoulement rules in Article 4(2) and (3) refer to the safety of third countries, the general rule in Article 4(1) refers to countries in general. The European Court of Human Rights and the EU’s Court of Justice have already both concluded that Greece in effect fails the standard set out in Article 4, and litigation in some Member States is also challenging the safety of Italy. So there could be a clash between the non-refoulement rule and the obligation to disembark in a Member State which is the host State, coastal State or place of safety, or in the territorial waters of which the applications were made.

Fourthly, as for those intercepted or rescued in the high seas or the contiguous zone (the Regulation does not contemplate the scenario of migrants being intercepted in the territorial waters of third States), the bulk of the EU’s asylum legislation does not apply. However, the EU’s qualification Directive does – since there is nothing in the text of that Directive to limit its territorial scope. But the wording of the Regulation is confusing in this regard, since it does not refer to the detailed text of that Directive but rather to general standards on non-refoulement, which are different from that Directive in some respects, as noted above (the omission of persons fleeing conflict, for instance).

Having said that, EU rules on asylum procedures and reception conditions do not apply to asylum-seekers who are intercepted or rescued in the high seas or the contiguous zone, and in that case the rules in the Regulation would apply. In effect, the rules summarised above provide for a highly simplified process – which might be dubbed the ‘maritime asylum procedure’ – for such cases. As noted above, though, the words ‘where necessary’ and ‘use all means’ arguably give Member States considerable flexibility not to apply these rules fully, and these rules are (understandably) subject to the requirement to give priority to the safety of all persons. This should mean that in the event of a risk to the safety of persons, if the application of the non-refoulement rule has not yet been assessed, the migrants must be taken to a (safe) Member State to avoid prejudicing the outcome of that assessment. Once the migrants enter a Member State’s territorial waters, EU asylum law will apply fully (arguably it applies even if the application was made before the vessel entered those waters; if not, then there is nothing to stop the asylum-seeker making a renewed application for asylum once the vessel is in those waters).

Since most EU asylum law does not apply to the high seas, the EU’s Dublin rules on asylum responsibility do not apply either, and it is an open question whether they would apply where a person made an application on the high seas and was then brought to the territorial waters of a Member State while the application was being considered. In order to limit somewhat the huge impact of these Dublin rules on the coastal Mediterranean States, EU Member States could agree between themselves on new rules for responsibility for asylum-seekers who make their application on the high seas. If necessary, this could take the form of an amendment to the Dublin rules, added to the proposed new amendment on the application of those rules to unaccompanied minors.

Finally, at the EP’s behest, the preamble to the Regulation clarifies (recital 7) that a shipmaster and crew should not face criminal sanctions for rescuing migrants and bringing them to a place of safety. This provision is welcome, but it would be better if the EU legislation on criminal sanctions for facilitating irregular migration were amended to confirm that there is no criminal liability in such cases.

Accountability of Frontex (and national authorities)

Article 13 of the Regulation, which was inserted at the EP’s insistence, states that Frontex must make annual reports on the application of the Regulation, including on Frontex’s own procedures and information on the application of the Regulation in practice, including ‘detailed information on compliance with fundamental rights and the impact on those rights, and any incidents which may have taken place’. Presumably this means that these reports will have to include full information on where migrants were disembarked and the assessments that were made of the safety of any third countries (and Member States) in each particular case. It would have been better to clarify the extent of these obligations expressly, although any provision on accountability is better than none. The EP also insisted on a recital in the preamble, which repeats text already in the Frontex Regulation regarding Frontex cooperation with third countries. 

When the Frontex Regulation was last amended in 2011, the EP insisted on many new provisions ensuring that Frontex would be complaint with human rights standards, and the application of these new rules has since been examined by the European Ombudsman. Frontex is still refusing to establish an individual complaints procedure for migrants who believe that it has violated the rules binding upon it, on the grounds that it only coordinates Member States’ authorities’ actions, so cannot be held directly responsible for those actions.

While it would be preferable to ensure that individuals could raise complaints that Frontex had not complied with its obligations as regards human rights, it is also true to say that Frontex only coordinates Member States’ authorities’ actions. In any event, those authorities take actions that are not coordinated by Frontex. Given that (as the new Regulation itself implicitly accepts) any control of the EU’s external border, including by means of patrols outside a Member State’s territorial waters, is linked to the application of the EU’s own rules on external border controls, the EU Charter of Rights is applicable to Member States’ control of those borders. And there have been allegations that Member States’ authorities have on some occasions been responsible for push-backs and ill-treatment of migrants at the external borders.

In this context, it is possible that Frontex has been serving for too long as a ‘lightning rod’ for critics of the EU’s external borders control policy, whereas attention should have focussed more on Member States’ authorities, whether they are being coordinated by Frontex or not. The ‘right to life’ in the European Convention of Human Rights entails, according to the European Court of Human Rights, an obligation to hold an independent investigation into losses of life that have arguably resulted from actions of the authorities. So arguably the EU is under an obligation pursuant to the EU Charter of Fundamental Rights to ensure that its Member States conduct such investigations into losses of life which are linked to the implementation of EU policies, in this case the EU external borders rules. Those authorities should also be held accountable for any alleged push-backs or other ill-treatment of migrants at the external borders.

To that end, the EU should agree upon a general framework for independent investigations into such alleged abuses, with the results of these investigations reported and assessed by the Commission as part of its twice-yearly report on the Schengen system. Furthermore, it is long past time for the Commission to stiffen its backbone and to bring infringement proceedings against Member States where there is sufficient evidence that their authorities are responsible for push-backs or other ill-treatment.


The existence of this Regulation is welcome, as its rules on search and rescue are valuable and its provisions on protection, disembarkation and accountability of Frontex are better than nothing at all. But the complex interplay of the provisions of this Regulation with EU asylum law has led to something of a ‘dog’s dinner’ of rules governing the asylum applications of people rescued or intercepted in the Mediterranean, and the rules on the accountability of Frontex are something of a ‘red herring’ in light of the allegations of serious misconduct in some cases by national authorities. Finally, the Commission’s continued unwillingness to bring infringement proceedings in this area (and in the face of its own documented breaches of other EU immigration and asylum law) is undermining the letter and spirit of the Charter by allowing Member States’ authorities to think that they can violate the Charter with impunity. 

Further reading: Council of Europe Parliamentary Assembly report, June 2014

UNHCR comments on the Commission proposal

Barnard & Peers: chapter 9, chapter 26

Unaccompanied minor asylum-seekers: a step in the right direction?

Steve Peers

The EU’s Dublin system on the allocation of responsibility for asylum-seekers, which sets out rules determining which single Member State has responsibility for considering an asylum-seeker’s application, has long been decried as unworkable and an infringement of human rights. These broader arguments will be addressed in a future post on this blog, but for the moment it’s a good time to have a look at yesterday’s proposal from the Commission to ameliorate the Dublin system’s effects a little.

The proposal only covers applications by unaccompanied minors, who make up a small proportion of the total number of asylum-seekers. Nevertheless, they constitute a particularly vulnerable group. Currently, the Dublin III Regulation states that the Member State responsible for an unaccompanied minor asylum-seeker is the Member State where a family member or sibling of the minor is legally present, if that is in the best interests of the child. A ‘family member’ is defined as a spouse, parent or guardian (if the minor is unmarried), or child of the asylum-seeker, if the family already existed in the country of origin. If the minor is married but his or her spouse is not in the EU, then the Member State where his or parents are legally present is responsible.

If there are no such persons on the territory of a Member State, then the Member State where a ‘relative’ of the unaccompanied minor is legally present is responsible, if that relative can take care of him or her.  A ‘relative’ is defined as an aunt, uncle or grandparent.

If there are family members, siblings or relatives in multiple Member States, then the responsible Member State is decided on the basis of the best interests of the child.

Finally, if no such persons can be found, the ‘default’ rule is that the Member State responsible is the State where the unaccompanied minor has applied for asylum, if that is in his or her best interests. But the legislation does not expressly address how this rule applies if the minor applies for asylum in more than one Member State.

When the Dublin III Regulation was negotiated in 2012, these provisions were hotly disputed between the European Parliament and the Council. The two institutions finally agreed to leave the default rule untouched, since there was a pending case on the interpretation of that rule before the CJEU. However, they adopted a joint declaration calling on the Commission to consider making a proposal to amend the legislation as regards that issue once the Court had delivered its ruling.

The Court gave its ruling in that case (MA) in June 2013, and the Commission has now presented its proposal. According to the Court, where an unaccompanied minor has already made an application for asylum in one Member State, and then proceeds to make an application in another one, the default rule should be interpreted to mean that the Member State where the most recent application was made is responsible for the application.

In the Court’s view, this interpretation was necessary first of all for literal reasons, as the legislation does not refer to the first Member State where the application was made. Secondly, unaccompanied minors were vulnerable persons, so the procedure to apply the Dublin rules should be as short as possible. Finally, this interpretation was most consistent with the ‘best interest of the child’, as required by Article 24 of the EU Charter of Fundamental Rights.

Yesterday’s proposal would first of all entrench the Court’s interpretation as an express rule in the Dublin Regulation, covering all cases where the asylum-seeker had applied in multiple Member States. This is the interpretation that the EP favoured, and had tried to fight for as an express rule back in 2012.

It would also set out a new rule which would apply when the unaccompanied minor has made an asylum application in one Member State, and is present in a second Member State, but has not made an application in the second Member State.  In that case, the second Member State should ‘inform’ the child of the possibility of applying for asylum there, and give him or her an ‘effective opportunity’ to apply.

If the minor takes up the opportunity to apply for asylum in the second Member State, that Member State becomes responsible for the application. If he or she does not apply, then the Member State where he or she has most recently applied is responsible.

The new proposal also specifies that the Member States concerned shall cooperate to establish what the best interests of the child are. Finally, it sets out an express rule on Member States informing each other of their decisions; the CJEU had already established such an obligation in the MA judgment.


In principle this proposal is very welcome, given that it exempts unaccompanied minors from the full rigour of the Dublin rules. While this might reduce the pressure to reform those rules, there is little prospect of a fundamental change in those rules (at least at the EU legislature's behest; the courts may be a different matter). So at least one category of vulnerable persons can escape from them. 

More particularly, the proposal both confirms the ‘applicant’s choice’ rule for unaccompanied minors without family members first established by last year’s judgment, and also extends that rule to more situations. However, there are certain limits on the proposal, and flaws which could be improved.

First of all, it should be noted that the proposal does not cover the position of those whose application for asylum has already been rejected in another Member State. But the Court of Justice referred to that issue in its MA judgment, noting that in such cases the second Member State has the option to treat the application as inadmissible in accordance with the EU’s asylum procedures rules. Presumably that rule would continue to apply here.

On this point, however, in principle there should not be many fast-track refusals of applications by unaccompanied minors, given the convoluted exceptions from some of the fast-track rules set out in the second-phase Directive on asylum procedures.

Secondly, the new amendment would not clarify what is meant by the obligation to ‘inform’ the child about applying for asylum and give him or her an ‘effective opportunity’ to apply for asylum. Although there is a rule in the Regulation already spelling out an obligation to inform asylum-seekers about Dublin rules, it only applies once the person concerned has applied for asylum. So it doesn’t cover the position of those who haven’t applied yet. There should be express rules on this issue, in order to ensure that the child is made fully aware of the choice of making a fresh application.

As for the effective opportunity to apply for asylum, there is a (rather vague) rule on this issue in the second-phase Directive on asylum procedures. But it isn’t clear if the proposed new rules in the Dublin Regulation would confer less, more or the same degree of protection than the rules in the Directive.

Thirdly, the prospect of Member States deciding between themselves on what is in the best interests of the child is problematic. Surely, in accordance with the EU Charter, the child has a right to express a view on this issue, especially since EU asylum law guarantees him or her the right to a representative for this very purpose? So the legislation should guarantee that any decision which Member States take which is purportedly in the best interests of the child should be open to challenge, following a prior right to a hearing and the disclosure of relevant information on this subject.

Next, the proposal may mean that more Member States in practice embark on the controversial practice of trying to determine the age of teenagers who claim to be 'minors' (the EU legislation defines this as being under 18). The asylum procedures Directive has rules on this issue, but there are no rules on what happens if the person turns 18 during the procedure.  

Finally, it must be noted that the rules will only be relevant for those minors who have the effective possibility of moving between Member States. They will need some knowledge of which Member States they might wish to apply to, and how to get there, as well as the money to arrange for their travel. Furthermore, they will obviously not be able to move to another Member State in order to apply there if they are in detention. On this point, it should be noted that the revised reception conditions Directive provides that unaccompanied minors can only be detained in ‘exceptional circumstances’.

Given the freedom of unaccompanied minors to make applications in another Member State, it cannot be argued that this possibility justifies their detention. In fact, the grounds for detention in the reception conditions Directive do not provide for such a case. Similarly, while the Dublin rules do allow detention if there is a ‘significant risk of absconding’, it is difficult to see how that ground for detention can apply to unaccompanied minors in light of these rules.  

Barnard & Peers: chapter 9, chapter 26 

Tuesday 24 June 2014

The CJEU ensures basic democratic and judicial accountability of the EU’s foreign policy


Steve Peers

In today’s judgment in European Parliament v Council, the Court of Justice of the European Union (CJEU) has ensured that a minimum degree of democratic and judicial scrutiny applies to the EU’s Common Foreign and Security Policy (CFSP).

The European Parliament (EP) was challenging the Council’s decision to sign and conclude (ie, ratify) a treaty between the EU and Mauritius, relating to the EU’s military operation as regards pirates in the Indian Ocean. As discussed further in the previous blog post concerning the Advocate-General’s opinion, one important aspect of this treaty (and similar treaties with other nearby States) was the inclusion of rules governing the transfer, trial and treatment of alleged pirates to Mauritius, once the EU caught them.

The dispute had two elements – substantive and procedural – which will be considered in turn.

Substantive issues

In the Council’s view, this treaty ‘exclusively’ concerned the CFSP. In that case, Article 218 TFEU (the clause which sets out the procedures for negotiating and concluding international treaties by the EU) states that the EP does not even have to be consulted. But in the EP’s view, it had the right of consent over this treaty, since it also touched upon issues of development policy and criminal and police cooperation. Article 218 provides that where a treaty concerns issues where the EU’s ordinary legislative procedure applies (such as those policy areas), the EP has a right of consent before a treaty can be concluded.

The Court assumed that the EP was not actually arguing that the EU/Mauritius treaty should also have been adopted by means of the development policy, police and judicial cooperation legal bases. So therefore it confined itself to ruling on a narrow legal point. Article 218 states that when a proposal to negotiate a foreign policy treaty is first made to the Council, the Commission and the High Representative for the CFSP shall make recommendations where that treaty relates ‘exclusively or principally’ to the CFSP.

But the CJEU ruled that the different wording (‘exclusively or principally’ when recommending a negotiation, and ‘exclusively’ when concluding a treaty) was immaterial. In its view, the basic rule in Article 218 is an underlying parallelism between the internal powers of the EP and its external powers. So where it has an internal power of consent and control (for instance, pursuant to the ordinary legislative procedure), the EP should have a corresponding external control over the conclusion of treaties.

This approach inevitably meant that the EP lost its substantive argument, since Article 36 TEU only gives the EP the power to be informed and consulted of the main elements of the CFSP.

The Court’s analysis contrasts with that of the Advocate-General, who assumed that the EP was arguing that the treaty should have had also the legal bases regarding development policy and police and judicial cooperation. (His interesting arguments were fully discussed in the previous blog post).

Procedural issues

The EP’s procedural argument was based on a new clause added by the Treaty of Lisbon (Article 218(10)), which states that the EP ‘shall be immediately and fully informed at all stages of the procedure’. In this case, the Council had informed the EP of its decision to sign and conclude the agreement over three months after it had taken it.

But before the Court could rule on the merits of this argument, it had to rule on whether it had jurisdiction to consider it. Article 275 TFEU rules out the Court’s jurisdiction as regards the CFSP, except as regards sanctions on individuals or ‘legal base’ arguments.

Nevertheless, the Court asserted its jurisdiction, arguing that the rule in the Treaty was its ‘general jurisdiction’, and so Article 275 was a ‘derogation’ which had to be ‘interpreted narrowly’. So the Court did have jurisdiction to consider issues of procedure as distinct from substance, since the procedural legal basis for CFSP treaties is Article 218, ie a provision outside the scope of the CFSP rules as such.

The Court then ruled in favour of the EP’s main argument.  Unlike the Advocate-General, it reached the common-sense conclusion that a three-month delay in informing the EP did not constitute ‘immediately’ informing it. Publishing the treaty in the Official Journal was not enough, since the purpose of the information requirement was to ensure ‘democratic scrutiny’ by the EP, including as regards legal base issues. The failure to inform the EP was an ‘essential procedural requirement’, ie a ground to annul the act concerned under Article 263 TFEU, since it was ‘an expression of the democratic principles on which the [EU] is founded’.  

This information requirement applied even though the treaty in question concerned the CFSP, where the EP had no power of consent or even consultation. The Court emphasised that the information requirement applied to any sort of treaty, and to any stage of the procedure.


If the EP had won its argument on the substantive issues, it would have gained the power of consent over most foreign policy treaties, given their links (which the Treaties in fact require) with other aspects of EU external policies. It does appear that the EP was trying to argue a more general legal base point, and so the Court’s decision to confine itself to a more technical analysis is disappointing. If the Court had ruled on the substantive legal base arguments, it would have been its first clarification of the external scope of the EU’s policing and criminal law powers, and its first clarification of the scope of CFSP powers since the Treaty of Lisbon came into force. So the missed opportunity is unfortunate. Perhaps next time, the EP will challenge a foreign policy treaty more unambiguously, so that its legal base arguments are decided on the merits.

It could have been argued that the exception (as regards CFSP) to the normal rule that the EP has consent or consultation powers over international treaties is also a ‘derogation’ from the normal rule which had to be ‘interpreted narrowly’, just like the rule limiting the Court’s CFSP jurisdiction. However, the Court adopted different rules of interpretation as regards these two issues. One could deduce that the Court is more concerned about asserting its jurisdiction than enhancing the EP’s rights.

But having said that, the Court’s ruling on the procedural point is quite significant. First of all, as regards its own jurisdiction, the Court’s ruling means that any CFSP measure can be litigated before it, as long as the legal arguments relate to a procedural rule falling outside the scope of the CFSP provisions of the Treaty (Title V of the TEU). For instance, it arguably means that the Court would have the power to rule on the compatibility of proposed CFSP treaties with EU law, since that jurisdiction is conferred by Article 218 TFEU and not expressly ruled out by Article 275. But such disputes might often include arguments about the substance of the measure concerned (for instance, whether it would breach the EU’s human rights obligations), and it could be awkward to distinguish between procedural and substantive issues in practice.

Secondly, as for the EP’s information rights, the Court’s reference to ‘all types of procedures envisaged in Article 218’ surely means that the EP must be given information at least about the recommendations to start negotiations, the decisions to open negotiations, the negotiating directives, the suspension of agreements and the adoption of EU positions in bodies set up by agreements, since all of those measures are referred to in Article 218. This allows the EP to indicate its point of view on draft or planned treaties or measures implementing them at an early stage of the process, allowing for further democratic input into the treaty-making process.

More broadly, the CJEU quite rightly rejected the absurd view that a three month wait could constitute an ‘immediate’ notification.  It applied the literal interpretation of the Treaty rule (applying to all treaties, and to every stage of the procedure) as well as the purposive rule (advocated in the previous blog post), agreeing that the purpose of the rule was to ensure democratic scrutiny and to allow the EP to ‘make known its views’. As I previously pointed out in that blog post, such public scrutiny would be an opportunity to discuss, for instance, whether the country in question had a problematic human rights record, and therefore to influence public debate over whether the treaty in question ought to be concluded.


Barnard & Peers: chapter 5, chapter 24, chapter 25

Thursday 19 June 2014

Pregnant workers and EU citizens’ free movement rights

Steve Peers

Jessy Saint Prix, a young French woman, came to live in Britain in 2006, in order to embark on a career as a school teacher. In the event, she has learned rather more than she probably anticipated about EU law and the English legal system.

Having worked for a year in the UK as a teaching assistant, she started a PGCE course, which would qualify her to be a teacher. In the midst of that course, she became pregnant.  So she dropped her course and registered with an agency. She had a few stints working in nurseries, but she found that work too difficult, now that she was nearly six months pregnant. Then she looked for other work, but didn’t find any, and so she gave up work about 12 weeks before the expected date of confinement.

In the meantime, her baby arrived. But despite her previous work as a teaching assistant and her later agency work, the UK government refused her claim for income support, as she was defined not as a worker, but as a  ‘person from abroad’. She challenged this in court, winning at first instance. But the government won at two further levels, and then the Supreme Court referred questions on the case to the Court of Justice of the European Union (CJEU). The CJEU finally ruled in her favour today. 

The main thrust of the Court’s analysis was the circumstances in which EU citizens retain the status of worker, and therefore access to benefits, having previously been employed in the host Member State (like Ms. Saint Prix). The EU’s citizens’ Directive states that a person retains the status of worker if they are ‘temporarily unable to work as a result of an illness or accident’, or in ‘involuntary unemployment’ (subject to certain conditions) or undertaking vocational training, which must be linked to their prior employment (as in Ms. Saint Prix’s case, before she quit her course) unless they are unemployed involuntarily. The relevant provision (Article 7(3) of the Directive) does not state whether this list is exhaustive or not.

But for the Court, that didn’t really matter. It reiterated its established case law that the concept of ‘worker’, referred to in Article 45 TFEU, is a matter of primary law. So the concept couldn’t be narrowed down by secondary law, although the Court confirmed that, in accordance with its sex discrimination case law, pregnancy could not be equated to an ‘illness’ under the Directive.

In this case, the definition of a worker could encompass someone in Ms. Saint Prix’s situation, who had previously been working in the host State, who gave up work less than three months before the birth due to the physical constraints of late pregnancy, and who began work again three months after childbirth, without leaving the Member State in question. Bizarrely, the Court compares her situation to an earlier judgment concerning prisoners who retained ‘worker’ status during their jail time.

However, retaining worker status in such a case is not unlimited. It is conditional upon returning to work or finding another job ‘within a reasonable period after confinement’. But what is a ‘reasonable period’? The CJEU said that this concept must take account of the ‘specific circumstances of the case’ as well as national law on the duration of maternity leave, in accordance with the EU Directive on pregnant workers.

That Directive provides for a minimum period of 14 weeks’ leave, at least two weeks of which must come before confinement. However, national law can set longer periods, and most Member States do. Although the Commission proposed back in 2008 that the Directive should be amended, in order to extend maternity leave to a minimum of 18 months, this proposal has not been agreed, and the Commission is now contemplating withdrawing it.

It should be emphasised that the CJEU made no reference to compliance with national law on the period of leave taken before confinement. Also, while the pregnant worker’s Directive allows Member States to set a qualifying period of work of up to a year before obtaining the right to maternity pay, there is no such condition as regards the right to maternity leave. So for the purposes of free movement law, the only qualification is that the person concerned can be defined as a worker.

The Court’s final point was to note that in order to obtain permanent residence status under the citizens’ Directive, which requires five years’ stay on the territory, a period of up to one year’s interruption of residence due to pregnancy or childbirth does not break continuity of residence. So a fortiori this should mean that worker status is retained during breaks due to pregnancy. However, the Court clearly does not mean by this that worker status would be retained during breaks lasting up to one year, in light of its earlier reference to the limits on the rule.   

Unlike the Advocate-General’s opinion, the Court of Justice made no mention of the Charter of Fundamental Rights or to sex discrimination issues more generally. Nevertheless, its judgment is obviously of great importance to migrant women who are EU citizens. It could also have a knock-on effect on the interpretation of EU law apply to third-country nationals. For instance, the meaning of ‘worker’ under the EU/Turkey association agreement is the same as under EU free movement law.

What are the limits and possibilities of this judgment for migrant women? One limit is that they have to qualify as a worker in the host State in the first place.  The Court’s insistence that it will define the concept of ‘worker’, despite the possible constraints of secondary law, leaves open the possibility that it will also override the limits in the citizens’ Directive relating to job-seekers’ access to benefits. However, the Court will be ruling on this issue separately in the near future (see the opinion in the pending Dano case).

Another limit is that they have to return to work (ie with their employer), or to find fresh work if they were temporarily unemployed at the time of childbirth, as in Ms. Saint Prix’s case. In fact, EU law makes it easy for a pregnant worker to retain her employment, since dismissal on grounds of pregnancy is banned by the pregnant workers’ Directive, except as regards ‘exceptional cases’ not linked to the pregnant worker’s condition. Anyway, pregnant workers would not normally need to apply for income support in such cases, since they can claim maternity pay or allowances under the pregnant workers’ Directive instead. But access to income support might still be relevant for those pregnant workers who have not yet satisfied the qualifying period for maternity pay under that Directive, as discussed above.

It’s obviously more problematic for a new mother who was unemployed by the end of her pregnancy to find work by the end of the national maternity leave period. Although Ms. Saint Prix was fortunately able to do so, not everyone will be able to. This could raise the awkward question of whether income support would have to be repaid in such cases.

There will likely be some unfortunate cases where ongoing medical problems arising from pregnancy and/or childbirth make seeking work impractical, even after the end of the maternity leave period. This issue has led to a mass of complicated and controversial case-law in the context of sex discrimination law. However, it should be recalled that under the citizens’ Directive, ‘worker’ status is retained by those former workers who are too ill to work. So because the CJEU’s sex discrimination case law treats medical problems which initially arose from pregnancy or childbirth as an ordinary illness, once the period of maternity leave is over, any women with such problems will still retain their ‘worker’ status (and therefore access to any relevant benefits) for the purposes of free movement law.  

Also, there will be some cases where the new mother does not find new work by the end of the maternity leave period despite her best efforts, or (more rarely) where her initial employer has become insolvent. For these difficult issues, it should be recalled that the Court refers to an obligation to consider the ‘specific circumstances of the case’, without further explanation. Arguably, this adds to, rather than qualifies, the reference to national law on maternity leave periods. In any event, the more general limits on expelling unemployed EU citizens would apply in such circumstances.  

It should be noted that the judgment logically applies also to those women who have previously worked in the host State, and then stopped work due to accident, involuntary unemployment, or vocational training, subject to the conditions in the Directive, since they all retain the status of ‘worker’. For Ms. Saint Prix, it wasn’t feasible to complete her course at the time she had initially planned, given the timing of her baby’s arrival. But for other women, whose baby is due after the end of their course, this could be of great practical importance.

Finally, today’s judgment should be seen in its broader context. EU law has long encouraged and facilitated the equal treatment of women on the labour market, including those women who are pregnant or who are mothers. It has also long encouraged and facilitated the free movement of workers. This judgment is the first opportunity for the CJEU to bring these two aspects of EU law together, and to confirm that migrant pregnant women also benefit fully from EU free movement law. Indeed, it is self-evident from the facts of this case that Ms. Saint Prix – like migrant workers generally – paid far more into the UK’s tax and benefits system than the benefits which she received from it.

Barnard & Peers: chapter 13, chapter 20

Wednesday 18 June 2014

Europe v Facebook: the beginning of the end for NSA spying on EU citizens?

Steve Peers

Since the revelations about the extent of spying by the American National Security Agency (NSA) revealed by Edward Snowden, doubts have increased about the adequacy of the data protection regime in the United States, in particular as regards its impact on EU citizens, who are subject to the more favourable regime established by the Data Protection Directive. One aspect of these doubts concerns the ability of the NSA to examine the content of communications processed by social media companies based in the USA, such as Facebook.

Today’s decision by the Irish High Court to send questions in the ‘Europe v Facebook’ case to the CJEU raises the possibility that the NSA’s access to EU citizens’ personal data might soon come to an end. But it’s not clear if the CJEU will address the most essential issues directly, because the case raises a number of complex legal issues that need to be examined in more detail.

As a starting point, the basic legal regime governing transfers to Facebook is the ‘Safe Harbour’ system, which takes the form of a Commission Decision finding that all American companies certifying their participation in a system for complying with basic data protection principles maintain an ‘adequate’ level of data protection. This is one of the ‘adequacy decisions’ that the Commission can make pursuant to the rules on the data protection Directive on transfers of personal data outside the EU (see further my recent blog post on the planned reforms to this system). Despite the doubts arising from the Snowden revelations, the Commission’s most recent report on the Safe Harbour system did not suggest that the system should be 

Not everyone accepts these assertions, however. An Austrian citizen, Mr. Schrems, complained about the transfer of his personal data as a Facebook user pursuant to the Safe Harbour rules to the Irish data protection authority, which was competent in this matter because Facebook has a subsidiary in Ireland. The national authority argued that it could not take a decision on this complaint, because it was bound by the Commission’s decision. Moreover, it argued that the complaint was ‘frivolous’.

Mr. Schrems then challenged the authority’s decision before the Irish High Court. In its ruling today, the national judge therefore decided to send a question to the CJEU. Essentially, the question is whether the national data protection authority is bound by the Commission’s Decision, and whether that authority can conduct its own examination.

The first obvious question in this case is whether the American system infringes EU data protection law. Basing itself on the recent Digital Rights judgment of the CJEU, in which that Court ruled that the EU’s data retention Directive was invalid, the national court clearly believes that it does. While acknowledging the important anti-terrorist objectives of the law, the judge, when examining national constitutional law states that it is ‘very difficult’ to see how such mass surveillance ‘could pass any proportionality test or survive any constitutional scrutiny’. Indeed, such surveillance has ‘gloomy echoes’ of the mass surveillance carried out in ‘totalitarian states such as the [East Germany] of Ulbricht and Honeker’.

The judge equally believes that the US system is a violation of EU law, with no adequate or accessible safeguards available to EU citizens, and no consideration of EU law issues built in to the review process that does exist.

Is this analysis correct? There are two fundamental issues here which the national court doesn’t consider: the scope of the data protection directive, and the derogations from that Directive. On the question of scope, the CJEU previously found in its Passenger Name Records (PNR) judgment that the EU/US agreement which provided for the transfer of data from airlines to the US authorities was outside the scope of the data protection Directive, because it regulated essentially only the activities of law enforcement authorities, and the Directive does not apply to the ‘processing of personal data…in the course of an activity which falls outside the scope’ of EU law, such as…public security, defence, State security…and…criminal law’. On the other hand, the CJEU ruled that the data retention directive was correctly based on the EU’s internal market powers, since it essentially regulated the activity of private industry, albeit for public security objectives. While in this case, it might be argued that the American law in question falls within the first type of law, the Safe Harbour agreement clearly falls within the second. So it is a sort of hybrid question, but on balance the issue falls within the scope of the Directive, since the measure at issue is essentially the Safe Harbour agreement.

Secondly, the external transfer rules in the EU Directive do not refer expressly to the issue of derogations from data protection rights on public security grounds. Yet presumably some such derogations can exist, given that the Directive itself provides for public security derogations as regards the standard EU rules. Surely the security exceptions applied by third countries don’t have to be exactly the same as those applied by the Directive. But some form of minimum standard must apply. For the reasons set out by the national judge, however, there is a strong argument that the US rules fall below the standard of anything which the EU can accept as ‘adequate’.  

Because the national judge takes these two issues for granted, there is no question sent to the CJEU on whether the American regime is either within the scope of the Directive, or violates the minimum standards of adequacy which the EU can accept as regards third states. But both these issues are absolutely essential in the debate over the post-Snowden relationship between the US and EU. It would therefore be desirable if the CJEU addressed them nonetheless.

Next, another problematic issue here is which set of EU data protection rules should apply: the external transfer rules, or the more stringent standard rules? The national court, along with the data protection authority, applies the external transfer rules, given Facebook’s certification under the Safe Harbour system. However, it is doubtful whether this is correct.

As is well known, in the recent Google Spain judgment, the CJEU ruled that the standard rules applied to Google’s search engine function, given that it had an ‘establishment’ in Spain, according to the Court’s interpretation of the rules. As I then argued on this blog, it probably follows from that judgment that the standard rules apply at least to some social networks like Facebook. In any event, the issue will arise again when the revised jurisdiction and external transfer rules, mentioned above, apply.  However, the complainant and the national court assume that the external transfer rules apply. Perhaps the CJEU should also examine this issue of its own motion.

Another problematic issue is the question of how to challenge the inadequacy of data protection in practice in the US, which is the subject of the only question sent to the CJEU. The Safe Harbour agreement addresses this point directly, since it allows national data protection authorities to suspend data transfers as regards an individual company, in accordance with existing national law, if either the US government or the US enforcement system has found a violation of that agreement, or if:

there is a substantial likelihood that the Principles are being violated; there is a reasonable basis for believing that the enforcement mechanism concerned is not taking or will not take adequate and timely steps to settle the case at issue; the continuing transfer would create an imminent risk of grave harm to data subjects; and the competent authorities in the Member State have made reasonable efforts under the circumstances to provide the organisation with notice and an opportunity to respond.

However, Irish national law does not provide for such a system, but simply sets out an irrebutable presumption that the Commission’s adequacy decision is sufficient. This rule may well have played a part in convincing Facebook and the subsidiaries of other US companies to set up in Ireland in the first place.
The challenge argued that the national data protection authority nevertheless had to exercise such powers, and so the national judge asked only whether this was possible. Logically, there can be only one answer, by extension from the NS judgment: Member States cannot create an irrebutable presumption that prevents the exercise of Charter rights, so the national data protection authority must have the powers in question.

In the alternative, or arguably additionally, it must be possible to challenge the validity of the Commission’s adequacy decision in the national courts, which would then have an obligation, if they thought that challenge was well-founded, to send questions on that point to the CJEU. (See the Foto-Frost judgment).

The next problematic issue is the role of the national constitutional protection for human rights. Clearly the national judge believes that the American system breaches the protection for the right to privacy guaranteed in the Irish constitution. Nevertheless, the national court proceeds to examine the issue primarily from the perspective of EU law. So if the CJEU rules against the challenge to the American law on the merits, or does not address those merits for procedural reasons, should the national court proceed to apply Irish law?

In principle, national constitutional law cannot apply here, since EU law, as the national court recognises, has extensively harmonised this issue. This means that, according to the Melloni judgment of the CJEU, only the EU’s human rights standards, in the form of the Charter, can apply. National constitutional standards cannot. But national courts in Ireland (and elsewhere) might be unwilling to accept that outcome.

National law would only apply if the CJEU rules that this issue falls entirely outside the scope of the Directive, as discussed above. If, on the other hand, the processing falls within a public security derogation from the Directive, the EU Charter would apply, by analogy with the CJEU’s recent judgment in Pfleger (discussed here), in which it ruled that the Charter applies to national derogations from EU free movement law. This parallels the argument (discussed here) that national data retention law falls within the scope of EU law, following the Digital Rights judgment, because it is a derogation from the EU’s e-privacy Directive.

Finally, the consequences of any future finding by the national data protection authority that transfers under the Safe Harbour decision must be suspended as regards Facebook must be considered. Assuming that the US had not changed its law in the meantime, Facebook would have a dilemma: should it comply with its US legal obligations, or face the suspension of transfers of data from Europe? Possibly it could avoid this dilemma by ensuring that it only processed EU residents’ data within the EU, potentially avoiding the scope of US law. But this might be expensive, and in any event the US might seek to extend the scope of its law to cover such cases. These issues would inevitably arise for other major US companies as well.

Any real prospect that Facebook transfers from the EU might be blocked would cause a major earthquake in EU/US relations, making the concerns about the recent Google Spain judgment look like a minor tremor. It may be that the only solution is for the US to take more seriously its ongoing discussions with the EU on data protection issues, with a view to reaching a solution that reconciles its security concerns with the basic principles of privacy protection.

Barnard & Peers: chapter 9 

The new guidelines for the Area of Freedom, Security and Justice: some critical comments

Emilio De Capitani
In the coming days the European Council will debate and adopt the long awaited Guidelines which will shape the future of the EU's Area of Freedom, Security and Justice for forthcoming years. These guidelines follow the end of the current Stockholm Programme (2009-2014) and come near the end of the last transitional period for the measures adopted before the entry into force of the Lisbon Treaty on police and judicial cooperation in criminal matters (what remains of the former intergovernmental  'third pillar' cooperation).
Regrettably the draft European Council Conclusions which have been circulated (see the Annex below) and the programme of the incoming “trio” Presidencies (Italian, Latvian and Luxembourg) which will implement them in the next 18 months confirm the worst provisions detailed in our previous post on this issue.
If anyone was searching for proof that European Strategies lack political vision and are a collection of bureaucratic and diplomatic choices, he or she will find in these documents the confirmation of this thesis.  
The emphasis of the European Council on the external dimension of the justice and home affairs polices by privileging soft law instruments such as the Global Approach on Migration or instruments such as mobility partnerships confirm two emerging trends since the entry into force of the Lisbon Treaty : 
- to transfer to the European Council the main EU political choices in the last area where the treaties still do not grant and effective parliamentary and judicial control. 
- to continue to avoid legally binding measures on which solidarity mechanisms can be established (Schengen, Frontex, and Eurosur being the exceptions which confirm the rule). 
Rhetorical declarations aside, the draft European Council guidelines confirm the choice for general (and generic) strategies such as the Internal security strategy or the anti-drugs strategy which are adopted without any debate between the European Council members nor with the European Parliament.
These Strategies should then be implemented by the so called “Policy Cycle” where EU agencies and the Member States representatives - instead of verifying their consistency with national internal security strategies deciding which areas the EU's intervention could add value in - pick and choose (on voluntary basis) some priorities which are approved without debate (as point A) by their ministers without (again) any European or national parliamentary debate.
Where choices and priorities are instead very clearly stated is on the role of the EU Agencies (Europol, Eurojust, EASO..) and bodies (the Anti Terrorism Coordinator) or where it is decided to go on with the establishment of an ambitious technocratic project such as the "smart borders" system (the feasibility of which is still to be proved even in the United States) or by creating an entry-exit system for third country nationals to control better the problems of the "over stayers" (those who remain after their initial permitted period of stay runs out) which apparently is one of the most dangerous threats to the EU. Needless to say that this idea is not new as it was raised by the US Congress years ago and  was considered “silly” in the US also by the former Homeland Security Secretary Chertoff under the BUSH administration.
Even worse, both the European Council draft Conclusions and the trio Presidency programme insist as one of their big priorities is the establishment of a "bona fide traveller" system which will discriminate between one traveller and another on the basis of de facto arbitrary criteria. They also reinstate their commitment to the creation of a European passenger name record (PNR) system.
These projects have in common the rather paranoic idea that any traveller is a potential danger. This is appalling in an European Union where there is still no permanent connection between the criminal records of the Member States, so that information on real criminals can be shared and where a terrorist (like the author of the attack to Jewish Museum in Brussels) can freely circulate even after been checked twice as a dangerous person on the Schengen information System.  
Selling out, for a false sense of security, the real fundamental rights of EU citizens, cannot be the real answer to the threats the EU will face in the coming years.
These inconsistencies can be solved by overcoming the 'silo' approach inside and between the MS and by better framing with a legislative measure the policy cooperation between the Member States (which still do not trust each other). True efficiency should then be measured if the threats are really supranational.
Even a project like  PNR could have its (crazy) logic if somewhere in Europe there were a central intelligence system which could filter these data against a massive intelligence analysis and profile, as happens in the USA. But as it has been designed, PNR will be only a policy laundering exercise where the European Union legislation is adopted to justify the collection of massive personal data at national level. Should we remember  that only on April 8th the Data retention directive, which followed the same logic, was annulled by the CJEU as a clear violation of the proportionality principle and of Articles 7 and 8 of the EU Charter?
The point is that selling out the personal data of EU citizens appears to the European Council less costly than building a real binding framework for police cooperation on the basis of Article 87 of the TFEU. The proof is given by the new Europol whose proposed legal basis (after amendments during negotiations) makes no more reference to Article 87 TFEU and which does not compel the Member States to share their security related informations.
These being the worrying projects on the European Council and Council side one can only hope that the newly elected European Parliament, in its July session, will challenge them and take the lead for a new alternative and legally sound policy which can shape in the next legislature an European area of Freedom, Security and Justice where the citizens' needs and not the administrations will be the real compass.

Barnard & Peers: chapter 25, chapter 26

Annex - draft Guidelines 

1.             One of the key objectives of the Union is to build an area of freedom, security and justice without internal borders, with full respect for fundamental rights. To this end, coherent policy measures need to be taken with respect to asylum, immigration, borders, police and judicial cooperation.

2.             All the dimensions of a Europe that protects its citizens and offers effective rights to people inside and outside the Union are interlinked. The success or failure in one field depends on the performance in the other fields as well as on synergies with related policy areas.  The answer to many of the challenges in the area of freedom, security and justice lies in relations with third countries, which calls for improving the link between the EU’s internal and external policies. This has to be reflected in the internal organisation of the EU institutions and bodies. Coordination with and within the Member States should be stepped up.

3.             Building on the past programmes, the overall priority is now to consistently transpose, effectively implement and consolidate the legal instruments and policy measures in place. Intensifying operational cooperation, enhancing the role of the different EU agencies and ensuring the strategic use of EU funds will be key. In further developing the area of freedom, security and justice over the next years, it will be crucial to ensure the protection of fundamental rights, including data protection, whilst addressing security concerns, also in relations with third countries, and to adopt a strong EU General Data Protection framework by 2015.

4.             Faced with challenges such as instability in many parts of the world as well as global demographic trends, an ageing population and skills shortages in Europe, the Union needs an efficient and well-managed migration and asylum policy. A comprehensive approach is required, optimizing the benefits of legal migration and offering protection to those in need while tackling irregular migration resolutely.

5.             To remain an attractive destination for talents and skills, Europe must compete in the global race for talent. Strategies to maximise the opportunities of legal migration should be developed, including the streamlining of existing rules and a dialogue with the business community. The Union should also support Member States' efforts for active integration policies which foster social cohesion and economic dynamism.

6.             The Union's commitment to international protection requires a strong European asylum policy based on the Treaty's principles of solidarity and responsibility. The full transposition and effective implementation of the Common European Asylum System (CEAS) is an absolute priority. This should result in high common standards and stronger cooperation, creating a level playing field where asylum seekers are given the same procedural guarantees and protection throughout the Union. It should go hand in hand with a reinforced role of the European Asylum Support Office (EASO), particularly in promoting the uniform application of the acquis. Converging practices will enhance mutual trust and allow to move to future next steps, including mutual recognition of asylum decisions.

7.             Addressing the root causes of irregular migration flows is an essential part of the EU migration policy. It is imperative to avoid the loss of lives of migrants undertaking hazardous journeys as well as to prevent and reduce irregular migration. A sustainable solution can only be found by intensifying cooperation with countries of origin and transit. Migration policies must become a much stronger integral part of the Union's external and development policies, applying the more for more principle and building on the Global Approach to Migration and Mobility. The focus should be on the following elements:
strengthening and expanding Regional Protection Programmes, in particular in the Horn of Africa, in close collaboration with UNHCR. In view of the protracted crisis in Syria, increase contributions to global resettlement efforts;
addressing smuggling and trafficking in human beings more forcefully, with a focus on priority countries and routes. Particular attention should go at present to the situation in Eritrea and the Sinai;
establishing an effective common return policy and enforcement of readmission agreements;
fully implementing the actions identified by the Task Force Mediterranean.

8.             The establishment of the Schengen zone, allowing people to travel without internal border controls, and the increasing numbers of people travelling to the EU require efficient management of the EU's external borders to ensure strong protection. This is in the first place the role of the Member States, which must fully take their responsibilities. At the same time the Union must mobilize all the tools at its disposal to support them in this task. To this end:
the integrated management of the external borders should be modernised to ensure smart border management with an entry-exit system and registered travellers programme and helped by the new Agency for Large Scale IT systems (EU-LISA);
Frontex, spearheading European solidarity in the area of border control, should reinforce its activities in terms of operational assistance and increase its reactivity towards rapid evolutions in migration flows, making full use of the new European Border Surveillance System EUROSUR;
the possibility of setting up a European System of Border Guards to enhance the control and surveillance capabilities at our external borders should be explored.
At the same time, the common visa policy needs to be modernised by facilitating legitimate travelling while maintaining a high level of security and implementing the new Schengen governance system.

9.             It is essential to guarantee a genuine area of security to European citizens by preventing and combatting organised crime, human trafficking and corruption. At the same time, an effective EU Counter terrorism policy is needed, whereby all relevant actors work closely together, integrating the internal and external aspects of the fight against terrorism. In this context, the European Council reaffirms the role of the EU Counter Terrorism Coordinator. In its fight against organised crime and terrorism, the Union should back the national authorities by mobilising all instruments of judicial and police cooperation, with a reinforced coordination role for Europol and Eurojust, including through:
the review of the internal security strategy;
the improvement of cross-border information exchanges, including on criminal records;
the development of a comprehensive approach to cybersecurity and cybercrime;
the prevention of radicalisation and extremism and addressing the phenomenon of foreign fighters, including through a legal instrument allowing for EU wide alerts.

10.         The smooth functioning of a true European area of justice with respect of the different legal systems and traditions of the Member States is vital for the EU. In this regard, mutual trust in each other's justice systems should be further enhanced. A sound European justice policy will contribute to economic growth by helping businesses and consumers to benefit from a reliable business environment within the internal market. Further action is required to:
promote the consistency and clarity of EU legislation for citizens and businesses;
-simplify access to justice; promote effective remedies and use of technological innovations including the use of e-justice; 
- examine the reinforcement of the rights of persons, notably vulnerable persons, in civil procedures to facilitate enforcement of judgements in family law and in civil and commercial matters;
- enhance mutual recognition of decisions and judgments in civil and criminal matters;
reinforce exchanges of information between the authorities of the Member States;
- fight fraudulent behaviour and damages to the EU budget by advancing negotiations on the European Public Prosecutor's Office;
- facilitate cross-border activities and operational cooperation;
- enhance training for practitioners;
- mobilise the expertise of relevant EU agencies such as Eurojust and the Fundamental Rights Agency (FRA).

11.         As one of the fundamental freedoms of the European Union, the right of EU citizens to move freely and reside and work in other Member States needs to be protected, including against possible abuse or fraudulent claims.

12.         The European Council calls on the EU institutions and the Member States to ensure the appropriate legislative and operational follow-up to these guidelines and will hold a mid-term review in 2017.