Wednesday 9 March 2016

Detention of asylum-seekers: the first CJEU judgment

Steve Peers

One of the most controversial aspects of immigration and asylum law is the detention of migrants: people who have broken no criminal law (other than, possibly a criminal law about migration control) but who are detained during their asylum application, or pending their removal from the country. The EU has had rules on detention of irregular migrants for some time, in the Returns Directive (on the CJEU’s interpretation of those rules, see my journal article here).  But it has only recently had rules on the detention of asylum-seekers, in the second-phase Directive on the reception conditions for asylum-seekers. (The UK and Ireland have opted out of both Directives).

Recently, in the JN judgment, the CJEU ruled for the first time on the interpretation of these new rules. In fact, this was the Court’s very first judgment on any aspect of the second-phase legislation, although it soon gave another judgment (on the rights of people with subsidiary protection, discussed here), and other cases are pending. The Court’s ruling addresses a number of key questions of interpretation of the detention rules, but left a number of issues open.

In general, the Court has limited the prospect of detaining asylum-seekers on grounds of ‘national security or public order’, and its ruling implicitly somewhat constrains the possibilities of detaining asylum-seekers on other grounds too. But in parallel to that, the judgment strengthens the rules in the Returns Directive on the detention and expulsion of irregular migrants. And the Court’s ruling is surprisingly open to the application of human rights ‘soft law’ as a means of interpreting EU law. Overall, while not mentioning the current ‘refugee crisis’, the judgment is an implied rebuff to those who would like to resort to extensive detention of asylum-seekers as a means to address that crisis.


The first phase reception conditions Directive (adopted in 2003, applicable from 2005) said little about detention of asylum-seekers. While the subsequent Returns Directive did regulate detention of irregular migrants, the CJEU made clear in Kadzoev and Arslan that those rules did not apply to asylum-seekers, because EU asylum legislation gives asylum-seekers the right to stay on the territory until a decision is made at first instance on their application, whereas the Returns Directive says that irregular migrants should be booted out as soon as possible. In Arslan, the Court clarified the relationship between the two sets of rules: an irregular migrant detained under the Returns Directive could not simply escape from detention by applying for asylum. Essentially the JN judgment returns to the same issue, and asks the Court to reconsider its position in light of the more detailed rules on detaining asylum-seekers which now apply.

So what are those rules? In the second-phase reception conditions Directive, the previous ban on detaining people solely because they have applied for asylum is retained. The Directive then provides generally for detention of asylum-seekers if ‘necessary’ after ‘an individual assessment of each case…if other less coercive alternative measures cannot be applied effectively’. Detention is permitted ‘only’ on six grounds: (a) ‘in order to determine or verify [an asylum-seeker’s] identity or nationality’; (b) to ‘determine the elements on which’ the application is based ‘which could not be obtained in the absence of detention, in particular where there is a risk of absconding’; (c) in order to decide on entry onto the territory; (d) when the asylum-seeker is detained pursuant to a planned expulsion under the Returns Directive, and there are objective grounds to show that he or she applied for asylum only to ‘delay or frustrate’ expulsion, despite having had an opportunity to access the asylum procedure; (e) ‘when protection of national security or public order so requires’; or (f) in accordance with the Dublin III Regulation on allocation of asylum-seekers between Member States, which provides for detention if there is a ‘significant risk of absconding’ before a Dublin transfer is carried out.

The grounds for detention must be ‘laid down in national law’, which must also lay down rules on alternatives to detention. There are detailed rules on procedural guarantees as regards detention, and on the conditions of detention. Those procedural guarantees and detention condition rules also apply to Dublin cases, and the Dublin Regulation moreover sets out precise rules on the length of detention. The CJEU has been asked to interpret the ground for detention in the Dublin III Regulation, in the pending Al Chodor case.

The judgment

Mr JN had made three prior applications for asylum. They were all unsuccessful, but nevertheless he was not removed from Dutch territory. Over a period of 20 years, he accrued more than twenty convictions for criminal offences. The case did not concern detention for those criminal convictions, as such detention falls outside the scope of the Directive (unless, arguably, the criminal conviction is related to immigration offences: more on that point below). Rather it concerned detention on grounds of ‘public order and national security’, which the Dutch government imposed in light of his criminal offences – but not as a penalty for them.

Obviously such detention is compatible in principle with the Directive, which expressly provides for detention on such grounds. So Mr. JN instead argued that the relevant provision in the Directive itself was invalid. It should be noted that another pending case asks the CJEU whether two other grounds for detention in the Directive are invalid: verification of identity or nationality, and determining the elements on which the application is based.

The Court began its analysis by reiterating its prior case law that the European Convention on Human Rights (ECHR), which includes rules on detention, does not bind the EU as such. Instead, it assessed the validity of the clause in the Directive in light of the EU Charter of Fundamental Rights – although this did entail some assessment of the validity of that clause in light of the ECHR as referred to in the Charter, as discussed below.

According to the Court, detention undoubtedly affects the liberty of the individual, as guaranteed by Article 6 of the Charter. So the question is whether this particular ground for detention was justified, in light of the general test for limiting Charter rights set out in Article 52(1) of the Charter. This test requires that limitations on Charter rights must: (a) be prescribed by law; (b) not infringe the essence of the right; (c) be aimed at protecting an objective of general interest, or the rights and freedoms of others; and (d) be proportionate – meaning that they are appropriate and necessary to achieve their objective.

Applying these tests, the Court first found that the possibility of detention on grounds of public policy or national security was prescribed by law, since it was set out in the Directive. It did not infringe the essence of the right to liberty, since it was based on individual conduct and applied in ‘exceptional circumstances’, circumscribed by the various general limits and guarantees relating to detention set out in the Directive. Detention on grounds of public order and national security meets a public interest, and moreover protects the right to ‘security’ of others.

The Court’s most detailed reasoning therefore concerned proportionality. Detention on public order or national security grounds was inherently ‘appropriate’ to the objective of ensuring public protection. It was ‘necessary’ for a number of reasons, which the Court elaborated in some detail. All restrictions on liberty have to be ‘strictly necessary’ and this particular ground to detain was ‘strictly circumscribed’ by the overall legal framework: detention on such grounds had to be ‘require[d]’; detention must be provided for in national law; the general limits and safeguards on detention in the Directive apply; the exception is limited by international human rights ‘soft law’; and the concepts of ‘public policy’ and ‘national security’ had to be narrowly interpreted.

The Court explored the latter two points further. As regards international human rights ‘soft law’, it noted that in the Commission’s original proposal for the Directive, it referred to a Council of Europe Recommendation on detention of asylum-seekers, as well as UNHCR guidelines on detention. It then applied some of the text of the latter guidelines: in particular detention of asylum-seekers must occur ‘only exceptionally’ in an ‘individual case’ as a ‘last resort’, where ‘necessary, reasonable and proportionate to a legitimate purpose’.

As for the detention grounds of ‘public policy’ and ‘national security’, the Court applied last year’s judgments in T and Zh and O (discussed here and here), in which it had ruled that ‘public policy’ exceptions in other EU immigration and asylum legislation had to be narrowly interpreted, consistently with the narrow definition of that exception in EU free movement law. In JN it said the same for the national security exception as regards detaining asylum-seekers, and furthermore as regards the grounds for entry bans longer than five years, as set out in the Returns Directive. So the exceptions apply only where there is a genuine criminal or security threat, not where there the authorities simply deem it expedient to detain people.

Next, the Court applied its interpretation of the Directive to the facts of this case. He was detained due to his prior offences and pending expulsion order, which was attached to a ten-year entry ban. Since entry bans for longer than five years can only be issued on grounds of a ‘serious threat to public policy…public security or national security’ it followed that detention could be ordered in the same circumstances – as long as proportionality was ‘strictly observed’ and those reasons are still valid.

The Court also made clear that the pending expulsion order could not lapse during consideration of JN’s asylum application. The national case law which provided for it to lapse had to be disapplied by the national court, in order to ensure the effectiveness of the Returns Directive (ie the expulsion of irregular migrants). So Mr JN was a sort of ‘Schrodinger’s migrant’: allowed to stay on the territory while his asylum application was considered (and so subject to the detention rules in the reception conditions Directive); but also simultaneously subject to an expulsion order under the Returns Directive, which was only temporarily suspended – and which continued to justify (in part) his detention under the formally distinct set of asylum rules.

Finally, the Court concluded by looking at the position under the ECHR, in the context of the Charter. The former was relevant to the latter because Article 52(3) of the Charter says that the ‘meaning and scope’ of Charter rights which ‘correspond’ to ECHR rights is the same as those ECHR rights.  However, the Court easily dismissed the ECHR argument by pointing out that in the recent judgment of the European Court of Human Rights (ECtHR) in Nabil v Hungary, an asylum seeker could still be detained pursuant to Article 5(1)(f) of the Convention (which allows detention ‘where action is being taken with a view to deportation’) because rejection of an asylum application would entail enforcement of an expulsion order. But the Court did refer to the safeguards in other ECtHR case law: there can be ‘no element of bad faith or deception by the authorities’, and detention must be proportionate.


As the CJEU’s first proper judgment on detention of asylum-seekers, the JN ruling may become seminal. That’s not because of the facts of this particular case: with three failed asylum applications and over twenty criminal convictions to his name, the grounds to detain Mr JN are stronger than they are for many other asylum-seekers. But much of what the Court said in its judgment has potentially wider impact.

I will analyse that possible impact from five angles: (a) the ‘public policy and national security’ ground of detention; (b) the application of other grounds for detention; (c) the Court’s use of ‘soft’ human rights law; (d) the interpretation of the Returns Directive; and (e) the role of law in the development of the EU’s asylum regime more generally.

The ‘public policy and national security’ ground of detention

The Court made clear that the public policy and national security grounds for detention must be narrowly interpreted, and interpreted consistently with EU free movement law. Mr JN’s detention was justified because of his prior criminal offences in conjunction with the underlying expulsion order. But are these tests cumulative or alternative? And are they exhaustive?

The Court does not address these questions. However, the requirement to interpret these grounds consistently with EU free movement law suggests that the two tests are exhaustive. Arguably criminal offences alone could justify detention, in light of the nature of this ground for detention. But the principle of proportionality must mean that detention would be harder to justify in the absence of an expulsion decision, and that the seriousness and number of the offences are also highly relevant. (Remember that detention under the Directive is distinct from detention ordered as a result of a criminal conviction, or pre-trial detention linked to the criminal proceedings).

Conversely, it seems unlikely that an expulsion decision alone could justify detention on this ground. If that were permitted, it would be too easy for Member States to justify the detention of almost all asylum-seekers, by issuing irregular migrants with expulsion orders as soon as they are apprehended, before they can apply for asylum. This would undercut the Court’s emphasis on the exceptional nature of detention of asylum-seekers. Furthermore, the Directive has a lex specialis on detaining asylum-seekers who had been subject to expulsion orders: the ‘last-minute application’ clause. If the drafters of the Directive had intended a broader possibility to detain asylum-seekers merely because they were subject to expulsion orders, they would have drafted that clause differently.

Other grounds for detention

While most of the JN judgment focusses on the particular ‘public policy and national security’ ground for detention of asylum-seekers, some of the Court’s reasoning casts light by analogy on the validity and interpretation of the other five detention grounds.

First of all, each of the other five grounds for detention of asylum-seekers restricts their liberty, so must be also justified under Article 52(1) of the Charter. Applying the Court’s analysis in the JN judgment by analogy, each of those other five grounds is ‘prescribed by law’, at least according to the CJEU’s approach to that concept. However, the other grounds are not so closely linked to individual conduct of the person concerned, although arguably the ‘last-minute application’ and Dublin III ‘serious risk of absconding’ ground have a closer link than the others (the Dublin III Regulation refers to ‘reasons in an individual case’ to suggest that an asylum-seeker may abscond). Nor is it clear how the ‘exceptional circumstances’ concept applies to the other grounds, although they are all also subject to the general limits and guarantees relating to detention set out in the Directive.

The public interest arguments for the other grounds of detention are less obvious, although the Court could probably find them: the efficiency of the asylum system, and (as regards the entry control and ‘last-minute application’ grounds) immigration control (see the Schwarz judgment by analogy). But the restrictions on liberty are not so obviously appropriate as is the case for public policy and national security (except as regards the ‘last-minute application’ clause, provided that there was an effective opportunity to apply for asylum).

As for necessity, the Court applied the ‘strictly necessary’ rule to all deprivations of liberty. Furthermore, the other grounds for detention are also subject to the general limits and safeguards set in the Directive, and the rule that detention must be provided for in national law. However, not all of the specific features which the Court discussed in JN apply to the other grounds for detention: there is no obligation that detention on the other grounds be ‘required’, and the interpretation of those other grounds under EU law and international human rights soft law will necessarily be different. That brings us neatly to the Court’s innovative use of that soft law.

The Court’s use of ‘soft’ human rights law

First of all, the Court’s use of international human rights ‘soft law’ is remarkable in itself. It’s only taken account of such rules once before in the immigration and asylum context: the El Dridl case, where the preamble to the Returns Directive referred to a Council of Europe Recommendation on detention of irregular migrants. But in JN, the explanatory memorandum to the original proposal is enough to trigger incorporation of the soft law into the Court’s interpretation of the Directive.

It’s not clear if this may have broader implications beyond the reception conditions Directive. I’ve checked the original proposals for the other second-phase asylum laws, and none of them refer to international soft law as far as I can see. (But note that the preambles to the legislation do refer to the Geneva Convention on refugee status, as well as the UN Convention on the Rights of the Child).

However, it does have a number of implications for the interpretation of the reception conditions Directive. I have read through both ‘soft law’ measures invoked by the Court, and noted some key points where they could be useful in interpreting the Directive. For the sake of readability, I have put some of the detail in an Annex to this blog post. But here are the highlights.

The soft law gives more precise explanations for detention on grounds of determining nationality or identity, or to determine elements of the claim. A crucial point here is a detailed interpretation of Article 31 of the Geneva Convention, which states that refugees ‘coming directly’ from persecution cannot be penalised for irregular entry if they breach immigration law for ‘good cause’ and contact the host State’s authorities ‘without delay’. This is a big issue in practice.  

Two years ago, in its judgment in Qurbani (discussed here), the CJEU said it would not interpret Article 31 of the Convention, unless EU legislation referred to it. Well, the preamble to the second-phase reception conditions Directive does refer to it, in the context of detention (so does the Dublin III Regulation, and EU anti-smuggling law). Since criminal penalties for irregular entry are affected by the Returns Directive (according to the CJEU’s case law), it must follow that they are also affected by EU asylum law, a fortiori because the EU legislators expressly refer to Article 31 of the Convention. 

What does this mean on the ground? While the JN judgment did not concern Article 31, the CJEU's willingness to apply the UNHCR guidelines on detention means that those guidelines could be relevant to its interpretation. First of all, those guidelines say that asylum-seekers are covered by Article 31, even if their refugee status has not yet been established. Next, the 1999 version of the guidelines say that ‘coming directly’ also covers cases where asylum-seekers travelled through other countries. The ‘good cause’ rule must be interpreted in context, and there is no strict time limit for contacting the authorities. Between them, these interpretations of Article 31 should limit asylum-seekers’ criminal convictions for irregular entry considerably. In any event, EU legislation and case law says that asylum-seekers are entitled to stay on the territory and are outside the scope of the Returns Directive since they cannot be considered irregular; subjecting them to a criminal prosecution for irregular entry would directly contradict this.

Furthermore, the soft law is relevant not only to the grounds for detention, but also alternatives to detention, judicial review of and the conditions for detention. On that latter point, it mentions the practice of religion in detention, as well as a broader measure of contact with the outside world. Asylum-seekers should have a complaints procedure concerning detention conditions. There are more details on detention of vulnerable persons.

On that point, I can never pass on an opportunity to comment on the quite obnoxious derogations permitted in the Directive, allowing Member States to waive the requirements for separate accommodation for detained families and detaining women separately from unrelated men, in ‘duly justified’ cases at the border. In light of the Charter rights to privacy, the rights of the child and the EU’s imminent signature of the Council of Europe Convention on violence against women (on which, see here), these derogations are surely either invalid or can only apply in cases of force majeure.

Interpretation of the Returns Directive

The Returns Directive says nothing explicitly on the lapse of return decisions. This judgment is the first time the CJEU has ruled on the issue. While the Court only addresses the specific point of return decisions lapsing due to an asylum application, it might be argued by analogy that the lapse of return decisions in other circumstances is also incompatible with the Returns Directive. Although Member States are allowed to set higher standards than the Returns Directive, that only applies if those standards are still ‘compatible’ with the Directive. As we saw in the Zaizoune judgment (discussed here), such higher standards cannot amount to a waiver of the obligation to return people. It’s implicit in the JN ruling that equally it’s not compatible with the Directive for return decisions to lapse as soon as an asylum application is made.

The role of law in the development of the EU’s asylum regime

The JN ruling came as the EU took further measures to reduce the numbers coming to or staying on the territory – most notably by reaching a controversial arrangement with Turkey (on which, see here). Overall, the judgment sends a clear signal that the CJEU is going to assert its legal authority to ensure that measures taken to deal with the refugee and migration crisis are compatible with human rights, in particular as regards asylum-seekers – although conversely the Court is keen to strengthen the obligation to expel those who have not established any need to stay.

More broadly, the EU’s refugee policies are obviously in a state of deep crisis. Rather than leave the issue entirely to populists at the EU or national level, it would be better for the EU ask a panel of respected international experts to recommend (quickly) how the EU, in the wider international context, should deal with the crisis. I would nominate (say) Mary Robinson, David Miliband, Madeline Albright and Carl Bildt for this task. In any event, we cannot go on as we are: the EU needs an asylum policy that is simultaneously fair, humane, realistic and coherent; but it is falling far short of that at the moment.  

Barnard & Peers: chapter 26
JHA4: chapter I:5
Photo credit: UNHCR, B. Szandelszky


Council of Europe Committee of Ministers Recommendation

Point 1 – the definition of ‘detention’ is taken implicitly from ECHR case law, and is more precise than in the Directive

Point 3 – a general provision says ‘the aim [of detention] is not to penalise asylum-seekers’. The ground of detention to determine nationality or identity is explained in more detail than under the Directive. It applies ‘in particular when asylum seekers have destroyed their travel or identity documents or used fraudulent documents in order to mislead the authorities of the host state’. The ground of detention to determine elements of the asylum claim is less detailed than under the Directive, which contains the following additional words: ‘in particular where there is a risk of absconding of the applicant’.

There is no parallel to two of the grounds for detention under the Directive: under the Dublin process (ie where there is a ‘significant risk of absconding’ during that process); and where there is an asylum application purely to forestall an expulsion decision, if the asylum-seeker had previously had an opportunity to apply for asylum. 

Point 4 – says there must be a ‘careful’ examination of the grounds for detention in individual cases, and detention shall be ‘non-arbitrary’.

Point 5 – discusses grounds for judicial review, which are not expressly mentioned in the Directive. If a maximum detention duration has not been provided for by law, the duration of the detention should form part of the review by the above-mentioned court (see the Mahdi judgment on the Returns Directive by analogy).

Point 6 – ‘Alternative and non-custodial measures…should be considered before resorting to measures of detention’. The Directive does not state this expressly.

Point 7 - Measures of detention should not constitute an obstacle to asylum seekers being able to submit and pursue their application for asylum.
Point 8 - Asylum applications from persons in detention should be prioritized for the purposes of processing. This is especially the case where a person is held in detention because of reasons resulting from the law pertaining to foreigners.
Conditions of detention
Point 15 - Detained asylum seekers should be allowed to practice their religion and to observe any special diet in accordance with their religion.
18. Asylum seekers should be allowed to contact and, wherever possible, receive visits from relatives, friends, social and religious counsellors, non-governmental organisations active in the field of human rights or in the protection of refugees or asylum seekers, and to establish communication with the outside world. Note that this is wider than Article 10(4) of the Directive.
19. Asylum seekers should be guaranteed access to a complaints mechanism concerning the conditions of detention. This issue is not mentioned in the Directive.
22. If minors are detained, they must not be held under prison-like conditions…If [placing outside detention] proves impossible, special arrangements must be made.
UNHCR guidelines
Guideline 2 - Interprets Article 31 of Geneva Convention – also referred to in preamble to the Directive. Refers also to UNHCR Executive Committee conclusions for more on when detention is ‘necessary’ under Article 31 of the Convention
Article 31 applies also to asylum-seekers, not just recognised refugees; 1999 version of conclusions: (point 4) ‘coming directly’ clause also covers cases where the asylum seeker transited through other States on way to State where they are now present. No strict time limit to the phrase ‘without delay’. ‘Good cause’ – must look at all the circumstances
Guideline 3 – must consider alternatives to detention first – same as in CoE recommendation.
Grounds for detention (i) to prevent absconding (matches Dublin III Regulation to some extent); (ii) manifestly unfounded or abusive claims (no match with Directive); (iii) to verify identity or security; no reference to nationality (so not as complete a correspondence as CJEU suggests); (iv) elements of the claim – explained in detail ‘within the context of a preliminary interview' (with further clarification); (v) public health (no match in the Directive); (vi) national security; or (vii) a 'last minute' application to frustrate expulsion (no match in the Directive) 
Point (d) of Article 8(3) of the Directive doesn’t appear here; ‘procedure to enter the territory’ does not apply.
General rule – cannot use detention as a deterrent, or to dissuade continuing with claims; not punitive or disciplinary, or for breach of rules at reception centres or camps.
Guideline 4.3 – more detailed rules on alternatives to detention than in Article 8(4) of the Directive.
Guideline 5 - detention cannot be discriminatory
Guideline 6 - there must be time limits on detention
Guideline 7(iv) – right of asylum-seeker or lawyer to attend hearing re review of detention; 7(v) – authorities have burden of proof re detention; 7(vi) not an obstacle to pursue the asylum application (as in CoE recommendation).  
Guideline 8 – like CoE Recommendation: religious diet, wider access to outside world; more details on basic necessities than in Directive (ie ‘dignity’); no prison uniforms or shackling; also refers to complaints procedure (like CoE Recommendation) but goes into further detail than that Recommendation
Guideline 9 – more details on vulnerable persons than in Art 11(1) of the Directive


  1. Dear Professor Peers,
    I am a student from the UK currently volunteering in Greece with the refugees currently stuck in Athens and elsewhere since the Greek borders closed. I was previously working outside Moria refugee camp on Lesvos which had been open to refugees to come and go until the EU-turkey deal, after which the camp became a detention centre for all current and new refugee arrivals. It is extremely unclear as to what the plan is to do with refugees (now effectively detainees) who are detained in Moria camp now. It appears they are taken there before being registered and sent to other camps (at first they were shipped to Kavala on the mainland but this and most other government run camps are now full) before being deported back to Turkey. It is unclear but seems very unlikely that they are given proper opportunity to seek asylum. Moria camp when I left to volunteer with refugees in Athens a few days ago had become a camp the detainees are not permitted to leave and no-one is allowed in except workers with specific NGOs and UNHCR workers. Barbed wire was put up on the fences about a week ago. I cannot confirm at all what the conditions are like inside but there have been reports from refugees inside that they have been denied food and water, have to wait hours to use the toilet and sleep on hard floors. I cannot say whether any of this is true and personally find it hard to believe but this is what the rumours are. Refugees have attempted to board the ferry to athens with fake papers out of desperation and are arrested and placed straight into Moria camp. In light of the EU-turkey deal I would very much appreciate any legal clarification you can provide as to whether such refugee detention is at all legal, and what if any, the chances are of highlighting its illegality and changing the situation to ensure asylum-seekers can indeed seek asylum before being treated as criminals.
    Thank you for the analysis you have so far provided on this blog and I await your reply eagerly!

    1. If people have applied for asylum detention may be legal, but only on the conditions set out in the reception conditions directive which the CJEU has so far interpreted only in this case (alongside the international soft law I referred to). For people who haven't applied for asylum the Returns Directive applies; it sets similar standards as regards conditions for detention but the grounds for detaining people are somewhat different. It would be hard to say more than that where the circumstances probably differ in individual cases. Certainly refusal of food, water and toilets would breach the right to dignity as guaranteed by the EU Charter, and the asylum procedures Directive requires that there must be an effective opportunity to claim asylum.