The Memorandum of Understanding between Italy and Libya: Does It Create Human Rights Obligations on the Part of Italy?

[Alessandro Pizzuti is co-founder of UpRights. He also worked as legal officer at the Special Tribunal for Lebanon as well as International Residual Mechanism for International Criminal Tribunals, International Tribunal for the former Yugoslavia and International Tribunal for Rwanda. Clare Frances Moran is a lecturer in law at Edinburgh Napier University, teaching and researching public international law, focusing on issues of responsibility in international criminal law, international human rights law and international humanitarian law. The authors would like to thank Luigi Prosperi and Paolo Busco for their help and suggestions for this post.]

In 2017, Italy and Libya signed a Memorandum of Understanding to strengthen their cooperation in the field of migration (MoU).  This post examines the human rights implications of Italy’s support to the Libyan Coast Guard (LCG) provided under MoU pursuant to the International Covenant on Civil and Political Rights (ICCPR), the European Convention on Human Rights (ECHR) and the Convention against Torture (CAT).

Migrants intercepted at sea by the Libyan Coast Guard (LCG) are subjected to various forms of mistreatment in Libya, which may amount to violations of the ICCPR, the ECHR and the CAT. To the extent that Italy’s support to the Libyan authorities is instrumental to such violations, these instruments may apply extraterritorially, attracting Italy’s responsibility for such acts.

On 4 August 2021, the Italian Parliament authorised again the financing Italy’s support to the LCG. In July, Uprights released a Position Paper concluding that Italy’s cooperation with Libyan authorities under the MoU is in violation of its human rights obligations and that the agreement needs to be reframed consistently in accordance with international law standards.

Background

The fate of migrants rescued by the LCG in the Mediterranean Sea is of public knowledge. International organisations (OHCHR/UNSMIL, UNHCR) and NGOs (AI, HRW) have reported that migrants rescued by the LCG have been subjected to various forms of mistreatment once returned to Libya, where they are held in detention camps manned by armed groups linked to the Libyan Government. There, migrants are subjected to systematic abuses including torture, sexual violence, and murder.

The support provided by Italy pursuant to the MoU has enhanced the capacity of the LCG to intercept migrants at sea and return them to Libya. Italy’s assistance can be categorised under two headings: remote support (provision of equipment, patrol boats, and trainings mostly carried out from Italy) and immediate support (Italian personnel, deployed in Libya, engages in the maintenance of the assets or in liaison/consultation activities with their Libyan counterparts). Such immediate support would also include the active involvement of Italian authorities in coordinating LCG’s rescue operations directly from Libya. Denied by the Italian Government (here, p.60), Italy’s coordinating role routinely emerges in judicial proceedings (here, pp.3, 20-21) and media (here, here, here). The CoE reported that the Libyan Joint Rescue Coordination Centre “is operated with the active involvement of Italian staff” (para.20).

Applicability of ICCPR, ECHR and CAT to Italy’s support to the LCG

The qualification of the abuses allegedly committed against migrants in detention camps as human rights violations seems straightforward. Such conducts may amount to, inter alia,   violations of Articles 6 and 7 ICCPR, Articles 2 and 3 ECHR, Articles 1 and 2 CAT.In terms of causation,Italy’s assistance to the LCG is linked to these violations. LCG’s ability to return migrants to Libya, where they are subsequently mistreated, is contingent on Italy’s support.

Apparently, Italy does not exercise effective control over the Libyan territory where these violations allegedly occur, nor it has control over the victims of these abuses. Accordingly, Italy’s responsibility for such violations can only be established under two conditions: (i) such support triggers the jurisdictional link of ICCPR/ECHR/CAT; and (ii) ICCPR/ECHR/CAT cover the notion of complicity in human rights violations. While not aligned, the practice of the Human Rights Committee (HRC), the European Court of Human Rights (ECtHR) and the Committee against Torture suggests that these questions may be answered in the affirmative. 

ICCPR

Extraterritorial application: The ICCPR extraterritorial application vis-à-vis the Italian support to Libya may be established on the basis of the functional approach envisaged in GC 36. Under the functional approach, the jurisdiction of the ICCPR covers the rights of individuals outside the state territory, provided that they are “nonetheless impacted by” the “military and other activities” of the State party “in a direct and reasonably foreseeable manner” (para.63). According to this doctrine, the ICCPR jurisdictional link extends to all those situations in which a State exercises power or effective control over individuals’ rights, wherever located. While GC 36 focuses on the right to life, the considerations related to the extraterritorial application should be considered of general character, pertaining to article 2.1 ICCPR, and thereby applicable also to the prohibition of torture. In any event, objections that the functional approach was designed specifically for Article 6 ICCPR in light of the nature of its protected interest should not prevent its application to Article 7 ICCPR. The protected values underpinning these provisions are of a comparable nature.

The HRC’s first application of the functional model in A.S. and others v. Italy resulted in a certain ambiguity concerning its actual breadth. There, the conclusion that Italy’s obligations under the ICCPR were engaged in respect of a vessel in distress in the Maltese SAR Zone was predicated on the notion of “special relationship of dependency” ensuing from a distress call received by the Italian MRCC and the proximity of an Italian boat to the vessel in question. Criticisms to such an approach pointed to the fact that the HRC focused on the potential, rather than the actual control over the rights of the individuals at stake (here, here). Arguably, the critical aspect of the application of the functional model to A.S. and others v. Italy emanates from the very fact that it is unclear how such model can fit squarely with States’ positive obligations.    

In the present case, the application of the functional model seems to be more straightforward. It does not hinge on an omission, but rather on positive activities performed by the State. Through its support to Libyan authorities, Italy actively exercises power over migrants’ rights rescued by the LCG. Indeed, the provision of assets and training (from Italy), and the maintenance, liaison/coordination of the LCG operations (from Libya) are activities that have a direct impact on the rights of migrants abused in the detention camps following their interception at sea. Moreover, according to the HRC, States may be found responsible for extra-territorial violations of the ICCPR “if it is a link in the causal chain that would make possible violations in another jurisdiction, where the risk of an extra-territorial violation is a necessary and foreseeable consequence judged on the knowledge the State party had at the time” (Munaf v. Romania, para.14.2). Here, the causal connection between Italy’s conduct and the abuses is undeniable. Lastly, the common knowledge concerning the fate of migrants returned to Libya leaves no doubt in terms of foreseeability.

Complicity: The fact that the violations allegedly committed against migrants are materially perpetrated by non-Italian agents should not exclude Italy’s responsibility under the ICCPR. GC 36 specifically establishes that “States also have obligations under international law not to aid or assist activities undertaken by other States and non-State actors that violate the right to life” (para.63). Again, while focusing on the right to life, the GC 36’s reliance on Article 16 Articles on State Responsibility suggests that such proposition applies to other provisions under the Covenant. Accordingly, Italy’s instrumental role in relation to these abuses may be still imputed as a violation of the ICCPR.

ECHR

Extraterritorial application: The ECHR extraterritorial application concerning Italy’s support to the LCG should be considered separately in light of the origin and the form of the support provided (remote/immediate support). The ECHR extraterritorial applicability to Italy’s support provided from Libya seems to be confirmed by the ECtHR’s jurisprudence.  According to the ECtHR “where, in accordance with custom, treaty or other agreement, authorities of the Contracting State carry out executive or judicial functions on the territory of another State, the Contracting State may be responsible for breaches of the Convention thereby incurred, as long as the acts in question are attributable to it rather than to the territorial State” (Al-Skeini, para.135; Banković, para.71). Considered under the personal prong (“State agent authority”), this form of extraterritorial model reflects functional features (lawful exercise of public functions abroad and attribution) that fit squarely in the present case. Indeed, the support provided by the Italian personnel in Libya amounts to executive functions and is carried out with the consent of Libya. Moreover, in the absence of any indication that the Italian officers have been placed at the disposal of the Libyan Government, their conduct should be considered directly attributable to Italy.

Arguably, the remote support provided to the LCG could also trigger the ECHR extraterritorial application vis-à-vis the violations suffered by migrants, despite being provided from outside the Libyan territory. Under the MoU, Italy has assumed responsibility to support LCG’s rescue operations in the Mediterranean, an issue which is currently under consideration in S.S. and others v. Italy. This support is a form of border externalisation which undermines a number of Italy’s obligations, including that of not returning individuals to a situation of torture or inhuman treatment, under Article 3 ECHR. As the ECHR has already been held to have extraterritorial effect even on the high seas, it is clear that this policies of border externalisation, in which the party with human rights obligations attempts to evade them through support of a non-party to the Convention, do not dissolve the jurisdictional link. The link remains because of Italy’s material support and the well-documented inhuman conditions to which the migrants are returned.

Complicity: There is a strong argument, based on the non-refoulement obligation under Article 3 ECHR, to support a case of complicity in respect of the Italian Government. Miles Jackson highlighted a pattern within the Article 3 jurisprudence requiring States to avoid complying with violations of human rights by allowing, or extraditing, individuals to countries in which they would suffer. Through this lens, Italy’s conduct can be seen as inconsistent with Article 3, even where direct commission is not evidenced. Under the practice of the ECtHR, States have thus been required to accept their responsibilities, which include the avoidance of tacit support for violations in other countries. Italy’s support goes beyond tacit compliance, but the point remains that the passing of responsibility in such a fashion to Libya has already been outlawed by the ECHR. Milanovic and Shah, in their amicus brief submitted for Ukraine and The Netherlands v Russia have also argued for a complicity basis in respect of State responsibility for rights violations. This demonstrates a wider understanding of the value of such an approach and the significance of holding States to their responsibilities, regardless of the methods employed to evade liability.

CAT

Extraterritorial application: The CAT can apply extraterritorially to the facts at issue. The principles under which CAT may operate extraterritorially slightly differ from ICCPR and ECHR. The CAT does not contain a general jurisdictional clause, rather such limitation is directly incorporated in its various provisions. Thus, the assessment of the scope of Italy’s responsibility is oriented by the scope of the different obligations under the CAT.

Notably, the practice of the Committee against Torture reflects that along positive obligations – including the prevention of acts of torture through legislative/administrative measures or establishing criminal jurisdiction (e.g. Articles 2.1, 5.1 and 7.1) –, the CAT also implicitly imposes negative obligations to refrain from committing torture (Ben Salem v. Tunisia, paras.16.4, 16.5; Slyusar v. Ukraine, paras.9.2, 9.4, Milanovic, pp.213-215, Zach, p.63). Arguably, while the responsibility for positive obligations is confined to acts occurring in the territory under the State’s control, there is no geographical limitation vis-à-vis the negative obligations (Milanovic, p.215). In this vein, the Special Rapporteur on Torture concluded that “States’ negative obligations under the Convention are not per se spatially limited or territorially defined, nor are its obligations to cooperate to end torture.” (para.28).

Under this perspective, CAT’s extraterritorial application mirrors the ICCPR functional model. State’s activities which impact the right of an individual not to be tortured would fall within the boundaries of the CAT, wherever the individual is placed. As such, the CAT should apply to Italy’s support to the LCG (both performed remotely or from Libya) that contributes to torture, wherever such acts are committed. The acts underpinning such form of assistance are inconsistent with the CAT negative obligations.

Complicity: The assistance provided to acts of torture can attract state responsibility under the CAT as a form of complicity, even if such acts are committed outside the State’s territory. First, complicity in torture falls squarely in the prohibition of torture under Articles 1, 4 CAT. Second, the practice of the Special Rapporteur (paras.20-25) and the Committee against Torture (para.34) indicates that complicity in acts of torture occurred abroad is inconsistent with the CAT. Thus, even if migrants are not subjected to torture by Italian agents, Italy may be in breach of Articles 1 and 2 CAT because of the support provided under the MoU, which has contributed to these violations.

Conclusion

To a certain degree, the MoU has been the result of the ECtHR Hirsi case. There, Italy was found in violation of the ECHR after a vessel of the Italian navy intercepted migrants in high seas and transferred them back to Libya. The control exercised by the Italian authorities over the migrants during their interception/transfer was crucial to establish that the relevant violations fell under Italy’s jurisdiction. As applied, the MoU attempts to avoid future violations of this nature by removing the overt role of the Italy in such interceptions, creating an artificial hiatus between the activity of the Italian authorities and the underlying violations suffered by the migrants attempting to cross the Mediterranean Sea.

However, a survey of the practice underlying the ICCPR, ECHR, and CAT suggests that despite these efforts, such hiatus is nothing more than a thin veil which has been drawn over the Italian government’s role in returning migrants to Libya. The principles governing the extraterritorial application and complicity for human rights violations of these instruments can, and should, pierce such a veil, identifying the responsibility of Italy for these human rights violations.

Under Article 5, the MoU requires Italy and Libya to execute the agreement in full respect of their respective human rights obligations, the only possibility for Italy to avoid incurring in these forms of responsibility is ensuring that their Libyan counterparts fully comply with such clause.