From Turkey to Libya: The EU Migration Partnership from Bad to Worse


From January 2014 to end of 9 February 2017, 12,397 people are reported to have lost their lives at sea in the Mediterranean. At the informal Summit held at La Valletta on 3 February 2017, the European Council agreed a Declaration (Malta Declaration) concerning the external aspects of migration and the Central-Mediterranean route. It states that the EU primary goal is to train and equip Libyan coastguard in order to bolster its capacity to stop people smugglers, increase search and rescue operations, and prevent departure of unseaworthy boats headed toward Europe.

The European Commission’s proposal to set up the so-called Migration Partnership Framework (MPF) was endorsed, in June 2016, by the European Council. The self-proclaimed objective of the EU MPF is to strengthen relationships with third countries to better manage migration. Short-term actions will be directed to save lives at sea and in the desert; dismantle traffickers and smugglers’ networks; increase returns of those who are not entitled to stay; avoid dangerous journeys for migrants and refugees who can thus stay closer to their home country; and open up legal ways to Europe for people in need of protection, in particular with resettlement of refugees. In the long-term, the main goal is to address the root causes of irregular migration and forced displacement by supporting the economic, social and political development of third countries. All these objectives will be attained through ‘compacts’, which are political ‘packages’ setting up clear goals and joint commitments, such as the negotiation of readmission agreements with countries of origin and transit. In this regard, dialogues and negotiations with Mali, Nigeria, Niger, Senegal and Ethiopia have already started (for commentary, see, Bauloz).

This contribution will briefly analyse the MPF focusing on the EU cooperation with Turkey and Libya. In so doing, it will also discuss the Malta Declaration and the interrelated Memorandum of Understanding (MoU) between Italy and Libya (February 2017) aimed to stem migratory flows by thwarting departures from North Africa. Whilst in 2016 the number of refugees crossing the sea to reach Europe plunged to 364,000 (one million in 2015), the number of those who died in the Mediterranean (7,495 persons) rose sharply. A significant drop in arrivals to Greece outweighed record migration to Italy, as a consequence of the EU-Turkey deal (signed in March 2016) and tighter border controls in the Western Balkans.

To address what has been labelled as a ‘migration crisis’, the EU has opted for an emerging strategy based on externalization of migration controls. Quite questionably, the EU and Member States have made financial and political support as well as cooperation on development conditional on third countries’ effective implementation of exit controls. As affirmed by the European Council, the new EU MPF is based on ‘effective incentives and adequate conditionality’ [but] ‘cooperation on readmission and return will be a key test of the partnership between the EU and [its] partners’.  This has been confirmed by the so-called ‘humanitarian readmission programme’ under the terms of the EU-Turkey deal. Questionable readmission agreements and enhanced cooperation on return have also been proposed with refugee-producing countries, such, Mali, Afghanistan, Nigeria, Senegal, and Ethiopia with the objective of offering financial support in exchange for accepting returns from Europe. The EU and Member States also plan to expand their schemes of cooperation on management of migration to other countries, such as Egypt, Tunisia, and Sudan. In January 2017, a Commission Communication for the Southern Mediterranean set out the goals to both step up the training programme of the Libyan coastguard to autonomously conduct search and rescue (including disembarkation) and strengthen Libya’s southern border to hinder migratory flows to Libya and from there into Europe. As the EU is politically and economically supporting enhanced bilateral cooperation between Member States and third countries to prevent unauthorized arrivals, it is within such a broad EU containment strategy that the Italy-Libya MoU should be read.

In a logic of border closure, which purports to trap tens of thousands of migrants and refugees in Libya, a war-torn country where they are exposed to the risk of torture and exploitation, the question arises over how far these policies can reasonably be pursued. The main conclusion here is that an externalization policy, which aims to outsource migration controls and curtail arrivals through technical and financial support to third countries, will end up jeopardizing refugees’ fundamental rights. Rather than insulating States against accountability, these practices of externalization may still engage the international responsibility of European States for breach of the principle of non-refoulement and the right to leave. Therefore, legal pathways to Europe, including visa on humanitarian grounds, should be promptly provided to avoid dangerous journeys and ensure safe access to international protection.

 The EU-Turkey deal

On 18 March 2016, the EU and Turkey reached an agreement providing for ‘rapid return of all migrants not in need of international protection crossing from Turkey to Greece and to take back all irregular migrants intercepted in Turkish waters.’  The joint Statement provides for a mechanism of systematic and ‘rapid return of all migrants not in need of international protection crossing from Turkey to Greece and to take back all irregular migrants intercepted in Turkish waters’ (for discussion on its legally binding status, see, Gatti). The agreement provides that migrants arriving in Greece will be duly registered and their asylum applications processed in accordance with theProcedures Directive 2013/32/EU.  Moreover, it establishes that for every Syrian being readmitted to Turkey from Greece, another Syrian will be resettled from Turkey to the EU prioritizing those migrants who have not previously entered or tried to enter the EU irregularly. Turkey will also take any necessary measures to prevent new irregular arrivals on Greek islands and will cooperate with the EU to this end. In turn, EU Member States will both accelerate the fulfilment of the visa liberalization roadmap in view to lifting the visa requirements for Turkish citizens and speed up the disbursement of 3 billion Euros to Turkey under the Facility for Refugees.

Readmission to Turkey and resettlement to the EU are two elements of the cooperation between Turkey and the EU. Despite the EU-Turkey Agreement explicitly refers to the non-refoulement obligation and the prohibition of collective expulsions, it lacks any reference to specific vulnerabilities of people landing on Greek islands. Additionally, in Greece, the capacity of the registration system is very limited, and refugees face the hardship of lengthy asylum procedures and overcrowded reception centres. Those people whose asylum applications are unfounded or inadmissible will be returned to Turkey. Following several Appeal Committees’ decisions rebutting the ‘safe third country’ presumption regarding Turkey, the Greek Parliament approved an amendment to its asylum law (L 4375/2016), which modifies the composition of Appeals Committees. It also sets out the necessary legal provisions both to apply the concepts of ‘safe third country’ and ‘safe first country of asylum’, and to ensure fast-track procedures for the assessment of asylum applications, including appeal procedures. On 6 April 2016, Turkey adopted a law and made specific assurances to guarantee that Syrian nationals returned to Turkey may request and be granted temporary protection.

Nevertheless, Turkey’s geographical limitation to the 1951 Refugee Convention makes unrealistic any possibility to request and receive refugee protection for those coming from non-European countries. These people can indeed only obtain a status of ‘conditional refugee’, granted on a temporary basis under the Turkish Law on Foreigners and International Protection, which came into effect in 2014. Syrian refugees and asylum seekers are instead removed from Greece to Turkey on the predicament that Turkey has implemented a temporary protection regime whereby all Syrians are automatically entitled to protection without going through a full asylum application process. The definition of Turkey as a ‘safe third country’ has been questioned by a number scholars (e.g., Chetail; Favilli; Labayle & de Bruycker; Roman & Peers; Ulusoy) and NGOs (e.g., Amnesty International, Human Rights Watch).  While it should be emphasized that Turkey is currently host to more than 2.7 million registered Syrian refugees, it has also been reported that migrants and refugees are often subject to detention and mistreatments (see, Abdolkhani and Karimnia v Turkey). Particularly after the instability created by the failed military coup in July 2016, and the authoritarian turn, the Parliamentary Assembly of the Council of Europe argued that returns of Syrian refugees to Turkey as a ‘first country of asylum’ may breach EU and/or international law, as Turkey may not ensure protection that is ‘sufficient.’ In the aftermath of the failed military coup, Turkey declared a state of emergency and submitted a formal notice of derogation to the European Convention of Human Rights (ECHR). It also notified the UN Secretary General that it might take measures, which could entail derogations from obligations under the International Covenant on Civil and Political Rights (ICCPR).

Additionally, Turkey has recently concluded 14 readmission agreements with countries of origin of migrants and asylum seekers, thereby increasing the risk of repatriation and refoulement. Turkey is formally and informally returning people back to their countries of origin, such as Afghanistan, Iraq, Pakistan, and Syria where they might face extreme danger to their life (see e.g., Siegfried). In the deal between the EU and Turkey, the dubious interlink between readmission and access to protection is not that veiled. First, returns take place under the 2001 readmission agreement between Greece and Turkey, replaced in June 2016 by the EU-Turkey Readmission Agreement.Second, such a human transfer is explicitly defined as ‘humanitarian readmission programme’, in the attempt to show, through the use of evocative words, the humanitarian purpose of this agreement – in other terms how the return of asylum seekers is de facto facilitating refugees’ access to protection.

By linking readmission of irregular migrants with resettlement, such ‘one to one’ programme – if used as a blueprint for cooperation with other third countries – might create an incentive on States of transit to readmit the highest number of migrants in order to enhance the resettlement commitment of the EU. If we then consider that Turkey is restricting entry visa requirements and is negotiating readmission agreements with several countries of origin, it can be assumed that the EU is increasingly delegating to Turkey the policy of containment of migration flow, despite the risk of chain refoulement of people in need of protection.


The Italy-Libya MoU

If it is said that history repeats itself, the relationship between Italy and Libya on migration control is no exception. It was February 2009 when the then Italian Ministry of the Interior and Libyan Colonel Gaddafi signed a Technical Protocol opening the way to the highly criticized push-backs of boat-refugees to North Africa (see e.g., Giuffré; Nascimbene; Terrasi; Trevisanut). Following the landmark Hirsi Jamaa v Italy case (2012) where the European Court of Human Rights (ECtHR) found Italy responsible for violations of Articles 3 and 13 of the Convention as well as Article 4 of Protocol 4, this questionable practice was suspended (for comments, see, Di Pascale; Giuffré).

However, ahead of the EU Council summit, on 3 February 2017, the Italian Prime Minister and the Head of the National Reconciliation Government of the Libya State signed a ‘MoU on cooperation in the development sector, to combat illegal immigration, human trafficking and contraband and on reinforcing the border security’. The 2017 MoU also brings to a revival of the full array of old agreements on migration control, which had seemingly been suspended during the Arab Spring and the escalation of the Libyan civil war. The accord erroneously refers to migrants with the term ‘clandestine’, which, since 2014, had been cancelled from all official documents of Italian authorities, as considered legally incorrect. Used even before people are able to lodge an asylum application, it contains indeed an aprioristic negative judgement on the intentions of the migrant.

Despite the current chaotic and dangerous situation in Libya following the overthrown of Colonel Gaddafi in 2011, the parties to the MoU agree on the need to find rapid solution to the problem of so-called ‘illegal’ migration to Europe in full respect of international human rights treaties (Preamble and Article 5). Italy thus accepts to fund the establishment of reception centres in Libya where migrants and refugees wait for their voluntary or forced return to their home countries (Preamble and Article 2). To this end, readmission agreements with states of origin should be concluded. Also relying on funds made available by the EU, Italy commits to provide (unquantified) technical and economic support to Libyan bodies and institutions in charge of the fight against illegal immigration, including the Defence Ministry border guard and the coast guard (Article 1).

It is also worth noting here that, on 9 February 2017, Italy has signed another similar framework agreement with Tunisia. While more stable than Libya, Tunisia is a fragile state with weak government institutions, which may not be capable to cooperate. Moreover, Tunisian law penalises irregular entry, stay, and exit, and does not guarantee adequate refugee protection.

Critical Assessment

A key goal of the EU policy of externalization of migration controls is to curb migratory flows, reduce human trafficking and combat the industry that smuggles people across both the Mediterranean and Aegean sea, thus preventing departures to seek protection in Europe. In this context, third countries are called to carry out effective exit controls and halt new arrivals on their territory. While such a dual goal is not explicitly stated in the EU-Turkey deal, it is instead clearly put in the Italy-Libya MoU and the Malta Declaration, which both purport to strengthen the capacity to manage Libyan southern borders. According to Spijkerboer, while push-backs at the Syrian-Turkish border have been documented since at least August 2015, it is not surprising that Turkey, following the deal with EU, has started using systematic violence against Syrian refugees attempting to cross the Turkish border.

Whilst altering geographies and moving borders, the EU and Member States have conditioned financial and technical support to third countries’ cooperation in preventing migrants and refugees’ access to Europe  – which in other words means catching up them in a system of outright containment. Although the EU has announced a €200 million plan to finance migration projects geared towards Libya, local authorities in Tripoli and other cities are mounting resistance to EU plans to curtail migration to Europe. Their main argument is that the EU should deal with the migration issue, without passing the burden to Libya, which does not have the capacity to manage all migrants and refugees who will remain on its territory, thus fomenting a climate of racism (see, Euobserver). On 14 February 2017, a group of Libyan lawyers and former politicians have also lodged a complaint to Tripoli Appeal Court claiming the unconstitutionality of the Memorandum, which has been signed by President al-Sarraj in Rome without seeking the approval of the Parliament and the Government (Cammilli). While up to 2,000 militias operate in Libya also ruling the coastline, the UN-backed government has limited control over the eastern region of the country. Nevertheless, the MoU is presented by the Italian Prime Minister as ‘just a piece’ of broader EU-wide plan aimed to seal off the migratory flows from Libya, while the Libyan President al-Sarraj told reporters that the deal includes ‘humanitarian repatriation’ of migrants.

If the EU and Italy are engaging in the ambitious project of training and equipping Libyan coastguard, it should first be asked whether Libya is a safe place for migrants and refugees ‘pulled back’ when rescued or intercepted in its territorial waters. A ‘place of safety’ is broadly defined as a place where the rescue operations are considered to terminate (para 6.12 2004 IMO Guidelines). Here, the lives of those rescued should not be under threat and their basic human needs, such as those relating to food, shelter, and medical exigencies, can be met. However,the 2004 Guidelines still fall short of a duty of disembarkation in any particular place, and a number of examples show how coastal States (including Libya) keep on bouncing responsibilities for disembarkation. Such an uncertainty has also led to episodes of non-rescue, delays (see, e.g., the MV SalamisandFrancisco y Catalinacases), the sinking of overloaded dinghies, or boats left adrift for several days with the consequent death of the passengers on board (see, e.g., the ‘Left-to-die’case)

A ‘rescue’ mission can be considered fully accomplished only when the stowaways are disembarked in a ‘place of safety’, not when initial succours to supply basic needs are completed. When dealing with refugees, ‘rescue’ includes disembarkation in a place on dry land where the lives and freedoms of those alleging a well-founded fear of persecution, torture, and ill-treatment would not be threatened, as stated by para 6.17 of the 2004 IMO Guidelines. In accordance with the principle of non-refoulementunder international human rights and refugee law, Article 4 of Frontex Regulation 656/2014 provides that no person shall be disembarked in, forced to enter, conducted to or otherwise handed over to the authorities of a country where his/her life can be put at risk, where he/she can be persecuted, and ‘from which there is a serious risk of an expulsion, removal or extradition to another country in contravention of the principle of non-refoulement.’ Therefore, by rejecting a fragmentary approach to maritime obligations, States should adopt a comprehensive approach to boat migration, which aims to reconcile ‘maritime security’ with human rights and refugee protection standards. This means that the different bodies of law operating at sea (human rights law, refugee law, law of the sea, and search and rescue) should be harmonized taking into account the diverging policy interests that are at stake.

As stated by the ECtHR – which broadly relied on reports of international human rights organizations and the UNHCR – Libya could not be considered a ‘place of safety’ because of the well-documented inadequacy of its response to flows of migrants and asylum seekers(Hirsi v Italy case, 2012). Considering the absolute nature of Article 3 of the ECHR, potential ‘problems with managing migratory flows cannot justify recourse to practices which are not compatible with the State’s obligations…’ (ibid, para 179). Therefore, any action ‘the effect of which is to prevent migrants from reaching the borders of the [would-be host] State’ (ibid, para 180) may amount to refoulement if there is a foreseeable ‘real risk’ of ill-treatment in the countries where migrants and refugees are trapped.

 The fact that a State is legally bound to disembark a person rescued at sea in a safe haven implicates the duty to consider the condition of reception and treatment of rescued migrants and refugees in the receiving country. But the situation of migrants and refugees in Libya has dramatically worsened since Gaddafi was ousted. People rescued in the Mediterranean report horrible conditions. They claim they would rather die at sea than going back to Libya. Therefore even if the coastguard were well trained in search and rescue, the refugees on the boat would panic as soon as they realize they will be returned to Libya. These situations can be very dangerous if we consider that more than 4000 deaths at sea in 2016 were caused through panic situations during rescue operations.

In the failed State of Libya with no stable government in place, the situation on the ground is so chaotic that no clear-cut rule and consistent practice can be identified. Although Libya acceded to the SAR Convention, it has not provided the required notification regarding the limits of its SAR region, nor has it designated a responsible rescue coordination centre. In several recent instances, Libya has reportedly endangered rescue missions in the Mediterranean, in particular those conducted by lifeboats sent by the Maritime Rescue Coordination Centre (MRCC) in Rome (see, e.g., the following incidents: Bourbon Argos; Sea-Eye; Iuventa).

Libyan authorities currently lack the personnel and equipment to patrol over 1,700 km of coastline. Helped by Libya-based smugglers, nearly 150,000 people have set off for Italy in unseaworthy boats in 2016. A group of 78 Libyan Coast Guard trainees and their mentors have recently started a training programme launched by the EU on the high seas as part of EUNAVFOR MED Operation Sophia. However, the EU aims to train around 1,000 Libyan coast guard members in total. The training program is expected to last 4 months at sea to promote skills of Coast Guard personnel in conducting rescue operations and intercepting boats of irregular migrants. The German humanitarian group Sea-Watch denounced an alleged attack on migrants by Libyan Coast Guard on 21 October 2016. However, Libyan authorities denied the attack and the EU decided to continue the training program.

While recognizing that Libya is not a stable country yet, the Malta Declaration emphasizes that priority will be given to training, equipment and support of the Libyan coastguard and other relevant agencies. This Declaration also mentions the need to both ensure adequate reception conditions for migrants in Libya together with international organizations, and support IOM in carrying out voluntary return activities to countries of origin. However, in a joint statement released by UNHCR and IOM, both organisations argue that widespread insecurity in Libya made it difficult to provide even basic services to people in need of protection. Accordingly, it cannot be considered a safe third country for return of migrants and refugees nor to establish extraterritorial processing of asylum-seekers in North Africa (ASGI).

The technical and financial assistance supplied by the EU aims to create a Libyan coast guard able to independently patrol the territorial waters by next summer, before the (migrant-departure) season begins. No one would deny that there is a need for a comprehensive search and rescue mission in the Mediterranean to save as many lives as possible. However, Libya, under the constant blackmail of violent and armed militias needs stabilization before any cooperation on the life of migrants and refugees can be set up. If migrants and refugees will be rescued by Libyan coast guard and disembarked in Libya – which is not a contracting party to the 1951 Refugee Convention and does not guarantee access to asylum – European States might engage international responsibility for breaching the fundamental rights of those rescued, such as the right to leave any country (e.g. Article 13(2) of the Universal Declaration on Human Rights, Article 12(1) of the ICCPR, Article 5(d)(ii) of the International Convention on the Elimination of All Forms of Racial Discrimination, and Article 2(2) of Protocol No. 4 of the ECHR) and principle of non-refoulement, which is part of customary law and treaty law (e.g., Article 3 of the Convention against Torture, Article 3 of the ECHR, Article 7 of the ICCPR, Article 33(1) of the Refugee Convention). Indeed, thousands of migrants and refugees rescued at sea by the Libyan coast guard and brought onshore are arrested and detained (e.g, in the al-Nasr detention centre) for migration-related offences. Trapped in a conflict ravaged country, they would face torture and other ill-treatment, forced labour and exploitation in centres run by the Department of Combatting Illegal Migration (DCIM) and by armed groups. There is also a risk of deportation to countries of origin.

To conclude, any proposals to place European coast guard patrols outside the extensive Libyan coast, in joint patrols with Libyan counterparts, who will then take intercepted migrant boats back to Libyan shores, will have grave human rights consequences. The same concerns apply to any EU-Libya cooperation initiatives designed to halt the movement of refugees into Libya by strengthening Libya’s southern border. Libya is not a ‘place of safety’ for disembarkation and is not a ‘safe third country’ for refugees, as recognized by international courts, international organizations, and NGOs. By viewing Libya only through the lens of migration, Italy (and the EU more broadly) fail to grasp that the lack of strong institutions along with a weak political, economic and security situation in the country (and in its neighbourhood) makes Libya an unreliable partner.

The EU’s own Fundamental Rights Agency has set out the range of options to allow greater safe access to protection. For instance, EU Member States could process humanitarian visa applications in their embassies, or extend other visa categories to those fleeing. Some do this quietly (e.g, the humanitarian corridors launched by Italy) but a coordinated move in this direction is legally warranted. In this regard, on 7 February 2017, Advocate General Mengozzi’s Opinion in the case C-638/16 PPU argues that EU Members States must issue a visa on humanitarian grounds where substantial grounds have been shown for believing that a refusal would place persons seeking international protection at risk of torture or inhuman or degrading treatment prohibited by Article 4 of the EU Charter of Fundamental Rights (for commentaries, see Gatta; Moreno Lax). Therefore, many small actions, including suspension of harsh policies of migration containment, could ensure refugees’ safety by enhancing legal routes to the EU and reducing demand for the services of smugglers.