Francesco Messineo is lecturer at Kent Law School, Canterbury.
For a country so obsessed with the bella figura (loosely translated as ‘making a good impression’), these are not good days for Italy – at least as far as international law is concerned. Only a few weeks after having been told by the International Court of Justice that its Court of Cassation had misapplied rules on state immunity against Germany, the government found itself in the position of having to rely on those same rules against India in an intricate case involving Italian soldiers allegedly killing unarmed fishermen off the coast of Kerala. On this occasion, the Italian Minister of Justice made some spurious assertions as to the lack of Indian jurisdiction over the killing of its own citizens which have been legally untenable at least since one day in September 1927 (ignore Lotus at your peril…).
But there is more. Yesterday, the Grand Chamber of the European Court of Human Rights strongly and unequivocally condemned the Italian policy of intercepting migrants’ boats in the Mediterranean sea and returning their unidentified passengers to Libya (Hirsi Jamaa and others v. Italy, 23 February 2012, available here; Hirsi hereinafter). And rightly so – the policy was such a flagrant breach of Italian and European obligations concerning international protection from refoulement that most observers and protection institutions had already unanimously expressed concern. To name but a few, the United Nations High Commissioner for Refugees, the Council of Europe’s Committee for the Prevention of Torture, the European Commission, Amnesty International, and Human Rights Watch had all previously said that such ‘push-back’ operations in the high seas were illegal both under international refugee law and under international human rights law (Hirsi, -).
The Hirsi case arose in the context of the 2007 bilateral anti-immigration cooperation agreement between Libya and Italy which was fully implemented in early 2009. When the policy of interception and rejection at sea was put into effect, it ostensibly achieved some of its stated aims. The Lampedusa ‘reception centre’ was suddenly nearly empty – a very different image compared to previous years (and subsequent events of 2011). In 2009 alone, more than 800 Somali, Eritrean and Nigerian citizens were returned to Libya before even touching Italian soil. They were returned to Tripoli without regard for the fact that, as unwanted migrants in Lybia, many of them faced a real risk of torture, physical violence, rape, indefinite detention in overcrowded and unhygienic conditions, as well as further expulsion towards their countries of origins. (Hirsi, -).
The applicants in Hirsi were 11 Somali and 13 Eritrean citizens belonging to a group of about two hundred migrants intercepted on 6 May 2009 about 35 nautical miles south of Lampedusa by the Revenue Police (Guardia di Finanza) and swiftly transferred to Tripoli on Italian military ships. None of them was identified by Italian authorities before being handed over to Libyan authorities. When ordered to board the Italian military ships, the migrants were told that they were being transferred to Italy. Two of them died in unknown circumstances after their arrival in Libya. Some of the others were granted refugee status by UNHCR in Tripoli. One of them was granted refugee status in Italy after making a second successful attempt at crossing the Mediterranean sea. It seems undisputed that they were genuinely in need of international protection.
There are three crucial reasons why Hirsi is a fundamental decision. The first is that it clarified that the Court’s Soering jurisprudence on non-refoulement under Article 3 ECHR also applies in the high seas. This is bound to have important consequences not only for the debates over the interpretation of the term ‘jurisdiction’ in Article 1 ECHR, but also, arguably, for the analogous long-standing debate over the applicability of non-refoulement obligations arising from the Geneva Convention on Refugees in the high seas. Secondly, it is only the second time (after Čonka v. Belgium, no. 51564/99, 2002) that the Court finds a State to have breached Article 4 of Protocol 4 ECHR, which prohibits the ‘collective expulsion of aliens’. In order to reach such a conclusion, the Court had to interpret the term ‘expulsion’ somewhat counter-textually. Thirdly, Hirsi stands as a landmark judicial reaffirmation of the long-standing jurisprudence of the Court on the protection of migrants from the risk of torture and inhuman treatment, and must be read in the context of European migration policy. I will very briefly address these three questions in turn.
I have elsewhere discussed the scope and consequences of international legal obligations on refoulement. There are many sources of international law, both customary and in treaties, that impose a duty not to expel, return, extradite or refuse access to a state’s territory to anyone whose most fundamental human rights (including the right not to be persecuted) would otherwise be at risk. We can find such prohibitions of refoulement both in international refugee law (most notably, Article 33 of the Geneva Convention on Refugees) and in international human rights law (e.g., Article 3 of the Convention Against Torture). However, it would be a mistake to blur all these obligations together into one general ‘principle’ of non-refoulement – when it comes to seeking remedies, it is often better to hold states accountable for each of their obligations separately.
In the context of the European Convention on Human Rights, the Court has consistently found that Article 3 must be interpreted as containing such a duty of non-refoulement, as the Grand Chamber reiterated in Hirsi:
113. According to the Court’s established case-law, Contracting States have the right, as a matter of well-established international law and subject to their treaty obligations, including the Convention, to control the entry, residence and expulsion of aliens […]. The Court also notes that the right to political asylum is not contained in either the Convention or its Protocols […].
114. However, expulsion, extradition or any other measure to remove an alien may give rise to an issue under Article 3, and hence engage the responsibility of the expelling State under the Convention, where substantial grounds have been shown for believing that the person in question, if expelled, would face a real risk of being subjected to treatment contrary to Article 3 in the receiving country. In such circumstances, Article 3 implies an obligation not to expel the individual to that country […].
But is this obligation also triggered in the high seas? The Italian government argued that the applicants were not under the ‘absolute and exclusive control’ of Italy, and that the operation was one of ‘rescue’ at sea, not a ‘maritime police operation’ (Hirsi, [64-65]), so the applicants would not be under the jurisdiction of Italy for the purposes of Article 1 ECHR. The Court, after recalling its recent Al-Skeini precedent (on which see here), held that the ‘nature and purpose of the intervention of the Italian ships on the high seas’ was wholly irrelevant to the application of Article 1. What mattered was that the applicants were in a military ship flying the Italian flag and fully ‘within the jurisdiction’ of Italy when they were handed back to Libyan authorities (Hirsi, [76-82]):
79. Moreover, Italy cannot circumvent its “jurisdiction” under the Convention by describing the events at issue as rescue operations on the high seas. In particular, the Court cannot subscribe to the Government’s argument that Italy was not responsible for the fate of the applicants on account of the allegedly minimal control exercised by the authorities over the parties concerned at the material time.
80. In that connection, it is sufficient to observe that in the case of Medvedyev and Others, cited above, the events at issue took place on board the Winner, a vessel flying the flag of a third State but whose crew had been placed under the control of French military personnel. In the particular circumstances of that case, the Court examined the nature and scope of the actions carried out by the French officials in order to ascertain whether there was at least de facto continued and uninterrupted control exercised by France over the Winner and its crew (ibid, §§ 66 and 67).
81. The Court observes that in the instant case the events took place entirely on board ships of the Italian armed forces, the crews of which were composed exclusively of Italian military personnel. In the Court’s opinion, in the period between boarding the ships of the Italian armed forces and being handed over to the Libyan authorities, the applicants were under the continuous and exclusive de jure and de facto control of the Italian authorities. Speculation as to the nature and purpose of the intervention of the Italian ships on the high seas would not lead the Court to any other conclusion.
82. Accordingly, the events giving rise to the alleged violations fall within Italy’s “jurisdiction” within the meaning of Article 1 of the Convention.
Once it was established that the obligation of non-refoulement also applied in the high seas because the applicants were ‘within the jurisdiction’ of Italy under Article 1 ECHR, the case was decided by the Court in a straightforward manner. Italian authorities ‘knew or should have known that, as irregular migrants, [the applicants] would be exposed in Libya to treatment in breach of the Convention and that they would not be given any kind of protection in that country’ (Hirsi, ). It was well known that ‘no rule governing the protection of refugees was complied with by Libya’ in the period under consideration (Hirsi, ). There was also a real risk that they would be returned to Eritrea and Somalia, where they would also be at risk of torture or inhuman/degrading treatment (Hirsi, [139-158]). Thus, there was indeed a violation of Article 3 ECHR.
Readers may wish to contrast this decision with the notorious US Supreme Court judgment of 1993 (Sale v. Haitian Centers Council, Inc., 509 US 155) which held that the protection from non-refoulement afforded in Article 33 of the Geneva Convention on Refugees was not applicable in situations of interception at sea (in that case, of Haitians trying to make their way to the US). Many commentators have since criticised that opinion, and the United Nations will no doubt be relieved that a different approach to extra-territorial application of human rights treaties was adopted by the European Court of Human Rights (for the UNHCR position, see generally here). True, the question of the extra-territorial application of Article 33 GCR must be kept separate from that of Article 3 ECHR, but it would be disingenuous not to recognize that Hirsi will be used as an important authority also in those debates.
What is particularly interesting here is that the European Court chose to solve the question of extra-territoriality also by reference to Italian domestic law (the Italian Navigation Code), which provides that a military ship flying the Italian flag is ‘Italian territory’. This became especially relevant also when construing the other Convention provision which Italy breached, that on the collective expulsion of aliens.
On the face of it, the provision in Article 4 of Protocol 4 (‘Collective expulsion of aliens is prohibited’) seems to refer to the expulsion of foreigners from the territory of a Contracting party. Could this obligation also apply in the high seas, in a situation where, as the Italian government put it, the migrants had not even ever crossed the Italian border? (Hirsi, ) In order to reach the conclusion that Article 4 P-4 did apply, the Court referred to the criteria in the Vienna Convention on the Law of Treaties (Hirsi, [170-175]), and then adopting a robust ‘evolutive’ interpretation:
177. […] The purpose of Article 4 of Protocol No. 4 is to prevent States being able to remove certain aliens without examining their personal circumstances and, consequently, without enabling them to put forward their arguments against the measure taken by the relevant authority. If, therefore, Article 4 of Protocol No. 4 were to apply only to collective expulsions from the national territory of the States Parties to the Convention, a significant component of contemporary migratory patterns would not fall within the ambit of that provision, notwithstanding the fact that the conduct it is intended to prohibit can occur outside national territory and in particular, as in the instant case, on the high seas. Article 4 would thus be ineffective in practice with regard to such situations, which, however, are on the increase. The consequence of that would be that migrants having taken to the sea, often risking their lives, and not having managed to reach the borders of a State, would not be entitled to an examination of their personal circumstances before being expelled, unlike those travelling by land.
178. It is therefore clear that, while the notion of “jurisdiction” is principally territorial and is presumed to be exercised on the national territory of States …, the notion of expulsion is also principally territorial, in the sense that expulsions are most often conducted from national territory. Where, however, as in the instant case, the Court has found that a Contracting State has, exceptionally, exercised its jurisdiction outside its national territory, it does not see any obstacle to accepting that the exercise of extraterritorial jurisdiction by that State took the form of collective expulsion. To conclude otherwise, and to afford that last notion a strictly territorial scope, would result in a discrepancy between the scope of application of the Convention as such and that of Article 4 of Protocol No. 4, which would go against the principle that the Convention must be interpreted as a whole. Furthermore, as regards the exercise by a State of its jurisdiction on the high seas, the Court has already stated that the special nature of the maritime environment cannot justify an area outside the law where individuals are covered by no legal system capable of affording them enjoyment of the rights and guarantees protected by the Convention which the States have undertaken to secure to everyone within their jurisdiction […].
The highlighted words are very interesting in light of the Court’s recent decisions on extra‑territoriality (Al-Skeini, above, and Al-Saadoon, no. 61498/08, 2010), which seem to proceed in the direction of closing the gaps of protection which had seemingly been opened by cases such as Banković (no. 52207/99, 2001) and, for other reasons, Behrami (no. 71412/01, 2007).
Overall, the Court once again reaffirmed its commitment to Article 3 as one of the cornerstones of human rights protection in Europe. It clarified that ‘cooperation’ policies with countries on the Southern Mediterranean coast cannot be conducted at the expense of non-refoulement obligations. This is a clear blow to more than a decade of ill-conceived migration policies in Italy and Europe, which have progressively seen the borders of Europe artificially shifted as far away as possible from the mainland. In 2004 and 2005, the Berlusconi government would deport foreigners after their arrival in Lampedusa without even identifying them (see here). In 2009, the government tried to prevent their arrival altogether. In both cases, these were only the most egregious facets of a European-wide project of border management which finds its most recent implementation in the EU FRONTEX Regulation (Council Regulation (EC) no. 2007/2004 of 26 October 2004), whose implementation should probably be reconsidered in light of Hirsi.
One should not be surprised by the reaction of the Italian Minister of the Interior of 2009, Roberto Maroni, who immediately declared that Hirsi was an ‘incomprehensible’ and ‘political’ decision of a ‘politicised Court’. Government lawyers had somewhat indelicately declared to the Court that the Hirsi applicants were engaged in ‘a political and ideological diatribe […] against the action of the Italian Government’ (Hirsi, ). In that regard, it seems that Italian politicians and their legal advisors rediscover arguments about the political nature of international law only when the policies of the Italian government contradict the law. While there is no doubt that the European Court of Human Rights is, in a sense, a political actor, so is every other ‘constitutional’ court in the world. Unfortunately, one cannot take international law seriously only when convenient – say, in order to ‘protect one’s sovereignty’ and request that Italian soldiers be deemed exempt from criminal courts in India.
(The author should disclose that he was the Refugee and Migrants’ Rights Coordinator of the Italian Section of Amnesty International between 2004 and 2006. He is obviously solely responsible for the opinions expressed here).