Tarakhel v Switzerland: Another nail in the coffin of the Dublin system?

 

Steve Peers

Introduction

Despite the EU’s purported adherence to high standards of human rights protection, the EU’s Dublin system, which allocates responsibility for each asylum-seeker’s application to a single Member State, has repeatedly run foul of human rights standards. Yesterday’s judgment of the European Court of Human Rights (‘ECtHR’, or ‘Strasbourg Court’) in Tarakhel v Switzerland, and the recent judgment of the same court in Sharifi v Italy and Greece, have further confirmed the problems in making this system compatible with ECHR obligations.

In fact, the Tarakheljudgment goes further than the prior judgments, which had merely exposed the lack of sufficient human rights protection in the EU legislation, as applied by Member States. Rather, it is now clear that the approach of the Court of Justice of the European Union (CJEU) in interpreting the Dublin rules is also incompatible with the ECHR.

Background

The Dublin rules initially appeared as part of the Schengen Convention, which bound only certain Member States. They were then set out in the form of the Dublin Convention, signed in 1990. This Convention was replaced by an EC Regulation (known as the ‘Dublin II Regulation’) from 2003. That Regulation was in turn replaced by the Dublin III Regulation, adopted in 2013, which applies to all applications made after 1 January 2014. Furthermore, the Dublin rules have been extended to the non-EU countries associated with the Schengen system, by means of treaties with Norway and Iceland on the one hand, and Switzerland and Liechtensteinon the other.

The previous leading cases on the compatibility of the Dublin regime with human rights were (for the Strasbourg court) the 2011 judgment in MSS v Belgium and Greece, and (for the CJEU) the judgment in NS, delivered later that same year. In MSS, the ECtHR ruled that Greece had violated Article 3 ECHR (the ban on torture or other inhuman or degrading treatment) in three ways: its treatment of the Afghan asylum-seeker in question in detention; its failure to secure adequate living conditions for him after release from detention; and its highly deficient asylum procedure. The evidence of these violations was found in numerous reports by NGOs and international bodies. The Court also ruled that Belgium had violated Article 3 ECHR because it had returned the same asylum-seeker to Greece (in accordance with the Dublin rules), even though it must have known of the situation there. For good measure, the Court also ruled that Belgium had violated Article 13 ECHR (the right to an effective remedy), since Belgium did not provide for sufficient reviews of the merits in cases such as this one.

Subsequently, the CJEU ruled in NS that asylum-seekers could not be returned to Greece, pursuant to the Dublin rules, because of systematic deficiencies in the asylum system in that country. Removals in such cases would constitute a breach of Article 4 of the EU Charter of Fundamental Rights (the equivalent of Article 3 ECHR). However, the Court distinguished such major breaches of fundamental rights from minor violations of EU or international rules relating to refugees, which would not require Member States to refrain from applying the Dublin rules.

Before the ECtHR could rule in Tarakhel, the CJEU clarified its position in its judgment in Abdullahi, delivered late in 2013. The Court started out by emphasising the presumption that all EU Member States protected human rights, noting that there was now second-phase legislation establishing the Common European Asylum System. It then characterised the Dublin rules as essentially regulating the relationship between Member States, referring in particular to the optional ‘sovereignty’ and ‘humanitarian’ clauses in the Dublin II Regulation, as well as the possibility of conciliation or separate arrangements between Member States. It followed that when two Member States agreed which of them was the Member State of first authorised entry (triggering responsibility under the Dublin rules), an asylum-seeker could ‘only’ challenge that decision by ‘pleading systemic deficiencies in the asylum procedure and in the conditions for the reception of applicants for asylum’ in the Member State which was deemed responsible for the asylum application.

Finally, the recent Sharifijudgment pf the Strasbourg Court established that Italy’s interception of asylum-seekers from Greece in the Adriatic, and their forced return to Greece, violated Article 3 ECHR as well as the ban on collective expulsions in the Fourth Protocol to the ECHR. In doing so, it confirmed a key corollary of the MSS ruling: Member States breach the ECHR if they stop asylum-seekers fleeing an unsafe country directly from crossing their borders. Although the rules on freedom to travel for third-country nationals in the Schengen Convention do not give asylum-seekers the right to move between Schengen States (unless, improbably, they have a visa or residence permit, or the visa requirement is waived for their country of origin), the ECHR nevertheless gives asylum-seekers the freedom to travel between Schengen countries (or any States) in such circumstances. Also, the right to move to another country extends beyond the three-month time limit on intra-Schengen travel, since asylum-seekers can in principle stay until their claim is finally rejected.

The Tarakhel judgment

Yesterday’s judgment concerned a family of eight Afghans, who entered the EU by crossing the Italian border first. This made Italy responsible for their applications under the Dublin rules. However, the family soon left the asylum-seekers’ reception centre which they were assigned to in Italy, on the grounds that conditions there were inadequate for families. They moved to Austria, which triggered the Dublin rules, asking Italy to take charge of them. Italy agreed, but before their transfer to Italy could be carried out, they moved on to Switzerland. That country in turn asked Italy to take charge of the family; Italy tacitly accepted.

However, they challenged their removal to Italy on the grounds that their treatment in that country, if they were removed there, would violate Article 3 ECHR. They lost their case in the Swiss courts, so asked the ECtHR to rule that their removal to Italy would constitute a breach of Article 3, as well as Article 8 ECHR (the right to family life). They also alleged a breach of Article 13.

The ECtHR rejected the Article 13 claim on the merits, since the Swiss courts had examined the merits of their legal arguments and they were allowed to stay on Swiss territory in the meantime. It held that it was not necessary to examine the Article 8 argument. Most importantly, by a majority of 14-3, it found that there was a breach of Article 3 ECHR.

Yet there are important differences between the MSS judgment and the Tarakhel judgment. Yesterday’s judgment does not state that Italy’s asylum system has effectively collapsed, as was the case in Greece. In particular, there were no allegations in Tarakhelrelating to flaws in Italy’s asylum procedures, or as regards detention. The argument instead was solely about living conditions in Italian detention centres.

The ECtHR began by reiterating its case law from MSS about reception conditions for asylum-seekers. While Article 3 ECHR did not guarantee a home or financial assistance, in cases involving EU Member States the Court took account of their specific obligations in that respect under the EU’s reception conditions Directive. Also asylum-seekers were an ‘underprivileged and vulnerable group’, and it was possible that extreme poverty could raise issues under Article 3. The Court also referred to other prior case law on the need to ensure that child asylum-seekers, who were in a position of ‘extreme vulnerability’, enjoyed ‘protection and humanitarian assistance’.

Next, the Court reiterated the usual rule that Article 3 prevents removal if ‘substantial grounds have been shown for believing’ that there is a ‘real risk’ of treatment contrary to Article 3 in the state of destination. The same rule could be used to rebut the assumption that countries applying the Dublin system were all safe. In this context, the ECtHR referred to the CJEU’s ‘systemic deficiencies’ test set out in the judgment in NS, but made no reference to the ruling in Abdullahi that this was the ‘only’ ground for challenging the application of the Dublin rules. However, the ECtHR also ‘notes’ the recent EM judgmentof the UK Supreme Court, which expressly stated that ‘systemic deficiencies’ were not the only ground for such challenges. Overall, the Court stated that these tests had to be applied by examining ‘the applicant’s individual situation in light of the overall situation prevailing’ in the state of destination.

Applying these rules to this case, the applicants had made three complaints about the situation of the Italian reception system. The first complaint, about the slowness of identification procedures, was dismissed out of hand, since the applicants had in fact been identified quickly. As for the second complaint, the Court accepted the evidence that there were not enough places for all asylum applicants. Thirdly, as for the reception conditions within the available facilities, a number of problems had been identified by the UNHCR and the Council of Europe’s Human Rights Commissioner.

Taken as a whole, then, the Court ruled that the ‘current situation in Italy can in no way be compared to the situation in Greece at the time of the MSS judgment’, where only a small fraction of asylum-seekers could be accommodated and ‘the conditions of the most extreme poverty…existed on a large scale’. So there could not be ‘a bar to all removals of asylum seekers to that country’. Having said that, the Court accepted that there was some risk that asylum-seekers might not get accommodation, or that the accommodation would be inadequate.

As for the individual position of the applicants, that was not comparable to the facts of the MSScase either.  The family in this case were taken care of immediately by the Italian government, rather than detained and then left to fend for themselves. But again, having said that, the Court was concerned that, in light of the vulnerability of asylum-seekers, and children in particular, there was no guarantee of (adequate) accommodation for families seeking asylum in Italy. So Switzerland could not send the family to Italy unless they obtained sufficient assurances on this point. This alone constituted a breach of Article 3 ECHR.

Interestingly, the majority judgment makes no reference to the alternative possibility of asylum-seekers obtaining private family housing at the expense of the State, which the CJEU developed in its recent Saciri judgment on the reception conditions Directive.

Comments

With great respect, there are many flaws with the CJEU’s judgment in Abdullahi. That judgment confuses Regulations (directly applicable in national legal systems) with Conventions (essentially governing relations between States). It places undue reliance on provisions of the Dublin II Regulation which were never applied in practice (conciliation) or were irrelevant to the case at hand (separate arrangements between Member States). It ignores the CJEU’s own case law on the ability to challenge Member States’ application of the Dublin II rules as regards unaccompanied minors (MA), humanitarian situations (K) or withdrawn applications (Kastrati). Its scope is unclear: does it only apply when Member States agree that the criterion regarding irregular entry is applicable, or in other cases as well? In any event, the judgment needs to be rethought in light of the Dublin III Regulation, which considerably expanded the procedural rights of asylum-seekers in the Dublin context. Why do that, if they can only challenge their transfer if there is a complete breakdown in the asylum system of the State responsible for their application?

But the most fundamental flaw in the Abdullahi judgment is exactly that: the CJEU’s statement that at least in some cases, the determination of the responsible Member State can ‘only’ be challenged if there are ‘systemic deficiencies’ in the asylum system of that State. Is that statement still correct after Tarakhel?

Certainly the statement is wrong if the CJEU meant (as it appeared to say) that both the asylum procedure and the reception conditions systems have to have failed in the responsible Member State, before a transfer to that State can be challenged. In Tarakhel, there is no issue raised regarding the asylum procedure in Italy. More generally, the Italian reception system is not in complete breakdown: the Tarakhel family faces neither extreme poverty nor vile detention conditions, but merely some risk that accommodation will either not be available or that it will be somewhat unpleasant. Accordingly, the Swiss obligations are nuanced: there is no ban on transfers, merely a procedural obligation to make arrangements with the Italian authorities.

It isn’t clear whether Tarakhelabandons the CJEU’s assumption that only ‘systemic deficiencies’ in the asylum system of a responsible State can justify a challenge to a Dublin transfer, or whether the judgment merely modifies the notion of ‘systemic deficiencies’ considerably, lowering the threshold for its application. On the first hypothesis, ‘systemic deficiencies’ are just one example of a situation that could lead to rebuttal of the assumption that another Dublin State is safe. Uncertainty about adequate reception conditions for families is another. But surely this cannot be an exhaustive list.

On the second hypothesis, a ‘systemic deficiency’ would not exist only where an asylum system had entirely collapsed, but where some particular aspect of the system was malfunctioning regularly to some extent. By analogy, a car needs to be fixed not only when the brakes entirely fail to work, but also when the windshield wipers occasionally malfunction. The risk is far greater in the first case, but the second case shouldn’t be ignored either. Again, the problems in Tarakhel cannot be the only example of a flaw in the asylum system of a responsible Member State that needs to be fixed before asylum-seekers can be transferred there.

There isn’t much difference between these two possible interpretations of Tarakhel. Although the first interpretation is in principle more open-ended than the second one, it shouldn’t take too much imagination to argue that any particular problem an asylum-seeker might face in the responsible Member State is ‘systematic’ in this very broad sense. The second interpretation does give the CJEU more leeway to back down from its head-banging judgment in Abdullahi, and explain that this was also what it had meant by ‘systemic deficiencies’ all along.

Of course, given the strong insistence on the efficiency of the Dublin system in the Abdullahijudgment, this is obviously not what the Court had meant at the time. Yet the clear message from the Tarakhel case is that there is not a simple binary distinction between cases when all Dublin transfers should stop, on the one hand, and cases when all Dublin transfers should go ahead at full speed, on the other. Instead, like a traffic light, yesterday’s judgment creates an intermediate category of cases in which national administrations must proceed with caution. This will undoubtedly make the Dublin system more costly and complex to administer, but that is often the only way to ensure that human rights are protected effectively.  

Barnard & Peers:  chapter 9, chapter 26

Read More

Non-EU benefit claimants and implications for the EU's relations with Switzerland

Steve Peers

Even the most xenophobic of British tabloids is unlikely to greet today's judgment of the Court of Justice of the European Union (CJEU) with the headline, 'Scrounging Swiss'. Yet the judgment, concerning the correct legal base of an EU measure extending revised social security rules to Swiss citizens, raises interesting questions concerning the extension of social security rules to third states in general, and on the EU's relations with Switzerland in particular.

Background

This the second in a series of three cases (to date) in which the UK has challenged the 'legal base' of Council Decisions which establish the EU's position as regards the extension of revised EU rules on social security coordination to particular third countries. The first case (C-431/11, UK v Council) concerned the European Economic Area (EEA), a treaty extending the EU's internal market to Norway, Iceland and Liechtenstein. The UK lost that challenge in September 2013. This case (C-656/11, UK v Council) concerns the extension of those social security rules to Switzerland, pursuant to the EU's free movement agreement with that country. A third case, still pending (C-83/13, UK v Council) concerns the extension of those rules to Turkey.

In all these cases, the Council has taken the view that the correct 'legal base' for its decision is Article 48 TFEU, which concerns the coordination of social security for employed and self-employed workers within the EU. Since the Treaty of Lisbon, this has provided for a qualified majority vote, although a Member State can pull an 'emergency brake' on proposed legislation if it believes that 'important aspects' or the 'financial balance' of its social security system will be affected by the proposal in question. The UK, on the other hand, has argued that the correct 'legal base' for these measures is Article 79 TFEU, which concerns the EU's common immigration policy.

This provision gives the UK (along with Ireland and Denmark) an opt-out. The UK intended to use this opt-out to reduce the cost of the provision of foreign benefits, although the total amount of the benefits paid out is surely not enormous. It may even be less than the cost of bringing this litigation.

It should be noted that the Council has taken a number of other decisions regarding social security rules for associated countries, based on Article 79 TFEU as the UK has advocated. In the Commission's view, set out in an interesting communication on social security and third countries, Article 48 should be used where the EU has a particularly close association with the country concerned, but otherwise Article 79 should be used.

Judgment

In the previous judgment concerning the EEA, the Court ruled that Article 48 TFEU had to be used because: the EEA was a particularly close association between the EU and the countries concerned; the decision aimed to extend the whole EU internal market to those countries; the relevant provisions of the TFEU and EEA treaty were the same; the EEA has particular rules on the legal effect of EEA law; the decision also applied to EU citizens in the third countries concerned; the decision merely updated prior commitments; and it would be difficult to ensure free movement in the event of parallel regimes. Article 79 TFEU could not be used, because it was 'manifestly irreconcilable' with the context and objectives of the EEA.

Most of those points were unique to the EEA, and so the UK had hoped that the previous judgment could be confined to its facts. However, the CJEU extended this case law to cover the EU/Swiss decision as well. It reasons were that: Switzerland has a 'vast' number of treaties with the EU which aim to strengthen the EU/Swiss economic relationship, even though the Swiss had voted not to participate in the EU internal market, via means of the EEA; the EU/Swiss treaty has the same wording as Article 48 TFEU; and the decision simply extends the revised EU rules to the third country concerned. 

The CJEU also ruled that Article 48 TFEU can apply to third-country nationals where an association agreement (which was the basis for the approval for the 'package' of seven EU agreements with Switzerland, in 1999) has already extended EU social security rules to the country concerned, and the decision in question merely aims to update the references concerned. Finally, the CJEU rejected the UK's argument that Article 48 could not apply to the rules on social security for those Swiss citizens not exercising economic activities (who are outside the scope of Article 48), on the grounds that this aspect of the new decision was purely ancillary to the rules on employed and self-employed persons.

So Article 48 TFEU can apply even if the relationship is not as special as the EEA, does not extend the full internal market to the country concerned, and does not contain the particular rules on the legal effect of EEA law. The most recent judgment makes no reference to the complication of having parallel regimes either. Moreover, the UK's best argument - that third country nationals not exercising economic activities were beyond the scope of Article 48 TFEU - has now been shot down by the Court.

It is surely likely, by analogy with this judgment, that the UK will lose its challenge to the Council decision regarding social security for Turkish citizens. Moreover, the Coumcil's decisions relating to social security for other associated third countries, based on Article 79, might even be legally questionable. This is because the Court's judgment today could be understood to mean that the implementation of any specific aspects of the EU's association agreements involve the use of the relevant legal base relating to internal EU law - given that, as the CJEU has always held, all association agreements aim to extend aspects of the EU's internal law to the third country concerned (Case 12/86 Demirel).

Broader implications

The CJEU makes no reference to the recent Swiss referendum, which established a legal requirement for the Swiss government to renegotiate the free movement treaty with the EU within three years, to establish a possible quota on free movement. Of course, the judgment does not concern termination of the EU's treaties with Switzerland, but rather their implementation. But it implicitly rejects any argument that the Treaty provisions on association agreements (now Article 217 TFEU) must always be used for decisions relating to such agreements. The Advocate-General's opinion in the previous EEA social security case had argued that Article 217 had to be used, but the Court did not follow that view.

It is therefore arguable that the EU could decide to denounce one or all of the other treaties in the 1999 package based on the particular legal base for the treaty concerned - such as the transport legal base for the treaty on aviation, for instance. If this is correct, most or all of these decisions to denounce could be taken by a qualified majority vote, so this would strengthen the EU's negotiating position with Switzerland when it comes time to renegotiate the main treaty on free movement of persons. It cannot be forgotten that the EU and the Swiss agreed in 1999 that the seven agreements were a package - and that the denunciation of one treaty would mean the denunciation of all of them.


Barnard & Peers: chapter 24

Read More

The Swiss vote against free movement of people and the implications for the UK's membership of the EU

Steve Peers

There are two implications of the narrow vote in the Swiss referendum to renegotiate treaties with the EU in order to permit a quota on labour migration? First, and most immediately, the vote will affect Swiss relations with the EU. Secondly, the vote foreshadows key aspects of the debate concerning the UK’s relationship with the EU, and the possible renegotiation of EU membership.

The Swiss implications

The details of what the Swiss voted for have been incisively analysed in the Kent EU rights blog post, here: http://blogs.kent.ac.uk/eu-rights-clinic/

So what are the broader implications? There is now a domestic constitutional obligation for the Swiss government to renegotiate its free movement treaty with the EU, so that labour quotas are allowed. Of course, such a renegotiation is technically possible, but will be politically difficult, since the EU insisted upon this treaty as a quid pro quo as part of a broader package that included treaties on six other issues, such as public purchasing and aviation access. The EU can, and probably will, insist on renegotiation of some of these other treaties as a consequence.

The EU should not be criticised if it demands a renegotiation of other treaties, as it had always insisted upon this link, which is set out expressly in all the treaties concerned. The Swiss public was also always aware of it. Indeed, undoubtedly the link with the other treaties explains why the Swiss public has voted for the free movement agreement in three previous referenda (once to approve it initially, and twice to extend it to new Member States). And it is clear that the supporters of a ‘no’ vote in the new referendum made the link clear to the voting public. We must conclude that the narrow majority who voted ‘yes’ thought that this would be a price worth paying.

After all, any agreement contains an element of quid pro quo. For instance, employees might like to be paid even if they don’t work, while employers might prefer it if employees worked without pay. Obviously both sides compromise; and a ‘pick-and-choose’ approach will have consequences. If employees start working for 3 days a week instead of 5, they won’t still receive full-time pay.

The implications for the UK

When re-negotiating with Switzerland, at least some Member States will be thinking about the UK. While it used to be the case that the cost of the UK’s net contribution to the EU was the main cause for Eurosceptics, that has been joined first by doubts about the EU’s democratic legitimacy and second by concerns about large-scale immigration from new Member States. Could the UK hold a referendum like Switzerland’s?

Legally speaking, no. The UK does not have a specific free movement agreement with the EU (linked to other treaties) like Switzerland does. Rather, free movement is part and parcel of our membership of the EU. If we want to be rid of it, we either have to renegotiate our entire membership or leave the EU. As a matter of domestic law, we could hold a referendum or otherwise change the free movement rules, and breach EU law while remaining a Member. But that course would be dishonest and disreputable. If the majority of the British people don’t like a key aspect of our arrangement with the EU, we should either leave or try to change that arrangement, while being aware of the consequences of doing so.

So we could ask to renegotiate our membership as far as free movement of people is concerned (among other things, of course). This is legally possible, but politically even more difficult than the Swiss case. Indeed, as I suggested already, the negotiations with Switzerland could serve as a proxy for the possible future negotiation with the UK – much as any EU negotiations concerning an independent Scotland would be a proxy for many Member States’ concerns about their separatist movements. Spanish politicians look at Edinburgh, and think of Barcelona.

It seems likely, then, that we are about to witness a ‘dry run’ for a possible British renegotiation process. This will provide a useful laboratory to test the theory that renegotiating the UK’s EU membership, or the UK leaving the EU, would only have (in Eurosceptics’ view) positive consequences for the UK.  


Barnard & Peers: Chapter 24

Read More