02 Mar Not All that Glitters Is Gold: the Human Rights Committee’s Test for the Extraterritorial Application of the ICCPR in the Context of Search and Rescue Operations
Paolo Busco is a member of Twenty Essex Chambers, where he practices in the field of public international law. All opinions are expressed in a personal capacity only.
On 27 January 2021 the Human Rights Committee published its decisions on two communications submitted in 2017 respectively against Malta (S.A. and others v. Malta, Communication No. 3043/2017), and Italy (S.A. and others v. Italy, Communication No. 3042/2017), concerning a tragic shipwreck that occurred in the Mediterranean Sea. The approach on the extraterritorial application of the ICCPR followed by the majority raises some interesting questions, both at the legal and at the policy levels.
The facts date back to 2013 and are reported in the decisions. In summary: on 11 October 2013, a vessel with hundreds of migrants that had set off from the coasts of Libya found itself in distress on an area of the high seas within the Maltese search and rescue zone. Those on board contacted the Italian Maritime Rescue Coordination Centre and were told that they should phone the Rescue Coordination Centre of Malta, given the location of the vessel. When they managed to make contact with Maltese authorities, they were informed that their vessel had been identified and that rescue units would soon arrive. The decisions go on to describe an appalling series of mistakes and possibly even deliberate actions, characterized by delays, coordination failures and attempts to pass off responsibility for the rescue operations. The decisions report that Italian authorities failed to provide any assistance until after the sinking ship capsized, despite Italy having naval units closer to the area of the shipwreck than Malta; and that Maltese authorities ran into a series of delays in identifying the sinking ship, and in coordinating and finally dispatching the rescue units. The shipwreck resulted in the tragic death of over 200 people, including many children.
The Committee found that Italy violated the right to life of those involved in the shipwreck, in breach of Article 6 of the ICCPR. The communication against Malta was found to be inadmissible, as the applicants did not exhaust local remedies. The two decisions, for what is relevant for the present analysis, must be read closely together.
Italy and Malta breached their obligations under UNCLOS, SOLAS and SAR
There is little doubt that conduct by Italy and Malta was in breach of fundamental provisions on search and rescue of UNCLOS, the 1979 International Convention on Maritime Search and Rescue (“SAR”) and the Regulations adopted pursuant to the 1974 International Convention for the Safety of Life at Sea (“SOLAS”). This is inexcusable. The question before the Committee, however, was only whether the ICCPR was breached. The preliminary question was of course whether the ICCPR applied to the events of 11 October. It is this question that I discuss here.
The test for the extraterritorial application of the ICCPR
There is an established jurisprudence that human rights treaties apply extraterritorially. Typically, this occurs when a State exercises jurisdiction abroad through effective control over territory, or exercises power and authority over an individual (General Comment, No 31 § 10). In the traditional sense, this involved occupation of territory (“spatial model”) or exercise of physical authority over persons (“personal model”); however, human rights bodies have rightly expanded the test to cover additional forms of exercise of authority by States that do not fall squarely into the two canonic models (see, however, the revirement in the recent decision by the ECHR in Georgia v Russia, discussed here). In addition, in 2018, the Committee adopted General Comment No 36, concerning the right to life under Article 6. According to § 63, persons are subject to the jurisdiction of a State for purposes of Article 2(1) of the ICCPR when that State “exercises power or effective control” over their “enjoyment of the right to life” (some authors have called this the “control over rights” doctrine). For the Committee, this brings under the State’s jurisdiction “persons located outside any territory effectively controlled by the State, whose right to life is nonetheless impacted by its military or other activities in a direct and reasonably foreseeable manner” (“impact model”). General Comment No. 36 is an articulation of the “functional approach” to jurisdiction: every time that a State has the power to exercise functions that have an impact on human rights of individuals in a direct and reasonably foreseeable manner, those individuals are within the jurisdiction of the State. More concretely: States must protect human rights in situations where they can do so.
The decision regarding Malta
In the decision regarding Malta, the Committee decided that the jurisdictional link existed on the basis that the ship was in the Maltese search and rescue region, and that in this area Malta has undertaken responsibility to provide for overall co-ordination of search and rescue operations, in accordance with SAR and SOLAS (Decision regarding Malta, § 67). The Committee also referenced the fact that Malta formally accepted to assume the coordination of the rescue efforts and that rescue operations had commenced.
In my view, from a legal perspective, this factual circumstance should have been the critical element in the reasoning of the Committee – both under more traditional tests for effective control, and under the functional approach. However, this does not appear to have been the determinative criterion. Rather, language used by the majority suggests an element of automatic consequentiality in the finding of jurisdiction: the individuals on the vessel “were within the Maltese search and rescue region and thus (…) subject (…) to the jurisdiction of Malta” (Decision regarding Italy, § 7.8). Some of the dissenting opinions reinforce this perception when they indicate that the SAR is “an area for which [a] State has assumed legal responsibility – and by implication, jurisdiction – under the law of the sea for search and rescue operations” (Dissenting opinion of Shany, Heyns and Pazartzis in the decision concerning Italy, § 5).
This is problematic in my view, because it is close to suggesting that the obligations undertaken by States in respect of their SAR area are able, in and of themselves, to ground an ipso facto control and hence a “presumption of jurisdiction” over situations occurring in the SAR region, dispensing from the requirement of assessing any actual control (including under the “impact model”) in the specificity of each case. As noted by Prof. Zimmerman, the decisions “will be perceived as providing for a general applicability of the Covenant as far as persons are concerned that find themselves in distress at sea (…) in the SAR zone of a State party” (Dissenting opinion of Zimmerman in the decision concerning Italy, § 4). However, a generalized “presumption of jurisdiction” of this sort would only exist “throughout a State’s territory” (Al-Skeini, § 131). The notion of jurisdiction in law of the sea conventions and the obligations contained therein lay the preconditions for the exercise of jurisdiction in the human rights sense of the term. However, the two notions of jurisdiction are not one and the same (Dissenting opinion of Zimmerman in the decision concerning Italy, § 6 – 7). For a finding of extraterritorial jurisdiction in the human rights sense, a fact-specific assessment on whether a State exercised actual control remains necessary (Papastavridis, p. 420). The key question is the extent of the necessary control – and, of course, under the “impact model” and “functional model” or other constructive approaches to jurisdiction (Trevisanut), the threshold may be lower than under other models.
The decision regarding Italy
How much lower was a question discussed in the decision adopted in respect of Italy. The Committee reasoned that a “special relationship of dependency” was established between Italian authorities and those on board the vessel. This was predicated on legal and factual circumstances, including: a) the duty to cooperate in search & rescue operations and to save life at sea, which binds all States regardless of where a shipwreck occurs; b) the fact that Italian authorities answered the first call from the vessel, and indicated to those on board that they would be rescued (without, however, indicating that Italian authorities were taking responsibility for the rescue); c) and the fact that an Italian naval unit – about one hour away – was closer to the sinking vessel than any Maltese units (Decision regarding Italy, § 7.8). On this basis, the Committee concluded that “the individuals on the vessel in distress were directly affected by the decisions taken by the Italian authorities in a manner that was reasonably foreseeable” and found that they were under Italy’s jurisdiction (Decision regarding Italy, § 7.8). While this approach is certainly functionalist in nature, it arguably goes beyond functionalism, establishing jurisdiction on the basis of a very low – in fact hypothetic – level of control, including under the “impact model”. It is perhaps telling that Prof. Shany, the theorist of the functional approach to jurisdiction endorsed in General Comment No. 36, was among the dissenters. In his (and others’) view, the majority failed to distinguish situations in which States have the potential to place individuals under their effective control, from situations involving the actual placement of individuals under effective State control (Dissenting opinion of Shany, Heyns and Pazartzis in the decision regarding Italy, § 5). The low threshold applied to establish control and jurisdiction is all the more difficult to reconcile with the Committee’s presumption of jurisdiction and control by Malta in its search and rescue zone.
The decisions by the Committee are interesting also from the policy perspective. The decision adopted in respect of Malta, and the suggestion of de jure control and ipso facto jurisdiction over events occurring in a State’s SAR zone may be a welcome policy development, despite the legal problems discussed above. This will hopefully promote more responsible conduct by States in performing search & rescue operations in their own SAR regions, or result in decisions to relinquish to other States portions of SAR zones that they are not able to manage. In the relationship between Italy and Malta, this has always been a topic of significant contention.
On the other hand, the decision adopted in respect of Italy, and the way the functional test has been applied outside a State’s SAR zone, raises some policy concerns. The crafting of a “special relationship of dependency” based mainly on factual circumstances such as having answered a distress call, or having kept radio contact with a sinking vessel for purposes of assisting a State with primary responsibility for rescue operations, may act as a disincentive for States to perform these critical activities. Several dissenting opinions noted this risk, emphasizing how the solution adopted by the Committee may end up “disrupt[ing] the legal order which the SOLAS and SAR Conventions attempted to introduce” (Dissenting opinion of Shany, Heyns and Pazartzis in the decision concerning Italy, § 6) and that “State parties of the Covenant (…) might even try to avoid coming close to vessels in distress so as to avoid any impressions of a “special relationship of dependency” having been created (Dissenting opinion of Zimmerman in the decision concerning Italy, § 6 – 7). It is interesting that on the same days the decisions by the Committee were published, the ECHR highlighted the risk of a chilling effect on important State activities (such as prosecutions of crimes committed abroad) that may derive from an excessively expansive approach to jurisdiction (Hanan v Germany § 135, and joint partly dissenting opinion of Grozev, Ranzoni and Eicke, § 5).
These observations also raise a more general question: whether the functional approach to jurisdiction and the impact model, quite apart from how the Committee applied them, are conceptually valid in the context of search and rescue operations occurring outside a State’s own SAR zone. These tests, as articulated in General Comment No 36, presuppose a State’s performance of an activity, and do not on their face regulate failure to perform an activity (General Comments 36 speaks of rights “impacted by a State’s military or other activity”). In most cases, a State’s interest in performing a certain activity will outweigh the perceived risk of triggering the test for the application of the Covenant. In these cases, it is not conceptually problematic to establish jurisdiction on the basis of the foreseeable impact of those activities (e.g. military operations, occupation of territory, etc.) on enjoyment of human rights. In other cases, the calculus may be different. Recent history demonstrates a regrettable reluctancy to perform search and rescue operations, mostly because of the implications on disembarkation of migrants rescued at sea. In these cases, the application of a very low jurisdictional threshold for purposes of human rights responsibility may risk being yet another disincentive. Human rights law, rather than being the enforcement mechanism of search and rescue obligations that some would like it to be, may result in even less compliance with these obligations.
A test for effective control that does not risk inhibiting search and rescue, coupled with the adoption of clear and fair rules on disembarkation, would probably be the best way to protect life and prevent tragedies like those addressed by the Committee in its recent decisions.