Updates, commentary, training and advice on immigration and asylum law

The EU-Turkey refugee deal: a Greek myth


Older content is locked

A great deal of time and effort goes into producing the information on Free Movement, become a member of Free Movement to get unlimited access to all articles, and much, much more


By becoming a member of Free Movement, you not only support the hard-work that goes into maintaining the website, but get access to premium features;

  • Single login for personal use
  • FREE downloads of Free Movement ebooks
  • Access to all Free Movement blog content
  • Access to all our online training materials
  • Access to our busy forums
  • Downloadable CPD certificates

It was hailed as an emergency solution to the refugee crisis but figures reveal that only 36 Syrians had been removed from Greece under the “EU-Turkey statement” by the end of 2018, write Taimour Lay and Theodoris Zeis.

It is over three years since the “EU-Turkey statement” on the Mediterranean migration crisis was signed and its supporters are still convinced it has been a success. Perhaps it was – for the signatories at least. In a recent BBC documentary, Inside Europe: Ten years of turmoil, Dutch prime minister Mark Rutte gushes over “killing the business model of the boat smugglers” and the drama of conducting last-ditch negotiations with Turkish prime minister Ahmet Davutoğlu behind the backs of the European Council.

It was Davutoğlu who, in March 2016, dangled the proposal of taking refugees back from Greece in exchange for £6 billion, accelerated EU membership talks and visa liberalisation for Turks in the Schengen area. Meanwhile, for every Syrian readmitted by Turkey from Greek islands, another Syrian from Turkey would be resettled to the EU member states.

“Nobody lost in this deal. It was a typical win-win solution”, said Davutoğlu, who resigned three months later after one disagreement too many with President Recep Tayyip Erdoğan.

Leaning heavily on the existing principles of “safe third country” and “safe country of asylum” embedded in the EU’s Asylum Procedures Directive, and taking advantage of gaps and ambivalences in the UN Refugee Convention, the arrangement was a tactical mix of law, politics and public relations. There is no risk of limbo, said the deal’s proponents. Syrians (and others) would be removed back to Turkey and granted Temporary Protection Status there. German think-tanker Gerald Knaus was one of the original designers of the deal. “People aren’t drowning, they don’t have to pay traffickers, they are not trapped in Greece for years,” he argued in 2016.

Except, in reality, key components were never implemented.

Hardly any refugees have been sent back to Turkey

Thousands of refugees trapped on Greek island “hotspots” suddenly found themselves undergoing “admissibility interviews” and being told that their asylum claims would instead be decided in Turkey. The effect of the EU-Turkey arrangement was to absolve Greece (and the EU) of responsibility for granting or refusing those post-March 2016 claims. This was coupled with amendments to Greek asylum law to further facilitate “fast-track” procedures on the islands and ongoing detention.

But the removals have largely proved a chimera. While the press releases confidently spoke of “rapid returns” from the islands to Turkey (and “1:1” exchanges in the other direction), between March 2016 and March 2019 a grand total of “1,836 migrants have been returned from Greece to Turkey under the EU-Turkey Statement”.

Even that small figure is misleading because over half were Pakistan and Bangladesh nationals effectively seeking voluntary return. 45% either did not claim asylum or withdrew claims so as to facilitate exit from the island hotspots, according to UN refugee agency figures.

As for Syrian refugees:

In total, between April 2016 and 31 December 2018, Syrian nationals account for 337 persons (19%) of those returned [under the EU-Turkey agreement]. 36 of them have been returned on the basis that their asylum claims were found inadmissible at second instance on the basis of the “safe third country” concept.

That’s just 36 forced Syrian returns under the EU-Turkey deal in two and a half years.

As for “resettlement in the other direction”, the so-called 1:1 system, figures are not readily available. They have been subsumed within the broader EU resettlement programme that Eastern European countries continue to resist.

Syrians are trapped in Greece instead

The net effect was to leave Syrian refugees who arrived from Turkey after the March 2016 statement – and who could not meet nebulous “vulnerability criteria” – stuck without status in Greece, their asylum claims deemed “inadmissible”, with no right to proceed on to the mainland and a well-founded fear of heading back to Turkey and elsewhere. As a result, Moria camp in Lesbos grew to four times over capacity (Samos three times) in desperate conditions characterised by overcrowding, fires, hunger strikes and protest. The island “fast-track” hotspots became detention centres.

Buried in the EU’s latest assessments is a recognition that the system is broken. In March 2019, the European Commission noted that

in the Eastern Mediterranean, numbers increased throughout 2018 and arrivals in Greece in 2019 are now more than 30% higher than a year ago. This trend puts continued pressure on the Aegean Islands, but also on the Greek-Turkish land border, as there are far too few returns to Turkey to alleviate this pressure and counter the business model of smugglers… The pace of returns to Turkey from the Greek islands under the Statement continues to be very slow, especially concerning Syrians.

Opponents of the deal may see this as a partial success – Syrian refugees and others are somehow resisting removal from Europe’s borders – but dysfunction is not an answer to the crisis on the islands when refugees are left without status. The effective breakdown of the arrangement should have instead led the EU to accept that refugees should have their asylum claims evaluated inside Europe’s border, with “burden sharing” across (not outside) the borders of the EU27. Instead, they chose stasis on the islands under the pretence of future removal to Turkey.

What has the EU-Turkey deal failed?

If admissibility hearings continue and refugees receive decisions, why are removals not being carried out? It seems little to do with the law.

Systemic challenges in the Greek appellate system eventually proved a dead end. After initial successes, mostly attempts to argue that individual appellants fell within the scheme’s “vulnerability” exemption, the senior courts have repeatedly affirmed that Turkey is “safe” and there is no breach of the Refugee Convention or the European Convention on Human Rights in returning prima facie refugees.

But there has been an evasion of the main issue. In the leading Greek Council of State case (2347-2348/2017), the judges upheld a decision that the Syrian nationals involved could be returned to Turkey.  But they did not do so by reference to the EU-Turkey deal, instead ruling that Article 38 of the Asylum Procedures Directive alone provided the basis for the rejection. In so doing, the Council of State sought to side-step any controversy over the status and reliability of the EU-Turkey statement, pointing out that the safe third country concept was already legitimately part of the EU’s common asylum framework pre-2016. This ignored that the statement had directly led to amendments to the fast-track procedure on the islands and has plainly been advertised as the means by which the EU was asserting Turkey was safe.

That hasn’t prevented more progressive decisions on their facts, with the 9th Appeals Committee as recently as 2018 finding that a Syrian from Afrin who spent just eight days in Turkey could not be said to have a meaningful connection with that country and thus shouldn’t be subject to a safe third country decision.

At a European level, the challenge in NF NG and NM v European Council also failed, with a ruling that the General Court had no jurisdiction. This relied on the tendentious technicality that the “statement” was supposedly not an act of the EU as an institution but rather a political declaration between individual members states and Turkey via the European Council. In September 2018, the Court of Justice of the European Union agreed.

Meanwhile, a full-frontal attempt to extinguish the deal has finally wended its way to the European Court of Human Rights. JB v Greece is due to be heard some time in 2019 with the parties arguing that Turkey cannot lawfully be characterised as a “safe third country” owing to deficiencies in its asylum procedures, with only “general assurances” offered by the Turkish government. In particular, it is argued that protection status in Turkey is not “comparable” or “equivalent” to that which a refugee is entitled to within the Refugee Convention in Europe.

But the mechanisms for implementation have ultimately unravelled in extra-legal ways. In July 2016 an attempted coup in Turkey led to a crackdown on President Erdoğan’s opponents and sections of the military and security apparatus. Turkish officers sent to the Aegean islands as part of the EU-Turkey statement found themselves recalled. 6,000 Turkish citizens have sought asylum in Greece to escape state reprisal for suspected support for the Gülen “movement”.

Meanwhile, appetite on both sides for removals has drifted: EU reluctance to waive strict criteria for visa liberalisation has meant the full “deal” was never likely to be implemented. President Erdoğan’s interest in accession talks has likewise waned.

At the same, the Greek authorities appear to have taken a pragmatic view over time. Syrians deemed not to be admissible, after many months trapped on the islands, have been given opportunities by the Greek Asylum Service to transfer to the mainland for reasons of vulnerability and administrative convenience. According to a senior judge, that policy means Syrians in reality move outside the confines of the EU-Turkey statement and are no longer at risk of removal.

Time to properly protect refugees in the EU

In practice, then, the statement is not being implemented, has evaded judicial scrutiny and has come at a cost of an ongoing crisis on the islands. There is no doubt that arrivals to Greece by sea have reduced, from a peak of 860,000 in 2015 to 170,000 in 2016, 30,000 in 2017 and 32,000 last year. But reductions in refugee flows across the Aegean have, since 2017, also been determined by changes in the dynamics of the Syrian civil war – in other words, those who were able to flee at the height of the war had already done so, although it remained in Turkey’s interests to trumpet its own efforts to “contain the smugglers” in the second half of 2016.

The nationality of refugees arriving by sea in 2018 had also shifted, with over a third from Afghanistan and 15% from both Syria and Iraq. At the same time, numbers passing by land from Turkey and through the Western Balkans increased by 284% in 2018.

EU decision-makers can’t have it both ways. If the EU-Turkey deal has “worked”, then the emergency is over and the temporary provisions need no longer apply. If it has failed, then admissibility decisions against Syrians should end. There needs to be a sincere debate about creating a legal regime that prioritises protection inside the EU’s borders.

Taimour Lay and Theodoris Zeis support the work of Refugee Legal Support Athens which has been providing advice to refugees in Greece for two years. The project is in need of funds to continue its work in 2019. Please consider donating.

Interested in refugee law? You might like Colin's book, imaginatively called "Refugee Law" and published by Bristol University Press.

Communicating important legal concepts in an approachable way, this is an essential guide for students, lawyers and non-specialists alike.

Relevant articles chosen for you
Picture of Taimour Lay

Taimour Lay

Taimour is a barrister at Garden Court Chambers specialising in refugee and immigration law. He previously spent 10 years as a journalist and has reported from East Africa and Somalia.