Sunday 21 December 2014

Could EU law save Paddington Bear? The CJEU develops a new type of protection


Steve Peers

Many readers may already be familiar with Colin Yeo’s brilliant blog post, imagining Paddington Bear’s unpleasant encounter with British immigration law. But could EU law save Paddington Bear, in particular by creating a new form of protection for those who have to flee their country of origin?


EU law has regulated in detail two forms of international protection: refugee status (based on the UN’s Geneva Convention) and subsidiary protection status, for those who don’t qualify for refugee status. (There’s also EU legislation providing for temporary protection, but the EU has never used it). But there are people who don’t qualify for either form of status. When do they have a right to stay?

This question is relevant in a number of cases. For instance, Paddington Bear could argue that he was an ‘environmental refugee’, fleeing catastrophic events in his country of origin that were not caused by humans (or even bears). (I’m converting Paddington into a human for the purpose of this blog post, just as Colin does in his). Others might lack a right to stay, but can’t be removed for practical reasons, for instance because they don’t have documents that prove their nationality, and so the country which they probably come from refuses to accept them back. Or they have a serious illness, but the developing country which they come from lacks the resources to treat it: so sending them back would, in practice, amount to a death sentence.

It’s the last of these scenarios which forms the subject of the CJEU’s two recent rulings in M’Bodj and Abdida. (Note that these judgments should be read in that order, and it would be highly misleading to read one but not the other). However, these judgments might well have implications for other groups of people as well – perhaps including even our marmalade-loving bear.


In the M’Bodj case, a non-EU citizen, having failed in his requests for asylum and stay on medical grounds in Belgium, was then the victim of a violent attack there. So he requested a disability allowance that was only available, on the facts of his case, if he could show that he was eligible for subsidiary protection. According to the EU’s qualification Directive, subsidiary protection must be granted if the applicant is facing any one of the following three situations: (a) the ‘death penalty or execution’; (b) ‘torture or other inhuman or degrading treatment or punishment of an applicant in the country of origin’; or (c) ‘serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict’. (The EU has more recently adopted a revised qualification Directive, but it didn’t change this particular definition).

The CJEU has previously interpreted the third of these grounds, in its judgments in Elgafaji and Diakite. But it has not yet interpreted the first two grounds. Mr. M’Bodj argued that the second ground applies, on the basis that the European Court of Human Rights (ECtHR) has interpreted Article 3 ECHR, which equally bans ‘torture or other inhuman or degrading treatment or punishment’, to mean that, in exceptional cases, people who would die if they were sent back to their country of origin, due to the inadequate medical treatment there, cannot be sent back.

However, the CJEU rejected its arguments. In its view, since the qualification Directive listed specific human activities as the source of persecution or serious harm, this form of ‘serious harm’ had to be the result of ‘a form of conduct on behalf of a third party’, so ‘cannot therefore simply be the result of general shortcomings in the health system of the country of origin’. This interpretation was bolstered by the preamble to the Directive, which says in effect that the Directive does not apply to those allowed to stay ‘on a discretionary basis on compassionate or humanitarian grounds’. It did make an exception for cases where the person concerned had been intentionally deprived of health care. Just because the ECtHR interpreted Article 3 ECHR to mean that people in Mr. M’Bodj’s situation could not be removed to their country of origin, that did not mean that they were necessarily entitled to subsidiary protection under the EU Directive.

The Directive does state that Member States can have ‘more favourable standards’ as regards qualifying for refugee or subsidiary protection status than those set out in the Directive. But such higher standards have to be ‘compatible’ with the Directive. In the Court’s view, it would indeed be incompatible with the Directive to give subsidiary protection status to Mr. M’Bodj, since there would be ‘no connection with the rationale of international protection’.

Read in isolation, the M’Bodj case means that those facing awful conditions in their country of origin that were not caused by humans have no right to refugee or subsidiary protection status. Their legal position is governed only by the ban on removal to their country of origin, set out in ECtHR case law, along with whatever limited benefits they may have under purely national law. There are a wide variety of such national law rules, as documented in a study for the European Migration Network. For Paddington Bear, this would mean that he could be sent back to his country of origin, unless the conditions there breached Article 3 ECHR standards. In that case, he could remain in the UK, but probably could not expect a very nice life.

However, the Abdida case, decided on the same day by the CJEU, suggests that in some cases, those who are unable to qualify for subsidiary protection or refugee status can use the EU’s Returns Directive to obtain a different type of protection. Mr. Abdida had similarly claimed that he should receive leave to remain on medical grounds in Belgium, and was appealing against the refusal of this application. While the appeal was pending, he was refused social assistance, on the grounds that it was only available to those whose challenge to their removal had suspensive effect; his challenge did not have that effect under Belgian law.

The national court asked the CJEU to interpret EU asylum legislation, and in light of its judgment in M’Bodj, the CJEU simply said that EU asylum law did not apply. While Member States have an option to apply EU legislation on reception conditions and asylum procedures to other types of applications for protection, Belgium has not chosen to do so.

The CJEU could have left the case there, but it didn’t. Instead, it decided to give the national court a detailed answer to questions it had never asked, on the interpretation of EU legislation that it had never referred to – the Returns Directive, which governs the process of removing irregular migrants from the territory.

First, the Court ruled on Mr. Abdida’s procedural rights. While the Returns Directive does not require legal challenges to removal to have suspensive effect, it was necessary to consider the impact of the EU Charter of Rights, Article 19(2) of which bans removals to States where the person concerned would face a serious risk of inhuman or degrading treatment. This had to be interpreted in light of the ECtHR case law on Article 3 ECHR discussed above, which bans removals on ‘medical grounds’ in exceptional cases.

How does this impact the interpretation of the Returns Directive? Article 5 of that Directive says that when they implement the Directive, ‘Member States shall take due account of: (a) the best interests of the child; (b) family life; (c) the state of health of the [person] concerned, and respect the principle of non-refoulement’. The CJEU ruled that Article 5, ‘taken in conjunction with Article 19(2) of the Charter’, means that Member States ‘may not…proceed with…removal’ where removal ‘would infringe the principle of non-refoulement’ on medical grounds.

This ban on removal had the consequential effect that the remedy against removal had to be suspensive, despite the optional wording of the Directive on this point, because otherwise Mr. Abdida could suffer irreparable harm if sent back to his country of origin before his appeal was decided.   

Secondly, the CJEU ruled on his social rights. Article 14 of the Directive says that except where irregular migrants are in detention, where they are given a time for voluntary departure or their removal is postponed, Member States must ‘ensure that the following principles are taken into account as far as possible…: (a) family unity with family members present in their territory is maintained; (b) emergency health care and essential treatment of illness are provided; (c) minors are granted access to the basic education system subject to the length of their stay; and (d) special needs of vulnerable persons are taken into account’.

In the Court’s view, Mr. Abdida qualified for this treatment because his removal had to be postponed under the Directive, which requires postponement where suspensive effect of an appeal has been granted. Oddly, the Court did not mention that the Directive also requires postponement where removal would violate the principle of non-refoulement, although this rule was obviously relevant to Mr. Abdida as well.

It should be noted that the Directive gives an option to Member States to postpone removal in other cases, too. On that point it specifies that ‘Member States shall in particular take into account: (a) the third-country national’s physical state or mental capacity;’ and ‘(b) technical reasons, such as lack of transport capacity, or failure of the removal due to lack of identification’. In these cases, the persons concerned could also invoke the safeguards set out in Article 14, since that clause applies regardless of the reasons for postponement of removal.

Yet on the face of it, Mr. Abdida’s specific claim for social assistance could not be successful, since such assistance is not mentioned in Article 14 at all. Indeed, the preamble to the Directive states that pending return, the ‘basic conditions of subsistence should be defined according to national legislation’. But the CJEU ruled that such legislation still had to be ‘compatible with the requirements laid down in’ the Directive. In this case, the right to the provision of health care would be ‘rendered meaningless if there were not also a concomitant requirement to make provision for the basic needs’ of the person concerned. However, that right only had to be provided ‘as far as possible’, on the condition that the person lacked the means to provide for his own needs; and it was up to Member States to ‘to determine the form’ which the provision of basic needs took.



At the outset, it should be noted that the qualification Directive and the Returns Directive have a different territorial scope. The former applies to all Member States except Denmark, but including the UK and Ireland. The latter doesn’t apply to the UK and Ireland, and it applies to Denmark and the non-EU Schengen associates (Norway, Iceland, Switzerland and Liechtenstein) only in certain cases.

So, in light of this recent CJEU case law, my advice to Paddington Bear would be to take the tube a few stops from Paddington to King’s Cross & St. Pancras station, and stow away again, this time on a Eurostar train to Brussels. (My apologies for facilitating a breach of immigration law, thereby infringing the EU’s Directive on that subject. And even greater apologies for inflicting upon the world the consequential change of name: it will be much harder to love ‘Gare du Midi Bear’.) Or alternatively, he could stow away on a train to Disneyland Paris – and create work for an army of copyright lawyers.

The Court’s judgments raise questions about the scope of international protection as defined by the qualification Directive, and about the new possibilities of protection arising under the Returns Directive. For the sake of brevity, let’s call that latter concept ‘alternative protection’ (technically, we should probably call it ‘Returns Directive protection’, but I can’t bring myself to use that term).

First of all then, the qualification Directive. The central point here is that there is no right to subsidiary protection merely due to the lack of decent health care in the country of origin, unless access to that health care has been blocked. The Court does not expand further on that exception, but presumably a parallel argument could be made for refugee status, where access to essential health care is blocked on one of the grounds (such as race, religion or political opinion) set out in the Geneva Convention, since that would surely constitute a form of persecution.

More broadly, the Court rules out a subsidiary protection claim unless the conditions in the country of origin have been caused by a ‘third party’ as defined in the Directive (namely the State, persons or groups controlling the State, or non-State actors). While the Court didn’t rule on this point as such, the logic of its judgment suggests that ‘non-State actors’ have to be humans. Also, it’s implicit that indirect causation isn’t sufficient. It could certainly be argued that the lack of sufficient health care in developing countries is ultimately the fault of corrupt dictators and elites who hoard those countries’ wealth, and/or the legacy of (neo-)colonialism and slavery. Equally, many believe that environmental displacement results from climate change caused by human activity. But it seems unlikely that the CJEU would accept this line of argument.

Furthermore, the Court has ruled out Member States using their power to apply more favourable rules for international protection in such cases. The CJEU had only ruled once before on this issue, in the case of B and D, when it ruled that States could not grant refugee status to persons who were excluded from that status. It was possible to interpret that judgment as meaning that Member States were only constrained from granting more favourable treatment when international law required them not to do so. But the M’Bodj judgment goes much further than this. So the CJEU hasn’t just refused to raise the floor for international protection – it’s significantly lowered the ceiling.

It should be noted that the requirement that more favourable standards be compatible with EU law applies in all other EU asylum laws, as well as in some EU migration legislation, such as the Returns Directive. The M’Bodj judgment might therefore have implications for these laws too. But the rule doesn’t apply to all EU immigration laws, and there’s no good reason to assume that a ‘compatibility’ requirement ought to apply in those cases where the EU legislature has deliberately decided not to insert one.

So what more favourable standards can Member States apply? In the case of the qualification Directive, the Court says that they must fall within the scope of international protection. It doesn’t elaborate further, but this presumably refers back to the key concept of persecution or serious harm caused by a ‘third party’. Higher standards can therefore apply only in relation to the Directive’s definition of ‘refugee’ and ‘subsidiary protection’. For instance, it should surely be open to Member States to grant subsidiary protection to persons fleeing indiscriminate violence even where the threat is not ‘individual’, or where the applicants are not civilians.

Secondly, what are the implications of the Abdida judgment for the new possibility of ‘alternative protection’? To understand this concept better, it should be compared to the elements of EU law governing international protection: the allocation of responsibility for an application (ie, the ‘Dublin rules’); the substantive grounds for protection; the procedures for determining whether there is a protection need; and the benefits which an individual receives during the determination process (and afterward, if his or her claim is successful).

Beforehand, though, some general points about the Court’s approach to the Returns Directive in Abdida.  When it was initially adopted back in 2008, that Directive was castigated by NGOs as the Devil incarnate, spurring on Member States to all sorts of inhumane treatment of irregular migrants. Over the years, the CJEU’s interpretation of the Directive has (generally speaking) ensured that irregular migrants are better treated than the Directive’s initial critics thought that they would be. The judgment in Abdida goes even beyond this, and transforms an instrument of repression into (in some cases) an instrument for protection. It’s like going into a vegetarian restaurant, and being served a bacon sandwich.

For meat-eaters – with the obvious exception of Ed Miliband – that would be a pleasant surprise. So what’s the problem? The problem is that it’s difficult to take a law that was drafted to expedite the removal of migrants, and using it to help them instead. The CJEU tried to overcome this in the Abdida judgment by performing a series of feats of legal alchemy: redrafting the national court’s questions from scratch; turning an obligation to ‘take due account’ of non-refoulement into a ban on removal; making an option to grant the suspensive effect of appeals into an obligation; transforming the ‘principle’ of obtaining health care into a rule; and adding social assistance to the list of benefits which Mr. Abdida has to receive.  Yet despite all this, a lot of effort will still be required to make the Returns Directive function as an instrument to grant individual protection.

To see why, let’s look at those elements of the protection process, starting with the allocation of responsibility for an application. There are no ‘Dublin rules’ applicable to alternative protection cases. So if Paddington Bear managed to make his way to Brussels or Paris, Belgium or France could not invoke formal rules requiring the UK to take him back. The position would be different though, if he had applied for asylum in the UK first, before making his way to another Member State. In that case, the Dublin rules would apply, so in principle the UK would have responsibility for him, since he crossed the UK borders without authorisation before he set foot in any other Member State. But this normal rule would have to be disapplied in his case, since unaccompanied minors can usually change the country which is responsible for their application (see the discussion here).

In fact, in real life, most of those using the ‘alternative protection’ route would likely have applied for asylum first (as is evident from the facts of the Abdida and M’Bodj cases), so the Dublin rules as such would apply to them. According to the CJEU, they could therefore only resist being sent back to the responsible Member State if there is a complete breakdown of the asylum system there. For Paddington, the demented attentions of Nicole Kidman’s character, or the grim reality of British immigration law as described in Colin Yeo’s blog post, would not meet that threshold.

Next, what are the substantive grounds for alternative protection? We know from the Abdida case that they include at least non-refoulement, and crucially also that the concept of non-refoulement in the Returns Directive and the Charter is wider than the concept of international protection. It covers at least ‘medical cases’, and arguably also other cases that fall outside the scope of refugee and subsidiary protection status but within the scope of Article 3 ECHR (for instance, persons excluded from refugee status or subsidiary protection status). It could also be argued that it covers environmental cases, such as the fictional destruction of talking bears’ habitat, or the very real scenario of sinking islands. Article 5 of the Returns Directive also refers to the best interests of the child, the state of health of the person concerned and family life, so logically these could also potentially be grounds for alternative protection. But it’s not possible to claim alternative protection on the basis only that it’s impossible to carry out a return decision in practice: se the Mahdi decision earlier this year (discussed here).

Moving on to the procedures for determining whether there is an alternative protection need, there are three elements here: access to the process; administrative procedure; and judicial protection. There’s an obvious problem with access to the process, for the Returns Directive cannot be ‘accessed’ unless there is a pending returns decision. So we have the odd scenario where people facing expulsion (Mr. Abdida) are better off in terms of health care than those with residence permits (Mr. M’Bodj). The Returns Directive requires Member States to issue return decisions to those whose stay is not authorised, but it leaves them an option to allow people to stay if they wish; in that case, the return decisions must be rescinded. Otherwise, the Directive is silent on whether return decisions expire or can be renewed or rescinded. There’s no provision for an individual to ask for a return decision to be issued, even where it’s possible that it might make the person concerned better off.

Next, the administrative procedure during the returns process has been addressed in the CJEU recent judgments on the right to be heard in Returns Directive cases, discussed here. It’s notable that the judgment in Boudjlida specifically requires national authorities, before issuing a returns decision, to consider any issues related to Article 5 of the Returns Directive. However, on the whole the concept of the right to be heard in Returns Directive cases, as developed by the Court is significantly weaker than the rules on interviews, et al, set out in the asylum procedures directive.

Conversely, as for judicial protection, the suspensive effect of removals set out in the Abdida case is stronger than the equivalent provisions of the asylum procedures Directive. It’s not clear if this would apply to other cases where an ‘alternative protection’ argument might be made though. Article 13 of the Returns Directive also contains other rules, on: the independence and impartiality of the appeals body (which actually need not be a judicial body); requirement for a merits review; access to legal advice; and legal aid. It might be arguable by analogy with the suspensive effect point in Abdida, that at least in some cases, Article 47 of the Charter confers stronger rights than the Directive requires.

Finally, the benefits which an individual receives during the determination process, if his or her removal is postponed, are (as noted above) family unity, essential health care, education for minors, protection for vulnerable persons, and written confirmation of their status. We know from the Abdida judgment that this list of entitlements is not exhaustive, and can be extended to protection of basic needs in individual cases. In that case, the CJEU implicitly accepted that there would be no point providing health care for Mr. Abdida if, while he received it, he was being left to starve to death in the streets.  This is surely not the only type of case where basic needs have to be provided for: should an unaccompanied minor be left without food and shelter also?

What about the two provisos mentioned by the Court: a means test, and Member States’ power to ‘determine the form’ which provision of basic needs take? On the first point, it should be noted that the employer sanctions Directive (discussed here) gives Member States an option – but not an obligation – to permit irregular migrants to take up employment. Possibly Member States could take into account other income being provided by families, friends or charities. On the second point, Member States’ discretion does not extend to an absolute power to detain the people concerned, since the grounds for detention and conditions for detention set out in the Returns Directive would apply. But as regards other aspects of basic needs, Member States would surely be free to decide on the details of housing and financial support. So sadly, Paddington Bear could not insist on staying with Hugh Bonneville’s family – and would have to settle for marmalade from Aldi, not Marks & Spencer’s.


Barnard & Peers: chapter 9, chapter 26


  1. Brilliant post, very informative, thanks so much.

  2. Fascinating take on Paddington. Should be made into a case study.