Duty to Allow Disembarkation of Rescued Refugees: Floating Between Sovereignty and Humanitarianism

(1)

Duty to Allow Disembarkation of Rescued

Refugees: Floating Between Sovereignty

and Humanitarianism

Master’s thesis

University of Amsterdam

Amsterdam Graduate Law School

Author: J.W.J. (Jasper) van Berckel Smit Student number: 10728686

Master track: Public International Law (International and European Law) Section: International Refugee Law

Supervisor: mw. prof. dr. M.Y.A. (Marjoleine) Zieck Date of submission: 19 July 2019 

(2)

Abstract

Increasingly, disembarkation of rescued refugees has been denied by Mediterranean States - re-luctant to cope with a full responsibility. As these incidents also reveal inconsistent views by var-ious stakeholders on the premise of the duty to allow disembarkation, the discussion concerning interpretation of this obligation requires a clear analysis of the legal framework, including the consideration of present-day reflections. To that end, this paper critically assesses to what extent States are obliged to allow disembarkation of rescued refugees on their territory.

Shipmaster’s obligation to provide help at sea forms the heart of the SAR regime. This duty is inevitably related to the shared obligation of States to co-ordinate and co-operate in order to release the master of his duties, by means of disembarkation to a place of safety. Yet, a place of safety does not necessarily have to refer to a place on the (land) territory of a State. Moreover, since this obligation has a divided character instead of an exclusive one, States will point at one another in their attempt to avoid the obligation. Additionally, the assurance of a special responsi-bility upon the State responsible for the search and rescue region (SRR) does not mean a swift disembarkation, as the time-frame issue of disembarkation is unclear. The temporal component of disembarkation is variable, depending on reasonableness and a case by case approach. Ac-cordingly, the State responsible for the SRR retains discretion, which could amount to survivors being held on an assisting vessel for days, even weeks, before disembarkation.

As the foregoing makes you wonder whether the object and purpose of the SAR regime are still served, this paper provides for a suggestion.Accordingly, this paper suggests the estab-lishment of a fair underlying relocation scheme in order to accurately live up to the priorities of the SAR system. 

(4)

1. Introduction

“All honour to the brave (…), who, at peril to their own, save the lives of the perishing, at sea!” 1

In the past, those who stayed behind held their hearts when a ship left for an overseas destina-tion. They hoped for mild weather and sea conditions, so their loved ones would return home safely. Nowadays, while technical improvement makes us better prepared to conquer the relent-less waves, the sea has retained its perilous nature. Especially when unsafe conditions aboard occur, due to an overcrowded or ill-prepared vessel, the dangerous characteristics of the sea are likely to cause a threatening situation for the crew and passengers.

The latter is a reality faced by numerous refugees trying to cross the Mediterranean sea. Statistics on the matter are illustrative. Since 2014, 18,400people have died while attempting to cross the Mediterranean. Although one could observe a general decline in deaths at sea since the 2

peak in 2016 , a mere look at this dwindled number disregards recent data on the Central 3

Mediterranean which exhibits a relatively high ‘risk of death’ amongst seaborne refugees. 4 5

Moreover, the absolute number of fatalities at sea has significantly increased since the inaugura-tion of the Conte government (incl. Salvini) in Italy, whereas the total number of arrivals has steadily decreased. 6

The Aldine 1874.

1

IOM 2019.

2

Ibid; see also press release Council of EU, 15 October 2018; press release FRONTEX, 4 January 2019.

3

I.e. total number of attempted crossings divided by total number of deaths during those crossings. See

4

ISPI, October 2018.

Fn 2. In 2018 (January-June), risk of death: 2.8% (10.575 attempted crossings, 359 deaths). In 2019

5

(January-June), risk of death: 5,1% (6,709 attempted crossings, 343 deaths). Attempted crossings include interceptions by Libyan Coast Guard.

ISPI, October 2018; see also press release UNHCR, 17 October 2018.

(5)

A probable explanation for the elevated fatality rate could be found in arguments relating to the search and rescue capacity in the Central Mediterranean. Search and rescue is based on the obligation to help the ones in need at sea, which is an age-old moral virtue and has been pro-claimed by the International Law Commission (ILC) to form part of customary international law. Moreover, Article 98 of United Nations Convention on the Law of the Sea (UNCLOS) has 7

codified the duty for Flag States to render assistance. Pursuant to the article's second paragraph, 8

Coastal States shall establish effective search and rescue (SAR) services in order to guarantee safety of life at sea.

Yet, the SAR services in the Central Mediterranean reduced.9 Crossing the central part of the Mediterranean became increasingly dangerous after EU Member States terminated the ‘Mare Nostrum’ rescue-operation and replaced it by Frontex's border control operation ‘Triton.’ While 10

operation ‘Mare Nostrum’ was predominantly aimed at rescuing persons in distress at sea, its 11

successor has the primary objective to strengthen border surveillance and control and is not ex-plicitly established as a SAR operation. The operation-shift thus caused an humanitarian aper12

-ture to be filled in. 13

In 2015, faced with an ever growing number of migrant deaths at the maritime borders of the EU, several NGO ships started conducting rescue operations in the Central Mediterranean. In the absence of state-led rescue programs, NGOs such as Médecins Sans Frontières (MSF),

Papanicolopulu 2016, p 494; Proelss 2017, pp 726-727; Barnes 2010, p 134; Goodwin-Gill 2007, p 278;

7

Oxman, Colum J of Transn L 1998, p 415; Nordquist 1995, p 171. Art 98(1) UNCLOS.

8

Press release UNHCR, 17 October 2018.

9

EC factsheet, 4 October 2016; FRONTEX is the Border and Coast Guard Agency of the European

10

Union.

This naval and air operation was a reaction by the Italian Government to the growing migration influx

11

during 2013. See info Mare Nostrum.

Andersson 2014; Gammeltoft-Hansen & Nyberg Sorensen 2013.

12

European Commission President Juncker, later acknowledged in a public speech: “it was a serious mis

13

(6)

Watch and SOS-Méditerranée played a life saving role. These organisations were highly essen14

-tial to the preservation of life at sea, as numbers of 2016 illustrate. They rescued 46,795 people and thus contributed for 26% in all rescue operations of that year. In comparison, the Italian 15

Navy conducted 21% of the rescues in 2016, the Italian Coast Guard 20% and EU NAVFOR Med 17% respectively. 16

However, contra intuitively, while one would suggest the foregoing humanitarianism pro-vokes widespread appreciation, rescue-service providers also received substantial criticism. For instance, it has been argued that the towering death toll in the Mediterranean Basin is caused by a growing number of rescue services, as additional operations allegedly serve as ‘pull factor’ for illegal migration. Further operations cause people smugglers to send larger numbers of mi17

-grants to Europe in increasingly unseaworthy vessels. This reasoning has been expressed inter 18

alia by Frontex , which accused NGOs for colluding with smugglers. Furthermore, similar 19 20

arguments were used by EU Member States to criminalise rescuing NGOs as well as other volun-teers, which manifested in seizure of vessels and detention of persons. 21 22

Report Arsenijevic, Manzi & Zachariah 2017; Report Steinhilper & Gruijters 2017; Report Heller &

14

Pezzani 2017; MSF Issue Brief 2017. Italian Coast Guard 2016.

15

MSF Issue Brief 2017.

16

See e.g. statement Government United Kingdom. The Guardian, 27 October 2014; see for discussion in

17

literature Del Valle, Ref Survey Q 2016; ECFR, 10 November 2015; Cusumano, Marine Policy 2017. See e.g. The Spectator, 6 September 2014; German Interior Minister claimed Mare Nostrum was “als

18

Brücke nach Europa”, Süddeutsche Zeitung, 31 October 2014.

Frontex’ confidential communication was obtained. See FT, 14 December 2016.

19

MSF, 23 December 2016.

20

See e.g. Iuventa, of the German NGO Jugend Rettet, that was seized (2 August 2017) by the Italian ju

21

-diciary under suspicion of ‘assistance to illegal migration’ and collusion with smugglers during three dif-ferent rescue operations. The seizure came only days after the NGO had refused to sign the 'Code of Con-duct’. Report Heller & Pezzani 2017; The Intercept, 20 April 2018; De Zeen, 20 March 2018; Stelka Mag, 21 March 2018; see also seizure of NGO ProActiva Open Arms. ANSA, 19 March 2018.

See e.g. accusations by Greece authorities of volunteers/fundraisers for people smuggling, espionage

22

and money laundering. HRW, 5 November 2018; See also Tunesian fishermen detained after disembarka-tion in Italy of 14 rescued migrants. Il Fatto Quotidiano, 4 November 2018.

(7)

Not surprisingly, human rights organisations outcried in response, as they claimed the intensified hostility towards rescue operators increased the number of fatalities at sea and there-fore should be stopped. The coin has another side though. Despite the fact that the there-foregoing 23

accusations are controversial , one should bear in mind the underlying political reasons affecting 24

the European asylum apparatus when trying to understand the hard stance of the EU Member States. Since the beginning of 2015, EU Regulation 604/2013, on determining the Member State responsible for examining an application for international protection (Dublin III Regulation), showed significant structural shortcomings, in particular the unfair distribution of responsibili-ties. Therefore, a reformation process was set in motion , in order to streamline and supple25 26

-ment the current rules with a ‘corrective allocation mechanism.’ If a EU Member State faces a disproportionate number of asylum-seekers, this mechanism would guarantee either way the re-location of people, or a ‘solidarity contribution’ per applicant. 27

Yet, hitherto the European Council has been unable to reach a consensus on Dublin Regu-lation reform. Particularly the Visegrad group has strongly rejected the proposals on the relo28 29

-cation mechanism, based on a balance of responsibility and solidarity. Moreover, the likelihood 30

of a prompt breakthrough is reducible, as recent negotiations have refuted to be fruitful. Conse31

Report Amnesty 2018, pp 14-16; see also DW, 8 December 2018.

23

The ‘pull factor’ claim is contentious, since the growing number of rescue operations cannot be deemed

24

as ‘pull factor’ for migration across the Mediterranean, given the lack of significant causality. ISPI, May 2018; With regards to the Greece accusations, five volunteers were cleared from charges. NYT, 7 May 2018.

Deficiency affirmed by Report EC 2015; see also COM(2016) 197.

25

Note proposal by EC, COM(2016) 270.

26

EPRS briefing, 1 March 2019.

27

Ibid.

28

Group composed of Poland, Hungary, Czech Republic and Slovakia.

29

For concerns about corrective allocation mechanism, see FT, 12 December 2017.

30

Progress report Council of the EU, 26 February 2019.

(8)

quently, as long as there is no fair relocation-mechanism in order to equally distribute the influx of refugees and asylum seekers among Member States, the anti-immigration rhetoric used by Eu-ropean ‘first-entry-States,’ such as Greece and Italy, is to some extent comprehensible. 32

The tension between the impossibility to cope with the full immigration burden in the ab-sence of solidarity - sovereignty issue- on the one hand, and the urgent need for increasing res-cue activities – humanitarian issue- on the other, takes a central place within this paper. The fric-tion between these approaches does not merely emerge with regards to NGO resuce operafric-tions, but forms a prominent connotation of the SAR regime as a whole, when confronted with migra-tory flows overseas. The infamous MV Tampa incident illustrates the latter point. After rescuing 33

hundreds of asylum seekers in the Indian Ocean, the Norwegian container ship found itself in an awkward position in between states that were reluctant to allow disembarkation on their territory. Despite the recommendation issued by UNHCR to share the burden between Norway (flag State) and Australia (coastal State), the latter agreed upon disembarkation of the survivors in New Zea-land, Nauru and Papua New Guinea. 34

Yet in this paper, the centre of attention is on NGO rescue activities in the Central Medi-terranean sea, as this relatively new and undiscovered phenomenon progressively triggers the friction between sovereignty and humanitarian concerns. Therefore it is interesting to see whe-ther the existing SAR regime is flexible enough to cope with large numbers of seaborne refugees in conjunction with rescue missions conducted by private actors. In this paper the anchor rests on the last part of the SAR regime – to be specific, the purported legal duty to allow disembarkation of the rescued refugees. The contemporary discussion on this alleged legal duty requires scrutiny, as the number of instances including a rejection of disembarkation has recently spiked.

See for discussion on burden sharing in relation to safety at sea, Klug, Intl J Ref L 2014.

32

See discussions of Tampa case by Bailliet, HRQ 2003; Fornari, Comunità Internazionale 2002; Math

33

-ew, Am J Intl L 2002.

See Judge North’s recitation of facts in MV Tampa case 2001.

(9)

rently, these incidents revealed inconsistent views by various stakeholders on the premise of a legal duty to allow disembarkation. 35

Accordingly, the legal discussion concerning interpretation of the duty to allow disem-barkation under International Law of the Sea requires a clear analysis of the legal framework, including the consideration of present-day reflections. The main purpose of this paper is to con-template the legal framework pursuant to International Law of the Sea, in order to critically an-swer the question to what extent States are obliged to allow disembarkation of rescued refugees on their territory. While addressing this central question, this paper uses a methodology that de-rives from classical legal research. The research determines what the law actually is, i.e. the legal framework concerning disembarkation of rescued refugees. However, this paper does not merely proceed from the internal perspective, as a suggestion is made on how the legal regime should look like in order to function well, which bears a rather normative character. Therefore this paper possesses a combination of natures: descriptive and prescriptive.

In the following, the legal framework of disembarkation is discussed (Chapter 2), starting from the duty to render assistance pursuant to UNLCOS (2.1). However, in order to detect the legal basis for the duty to allow disembarkation one is to find clarity in other legal sources (2.2). Based on Reg. 33(1-1) of the SOLAS Convention, a systematic assessment is made on the oblig-ations incumbent upon shipmasters, coastal States, port States and flag States, respectively (2.3). In accordance with the findings of the foregoing inquiry, concluding remarks concerning the duty to allow disembarkation of rescued refugees are made (3.1), as well as a further suggestion on how the SAR system should look like in order to function in conformity with its priorities (3.2).  

Note similarities with the MV Tampa case 2001, which equals the current incidents to the extent that

35

this case raised issues with regards to the question which State could be held responsible for providing the ultimate gangway on its land.

(10)

2. Legal Framework of Disembarkation

Introduction

In order to conduct an assessment of the duty to allow disembarkation of rescued migrants, the possible legal basis of this duty has to be identified. Therefore the relevant provisions under In-ternational Law of the Sea are considered in the following paragraphs, starting from the duty to render assistance to persons in distress by virtue of UNCLOS. Although the duty to render assis-tance forms one of the core principles, this analysis detects other international conventions that come closer to the heart of the disembarkation discussion. Accordingly, the scope of these con-ventions is scrutinised for its concrete contribution to the discussion of the duty to allow disem-barkation of rescued refugees.

2.1. Render Assistance

The obligation to provide assistance at sea has been codified in Article 98 of UNCLOS. Article 98 UNCLOS imposes two duties upon States. The first one is directed to every flag State and requires the master of the ship to render assistance to those in distress at sea. The second obliga-tion requires all coastal States to “promote the establishment, operaobliga-tion and maintenance of an adequate and effective search and rescue service.” The geographical scope of the duty to render assistance encompasses all maritime zones, even though Article 98 is primarily focused on the High seas. Additionally, a State cannot refuse to render assistance within its territorial sea, sim36

-ply by referring to his sovereign powers. However, the shipmaster’s duty to render assistance is 37

not unconditional. 38

Firstly, the literal wording of the text requires a master of a ship to take measures ‘so far as he can do so without serious danger to the ship, the crew, or the passengers.’ And he is

Pursuant to arts 58 (2) and 18 (2) UNCLOS, the duty to render assistance has effect in the EEZ and the

36

territorial sea.

See Article 18(2) UNCLOS and commentary in Nordquist 1993, p 162; Barnes, ICLQ 2004, pp 50–52.

37

Proelss 2017, p 727.

(11)

quired ‘in so far as such action may reasonably be expected of him.’ Therefore, the conditions provide for a requirement of the master of a ship to make a discretionary judgement based on all relevant circumstances of the case. As the shipmaster should consider his primary duty to en39

-sure the safety of the ship, crew and passengers, it might in general be complicated to establish a violation of the obligation to render assistance. Besides, paragraph two of Article 98 contains a 40

best effort duty (‘promote’) and therefore lacks a compelling nature. It merely stimulates States to set up SAR services, if necessary to be accomplished through regional arrangements with neighbouring States.

Nevertheless, the duty to render assistance reflects customary international law, as em-phasised by the International Law Commission (ILC).41Article 98 therefore “gives expression to the general tradition and practice of all seafarers and of maritime law regarding the rendering of assistance to persons or ships in distress at sea, and elementary considerations of humanity.” 42

Moreover, additional significant international conventions state the obligation to render as-sistance to persons in distress, inter alia: the 1974 International Convention on the Safety of Life at Sea (SOLAS Convention); the 1979 International Convention on Maritime Search and Rescue (SAR Convention); and the 1989 the International Convention on Salvage (Salvage Convention).

Following article 10 of the Salvage Convention, shipmasters have the obligation to assist any person in danger of being lost at sea, provided such action doesn’t prejudice the safety of the vessel or crew. The responsibilities pursuant to the SOLAS and SAR conventions are also simi43

-lar to UNCLOS. Additionally the latter two conventions amplify a paramount condition, subtly

Nordquist 1995, p 175.

39

Proelss 2017, p. 727.

40

II YB ILC 1956, pp 253, 281. ILC art 36 on duty to render assistance was based on three conventions:

41

(i) Salvage Convention, art 8; (ii) Collision Convention, art 8; (iii) SOLAS Convention, Annex, Ch V, Reg 10.

Nordquist 1995, p 171.

42

Salvage Convention art 10; note however that the Salvage Convention is primarily a private law in

43

-strument and might not be the most appropriate vehicle for addressing the public law issue of treatment of persons at sea. See MSC 76/22/13.

(12)

included in UNCLOS. These conventions provide for the obligation of shipmasters to proceed with all speed to the assistance of persons in distress at sea, regardless of the nationality or status of such persons or the circumstances in which they are found. The Council of the International 44

Maritime Organization (IMO) reiterated the non-discriminatory character of the rescue exer45

-cise. The UNHCR also emphasised “the fundamental obligation under international law for 46

shipmasters to rescue all persons, including asylum-seekers, in distress at sea.” Besides, con47

-trary to the best effort duty according to Article 98 paragraph 2 of UNCLOS, Contracting Gov-ernments have a concrete duty to ensure arrangements for coast watching and for the rescue of persons in distress at sea round their coasts by virtue of the SOLAS and SAR conventions. 48

But, what happens to the assisted persons after they embarked the rescue vessel? How far reaches the scope of shipmasters’ and coastal States’ obligations? Does a rescue operation in principle involve disembarkation of the rescued people to a safe place? Taking the aforemen49

-tioned provisions into account, it is not clear whether the duty to render assistance contains the obligation for states to allow disembarkation. As Article 98 UNCLOS does not provide for a clear cut answer, one is to find clarity in other legal sources within the Law of the Sea sphere, such as the conventions just mentioned above. To that end, a special interest goes out to the amended SAR and SOLAS conventions, as some their provisions specifically aim at the second phase of the SAR operation, viz. disembarkation. Hence, the relevant provisions within these two IMO conventions are analysed in-depth, drawing on the comprehensive interpretation technique of Articles 31 and 32 of the Vienna Convention on the Law of Treaties (VCLT).

Annex SOLAS Convention, Reg V/33.1; Annex SAR Convention, 2.1.

44

The IMO is the United Nations specialised agency with responsibility for the safety and security of

45

shipping and the prevention of marine and atmospheric pollution by ships.

C 54/17(d) (IMO Council); cited by 1985 Nordquist 1995, p 175; De Londras and Mullally 2017, p 59.

46

Addendum to Report UNHCR 1985, p 32.

47

Annex SOLAS Reg V/7; Annex SAR Convention, 2.1.

48

Papanicolopulu 2016, p 499.

(13)

2.2. SOLAS and SAR Amendments and Guidelines

In the wake of the MV Tampa incident in 2001, the IMO Assembly adopted a resolution on ‘Re-view of safety measures and procedures for the treatment of persons rescued at sea’, which rec-ommends reviewing measures and procedures for the treatment of rescued persons, ensuring that the life of persons on board ships is safeguarded at that coastal communities should deal with such people satisfactorily. This Resolution requested various IMO bodies to review selected 50

IMO Conventions to detect ‘any gaps, inconsistencies, ambiguities, vagueness or other inade-quacies’ within the SAR regime. To that end, the issues were to be approached in a holistic 51

manner, utilising information provided by UNHCR, as part of an inter-agency mechanism. Ac52

-cordingly, the Assembly kicked off the further development of conventions concerned with treatment of persons rescued at sea.

The IMO is an organisation with the responsibility of maintaining maritime conventions 53

- a competence that in general stems from its constitutive instrument and more in specific from 54

other relevant international maritime conventions. As part of the IMO, the Maritime Safety 55

Committee (MSC) deals with safety issues at sea. For that reason, talks on the treatment of per-sons rescued at sea took place within the MSC and subcommittees. After preliminary consid56 57

-eration of the Assembly’s requests, gaps in the SOLAS and SAR conventions were identified 58

IMO Res A.920(22).

50

Ibid, para 1 (under ‘REQUEST’).

51

Ibid, preambular paragraph (under ‘NOTING’); COMSAR 6/22, para 8.58; MSC 75/24, para 11.45;

52

COMSAR 7/23, paras 8.10, 8.13; MSC 77/26, paras 10.28-10.30. Rothwell 2015, p 421.

53

Art 2(b) IMO Convention.

54

E.g. Article VIII, SOLAS Convention; Article III, SAR Convention.

55

IMO MSC 75th, 76th and 77th session.

56

IMO COMSAR 6th, 7th and 8th session.

57

The consideration of SOLAS and SAR conventions was prioritised by the Committee, in MSC 75/24,

58

(14)

with regards to, inter alia, the obligations of shipmasters and States. Hence, a draft text was 59

established, which was considered a good basis for further development, as the carefully chosen wording reflected a well balanced compromise between the stakeholders. Following discus60

-sions on the proposed drafts, the MSC adopted amendments to Chapter V (‘Safety of navigation’) of the Annex of the SOLAS Convention. Similarly, Chapter 2 (‘Organization and 61

co-ordination’), 3 (‘Co-operation between States’) and 4 (‘Operating procedures’) of the Annex of the SAR Convention were amended by the MSC. 62

Amendments to both conventions are considered to have been accepted on 1 January 2006, as no more than one third of the State Parties have notified objections to the amendments. This is in accordance with the so-called tacit acceptance procedure. The number 63 64

of ratifications elucidate the global relevance of amendments to the Annex of the SOLAS Con-vention, as this convention has been ratified by 165 States, which reflects 99.04% of the world tonnage. Also the SAR Convention has a relatively high number of ratifications (113), reflect65

-ing 80.41% of the world tonnage. 66

Among the amended chapters of the annexes of the SOLAS and SAR conventions, this paper centres Regulation 33 paragraph 1-1 (Chapter V) of the SOLAS Convention (hereafter ‘Reg. 33(1-1) SOLAS’), because this paragraph pays particular attention to the duty to permit

COMSAR 6/22, paras 8.74 - 8.78.

59

MSC 76/22/11; COMSAR 7/23, paras 8.6, 8.17.

60

By virtue of art 28(b) IMO Convention, in conjunction with art VIII(b) SOLAS Convention.

61

Pursuant to art 28(b) IMO Convention and art III(2)(c) SAR Convention.

62

Only three and two objections have been communicated to the IMO concerning the amendments to the

63

SOLAS Convention and the SAR Convention, respectively. See Report on Status IMO treaties; IMO Res MSC.153(78); IMO Res MSC.155(78); art VIII(b)(vi)(2)(bb) SOLAS Convention; art III(2)(f) SAR Con-vention.

Rothwell 2015, p 421; See also for explanation of this procedure Adede, Va J Intl L 1977.

64

Status of treaties in IMO’s Global Integrated Shipping Information System (GISIS).

65

Ibid.

(15)

disembarkation. For that reason, the scope of this paragraph is examined under heading ‘2.3. Duty-bearers’, in the light of its object and purpose. Note however that by means of harmonising the obligations with regards to treatment of persons rescued at sea, Reg. 33(1-1) SOLAS and paragraph 3.1.9 of the SAR Convention possess a similar wording and also share the same draft-ing history. Therefore, when this paper refers to Reg. 33(1-1) SOLAS in the followdraft-ing, one is 67

to read paragraph 3.1.9 of the SAR Convention.

Apart from the discussed amendments, the non-binding MSC Resolution 167(78) sets out the Guidelines on the treatment of persons rescued at sea, as referred to in Reg. 33(1-1) SOLAS. The Guidelines provide guidance to coastal States and shipmasters with regard to the SAR 68 69

legal framework. They are intended to assist Governments and masters to better understand 70

their obligations under international law and provide helpful guidance concerning compliance with these obligations. Moreover, the Guidelines do not merely serve as directive for ad71

-dressees, but they actually hold essential interpretative value for the scope-assessment of Reg. 33(1-1) SOLAS.

Although guidelines have in principle a ‘soft law’-character within the international legal realm, the present Guidelines differ in the sense that the second sentence of Reg. 33(1-1) SOLAS (see under heading ‘2.3. Duty-bearers’) refers in its text to ‘guidelines’, as a factor to be taken into account. In spite of the lack of concrete reference , one should bear in mind that the Guide72

-lines have been adopted simultaneously with the amendment the SOLAS Convention. Further-more, throughout the drafting history, multiple remarks have been made on the Guidelines as

Acknowledged by delegations in COMSAR 7/23, para 8.25.

67

Where term (Member/flag/coastal/port) ‘State’ is used, it should be read to mean ‘Contracting Govern

68

-ment’ to the SOLAS Convention, or ‘Party’ to the SAR Convention, respectively.

Read: shipmasters, flying the flag of a Contracting Government or Party (see previous fn).

69

IMO Res MSC.167(78), 1.1.

70

Ibid, 1.2.

71

The SOLAS and SAR paragraphs refer to ‘guidelines’, without explicitly mentioning the name or char

72

(16)

flecting a harmonised interpretation, as well as being a necessary supplement to the amended ar-ticles of the SAR and SOLAS. Accordingly, the Guidelines are intended to be regarded an inte73

-gral part of Reg. 33(1-1) SOLAS and provide content to the legally binding obligations set 74

forth in this regulation. 75

2.3. Duty-Bearers

Reg. 33(1-1) SOLAS:

“[States] shall co-ordinate and co-operate to ensure that masters of ships providing assistance by embarking onboard persons in distress at sea are released from their obligations with minimum further deviation from the ships’ intended voyage, provided that releasing the master of the ship does not further endanger the safety of life at sea.

The [State] responsible for the search and rescue region in which such assistance is rendered shall exercise primary responsibility for ensuring such co-ordination and co-operation occurs, so that survivors assisted are disembarked from the assisting ship and delivered to a place of safety, taking into account the particular circumstances of the case and guidelines developed by the Organization.

In these cases the relevant [States] shall arrange for such disembarkation to be effected as soon as reasonably practicable”.

In the upcoming paragraphs an assessment is made of the obligations incumbent upon the ad-dressees of Reg. 33(1-1) SOLAS. Firstly, the shipmaster’s duties are examined. This is followed by the evaluation of obligations upon coastal States, port States, flag States and States responsi-ble for the search and rescue regions (SRRs). Various interpretative sources are utilised in order

E.g. MSC 77/26, para.10.26. Authoritative value of Guidelines from perspective of delegations is fur

73

-ther exemplified by statement Malta during MSC 78. Malta expressed its reservation on proposed amend-ed SAR and SOLAS articles, subject to proposamend-ed associatamend-ed Guidelines being appropriately amendamend-ed. See MSC 78/26, paras. 16.46-16.56; see also COMSAR 7/23, paras 8.22, 8.23, in which delegations stress importance of additional guidelines so to clearly understand obligations pursuant to Reg. 33(1-1), SO-LAS.

II YB ILC 1966, para 14, p 221; see also Ambatielos case 1952, p. 44.

74

Guidelines, para. 1.1.

(17)

to determine the scope of Reg. 33(1-1) SOLAS, pursuant to the rules of interpretation under the VCLT. Besides the literal wording , the Guidelines and the Travaux Préparatoires of Reg. 76 77 78

33(1-1) SOLAS form important points of reference. They represent an authentic interpretation by the parties of the ordinary meaning to be given to the terms. Only in certain conditions, the 79

VCLT prescribes to take into account the interpretative relevance of alternative applicable in-ternational law. However beware that systemic integration of other inin-ternational law provisions 80

could only be properly and adequately utilised when the contextual nuances in the relevant in-ternational conventions are duly taken into account. While conducting a scoping exercise on 81

Reg. 33(1-1) SOLAS specifically aimed at the duty-bearers, it should become clear to what ex-tent States possess a duty to allow disembarkation of rescued refugees.

2.3.1. Shipmasters

The most essential obligation incumbent upon shipmasters is that of rendering assistance to peo-ple in distress at sea, as discussed under heading ‘2.1. Render Assistance.’ But also subsequent to the initial rescue service, shipmasters have fundamental obligations to uphold. Besides taking good care of the survivors , shipmasters first and foremost must ensure that survivors are not 82

disembarked to a place where their safety would be endangered. This imperative is just as 83

much directed to shipmasters as it is to States and ensures survivors to be delivered to a 'place of safety’. According to the text of Reg. 33(1-1) SOLAS, a place of safety is not aboard the

Art 31(2), VCLT, as the authentic expression of the intentions of states, see AnnIDI 1956, p 359; II YB

76

ILC 1966, para 2, p 218. Art 31(3)(a), VCLT.

77

Ibid, Art 31(2)(a).

78

II YB ILC 1966, para 14, p 221.

79

Art 31(3)(c), VCLT.

80

Accordingly, Article 31(3)(c) VCLT “should not be understood as the legal basis upon which to align

81

the contents of a treaty with the contents of other treaties.” Rachovitsa, ICLQ 2017, p 588. Guidelines, paras 5.1.1-5.1.3.

82

Ibid, para 5.1.6.

(18)

ing ship. This comes as no surprise, since the provisions endeavour to release the master of its 84

obligation to take care of the rescued persons. Hence, the place of safety should be guaranteed by a coastal State. The Guidelines provide clarity with regards to the term:

“A place of safety (…) is a location where rescue operations are considered to terminate. It is

also a place where the survivors’ safety of life is no longer threatened and where their basic hu-man needs (such as food, shelter and medical needs) can be met. Further, it is a place from which transportation arrangements can be made for the survivors’ next or final destination”. 85

The Guidelines emphasise the fact that assisting vessels can only temporarily serve as a place of safety for the survivors , which implies the longer lasting character of a place of safety, as re86

-ferred to in Reg. 33(1-1) SOLAS. Nonetheless, the concept is not indefinite, as one should bear in mind that the place of safety should allow for the ability to make transportation arrangements for the survivors’ next or final destination. Note as well that the Guidelines do not exclude the 87

possibility of a place of safety to be on another vessel or other facility at sea, as long as it is not the initial assisting vessel itself. 88

The imperative upon shipmasters (as well as States) to deliver survivors to a place of safety is particularly interesting with regards to seaborne refugees. In the end, one of the essen89

-tial characteristics of a place of safety is its reflection of the non-refoulement principle - the core element of the international refugee protection regime. This is a well-established principle of in-ternational law in which States are prohibited from expelling or returning a refugee to the

Ibid, para 6.13. 84 Ibid, para 6.12. 85 Ibid, para 6.13. 86 Ibid, para 6.13. 87 Ibid, paras 6.13, 6.14. 88 Ibid, para 6.17. 89

(19)

tiers of territories where his or her life or freedom would be threatened. The non-refoulement 90

principle renders deliveries to unsafe States, such as Libya, unlawful. 91

2.3.2. Coastal States

Inevitably related to the obligation of the shipmaster to provide help at sea is the shared obliga-tion of States to release the master of his duties, by means of disembarkaobliga-tion to a place of safety. This obligation is subject to some conditions, namely the release cannot excessively retard the assisting ship and the release cannot deteriorate the survivors conditions. This paper subsequent-ly considers these prerequisites. But prior to a substantial discussion of the duty to release the shipmaster, it is important to stress that the first sentence of Reg. 33(1-1) SOLAS leaves the na-ture of the State involved (e.g. coastal or flag State) undetermined. All it says is that States are obligated to release the shipmaster of his humanitarian burden. As the first sentence of Reg. 33(1-1) SOLAS reflects a general State obligation to liberate the shipmaster, this duty is applica-ble to States in all their appearances - also as coastal State. 92

First of all, the first sentence of Reg. 33(1-1) SOLAS indicates the mandatory nature (“shall (…) ensure”) of the co-ordination and co-operation by the States. This sentence obliges to make something certain to happen , i.e. entails a duty of result. Namely, the co-ordination and 93 94

co-operation by coastal States should amount to the liberation of burden on the master’s side. The Travaux Préparatoires provide for the rationale of relieving the master by delivering

See treaty law, art 33.1 Refugee Convention; art 3 ECHR; and art 19(2) CFREU; see also arts 6 and 7

90

ICCPR; art 3(1) Convention Against Torture; art 2(3) OAU Convention; see also declaration of perempto-ry character of non-refoulement in UNHCR Conclusion on Non-Refoulement, para 70(1).

Hirsi Jamaa case 2012, para 36; Report HRW 2019; Report OHCHR 2018.

91

This conclusion was intended by the Member States. COMSAR 7/23, para 8.26; MSC 77/26, para

92

10.32; COMSAR 8/18, para 8.4; MSC 78/26, paras 16.48, 16.49; see also Guidelines, paras 6.3, 6.4. Meaning of “ensure”, Cambridge Dictionary.

93

Note difference with duty pursuant UNCLOS, which merely obliges coastal States to “promote”, art

94

(20)

vivors to a place of safety. From the preliminary stage of the drafting history onwards, the del95

-egations stressed the importance of avoiding the situation in which the shipmaster would be left alone with his humanitarian responsibilities. This predicament could discourage the shipmaster, 96

since the duty to rescue persons in need at sea would inevitably be associated with the loss of time and income and would likely create an undesirable incentive to the shipmasters. Conse97

-quently, this would mean the end of an effective humanitarian regime, as the aforementioned negative impetuses undermine the integrity and effectiveness of the SAR system.

Based on the literal wording of the text, disembarkation of the survivors to a place of safety is the only (sensible) way for coastal States to cease the shipmaster’s burden. Moreover, the Guidelines emphasise that a ‘place of safety' is a location where rescue operations are con-sidered to terminate. However note the practical consequences of an unpredictable place of dis98

-embarkation. Due to the establishment of a mere shared duty to allow disembarkation, coastal States will likely point at one another in their attempt to outrun the duty. Therefore, despite the 99

existence of a duty to allow disembarkation upon coastal States, “the use of the passive voice leaves unanswered the critical question of by which state rescued persons must be disembarked in the absence of agreement among affected States”. 100

The second remark in relation to the master’s relief concerns the extract “with minimum further deviation from the ships’ intended voyage” (first sentence of Reg. 33(1-1) SOLAS), which underlines the importance of mitigating the disturbance of the shipping industry. Although the objective is vivid - namely to reduce the cost of the virtuous act - , this extract provokes the

See identified basic principles in COMSAR 6/22, para 8.73.

95

Ibid.

96

Ibid, para 8.68; see for significant role of financial incentives, Davies, Pac Rim L Policy J 2003, p.

97

128–133; see also BIICL project on private vessel implications. Guidelines, para 6.12. See also Moreno-Lax, IJMLC 2010.

98

During the drafting process, Norway stressed this point emphatically. MSC 77/10/2.

99

Proells 2017, p 729; see also Moreno-Lax, IJMLC 2010.

(21)

question whether it matters when an assisting vessel does not have a clear cut intended voyage. What if the vessel does not have a clear destination, but instead sails around the seas in anticipa-tion of a rescue signal? When a private rescue vessel, e.g. a NGO ship, has not yet received an-other distress signal, would that make it defenceless towards disembarkation delays? The latter would be contrary to the intentions of the Member States, because the assisting vessels “should not be subject to undue delay, financial burden or other related difficulties after assisting persons at sea”. Yet, observe that the wording ‘should’ is hortatory and leaves some room for discre101

-tion.

Thirdly, the literal wording of the text attaches a humanitarian safeguard to the release of the master, namely the latter cannot imply the survivors to be abandoned at sea or to be replaced to an unseaworthy vessel. Thus, coastal States cannot solve the issue by simply shifting the prob-lem to another vessel lacking capacity or facilities. In spite the liberation of the shipmaster from his duty, a deterioration of the survivors’ circumstances is to be avoided at all times. This conclu-sion is confirmed by the Guidelines. 102

2.3.3. Port States

The recent stand-off between the Sea-Watch 3 vessel and the Italian authorities - which result103

-ed in the arrest of the captain of the NGO vessel and further diplomatic tensions between Italy and other EU Member States - raises questions as to what to do when private rescue vessels try-ing to enter port. Are port States obliged to provide access? Not the preparatory works, nor the Guidelines dwell on the duty to grant access to port, or any corresponding shipmaster right.

Guidelines, para 6.3.

101

Notice earlier discussion under heading ‘2.3.1. Shipmasters’, incl. references, with regards to delivery

102

to a place of safety.

DW, 29 June 2019; The Guardian, 29 June 2019.

(22)

Pursuant to international law, no general right of access to port exists. The port is con104

-sidered to be situated within the internal waters, which guarantees full sovereignty upon the port State. “Limitations to sovereignty cannot be presumed and there is no evidence of any limita105

-tions in State practice in relation to sovereignty over ports.” Yet, there is one broadly recog106

-nised exception to the wide discretion of port States, namely in case of a vessel in “distress”. This plight essentially requires a threatening situation to the ship or its crew/ passengers. Re107

-mark that distress is to be interpreted narrowly. Port States still have the power to deny access 108

when the requesting vessel poses a threat to its marine environment. Only in cases where the 109

access to port is needed for preservation of life, States are obliged to grant refuge. 110

In principle one could argue there is a high chance of deterioration of human health when a rescue vessel is not allowed to disembark for several days. Especially when there is a shortage of food, fresh water or medications, a stand-off situation could endanger life and limb onboard. Furthermore one should bear in mind the often poor conditions of survivors after recovery. The journey overseas has probably weakened them. Moreover, particularly refugees and asylum-see-kers frequently need extra medical or psychological treatment, due to recent traumatic experien-ces in detention centres or at the hands of human smugglers. 111

Lowe, SDLR 1977, p 622; Morrison 2016, p 168; Rothwell 2015, p 284; De La Fayette, IJMCL 1996;

104

see also Nicaragua case, par 213. Arts 2(1), 8 and 25(2) UNCLOS.

105

De La Fayette, IJMCL 1996, p 1; cited by Sohn 2014; Baatz 2014; Rothwell 2015; Attard 2014; Mar

106

-ten 2014.

Morrison 2016, p 168.

107

Tanaka 2015, p. 83; Coppens & Somers, IJMCL 2010, p 399; Vitzthum 2006, p. 89; Churchil & Lowe

108

1999, p. 63.

Long Lin case, paras 354, 357; ACT Shipping case, para 426; see also discussion of incidents of the

109

Erika (1999), the Castor (2000) and the Prestige (2002) in Nordquist 2007, pp 497, 505-19.

Churchill 2014, p 13; See also Morrison 2016, p 169; Morrison 2012, p 77.

110

Report HRW 2019.

(23)

However, even if a vessel is deemed to be in distress and therefore has a right to refuge, that does not ipso facto imply the allowance of disembarkation. To that end, it is essential to no-tice that the right to refuge does not entail the duty to allow disembarkation to rescued refugees, as is reiterated in literature. Worth mentioning is also the practical differentiation, illustrated 112

by a recent Italian block-off practice. Last August, after the Italian coastguard ship Ubaldo Di-ciotti picked up migrants from an overcrowded vessel near Lampedusa, Italian interior minister Salvini asserted Malta was responsible for disembarkation, because the boat full of migrants first passed through Malta’s search and rescue region. Malta refused as well, which caused an im-passe. Subsequent to long diplomatic negotiations, Italy granted access to its port in Sicily. Yet, Salvini did not let the migrants disembark and deemed the docking as a “technical stop.” The minister added that disembarkation could only be possible if pledges of fair burden sharing by other EU Member States were made. 113

2.3.4. Flag States

Via the shipmaster, the flag State bears the obligation to render assistance to persons in distress at sea. The latter duty has not been made explicit by the literal text of Reg. 33(1-1) SOLAS, but de-rives from UNCLOS and SOLAS. Furthermore, the flag State should monitor whether the 114

masters of vessels flying its flag discharge these duties. This is a due diligence obligation , 115 116

which also includes the administrative control over relevant public and private vessels. More117

-over - similar to coastal States - Reg. 33(1-1) SOLAS provides flag States for the general obliga-tion to promptly release the shipmaster. To that end, the same consideraobliga-tions as menobliga-tioned in the

Proells 2017, p 729.

112

The Guardian, 21 August 2018; ibid, 24 January 2019.

113

Art 98(1) UNCLOS; SOLAS Reg V/33(1).

114

Ibid.

115

UNHCR Background Paper 2014.

116

Pulp Mills case 2010, para. 197; Illustrative is the current controversy with regards to the administra

117

-tive/practical requirements private rescue vessels have to comply with in order to conduct rescue activities in the Mediterranean sea. See Sea-Watch press release, 2 April 2019; 7 May 2019.

(24)

discussion of coastal States apply to the flag State-duty to co-ordinate and co-operate. This is in line with the Guidelines, which demand flag and coastal States to set up effective arrangements to promptly support shipmasters. 118

Not surprisingly, Italy has consistently used the latter as an argument to vouch the legal basis for their push-off activities with regards to rescue operations of NGOs. The Mediter119

-ranean State repeatedly asserts the private rescue vessel’s flag State to bear the responsibility of dealing with the rescued refugees. Last January, Italy pointed at the Netherlands to allow dis120

-embarkation of 47 migrants onboard a NGO vessel flying the Dutch flag, being blocked from Italian ports. The Dutch Government however said there had to be found a long-term solution 121

for the mixed migration problem first. Furthermore the Ministry rejected responsibility, since 122

the vessel had acted "on its own initiative (…). It was up to the captain of Sea Watch 3 to find a nearby port to disembark the 47 migrants he had on board”. The foregoing demonstrates the 123

reluctance (and maybe inability) of flag States to exercise effective control over ships flying their flag. Arguably, the rationale for the reluctance of flag States is the unwillingness to create 124

precedents with regards to disembarkation of rescued refugees and/or their relocation. A casual green light by the flag State could be used against it at subsequent similar instances. Cases in-volving NGO rescue vessels are particularly sensitive to that end, given the NGO’s perpetuating presence in the Mediterranean sea.

118

EU Migr L Blog, 9 July 2018; France 24, 3 April 2019; The Local, 29 January 2019; The Guardian, 13

119

August 2018. Ibid.

120

The Local, 29 January 2019.

121

Reuters, 28 January 2019.

122

The Local, 29 January 2019.

123

UNGA Oceans Report 2005, p 3.

(25)

But it is doubtful whether all the assertions made by the Dutch Government are correct. Even though the flag State in principle lacks responsibility for conduct of private actors , the 125

Dutch claim, with regards to the shipmaster’s autonomy in the quest for a place of safety, is not viable. In the end, flag and coastal States are to establish effective arrangements to promptly support shipmasters and to ensure disembarkation to a place of safety. While shipmasters pos126

-sess discretion to some extent , that mere fact does not exempt flag States from the obligation 127

to actively co-ordinate and co-operate the disembarkation of the rescued refugees. Therefore, flag States cannot simply point at the shipmaster’s discretion to find a place of safety to disem-bark the survivors, in order to avoid their own responsibilities. Instead they should actively par-take in the co-ordination and co-operation process. Yet, no legal basis for an exclusive duty to permit disembarkation upon the flag State is provided by International Law of the Sea. Thus, 128

despite the flag State’s intrinsic special link with the vessel at hand, the conclusion reads the same as with regards to coastal States: the divided character of the obligation to allow disem-barkation causes States to point at one another in their attempt to avoid the obligation. Regard-less, the latter does not take away the ability of States to reach an effective bilateral agreement with the flag States with regards to non-SAR considerations, such as burden-sharing arrange-ments after disembarkation of the recovered persons. However, these considerations are out129

-side the scope of the SAR regime. 130

UNHCR Background Paper 2014, p 13.

125

126

Reg V/34-1, SOLAS Convention. Yet, remark that the master should also comply with relevant re

127

-quirements of the State responsible for the SRR, according to Guidelines, para 5.1.7.

Note Spain’s proposal to place more responsibility on flag States. COMSAR 7/8. Yet, the proposal did

128

not reach a majority in favour. COMSAR 7/23, para 8.8.

See e.g. The Local, 28 June 2019; The Telegraph, 30 January 2019.

129

COMSAR 7/23, para 8.21.

(26)

2.3.4. State Responsible for SRR

The State responsible for the search and rescue region (SRR) in which the recovery action took place is the main addressee by virtue of the second sentence of Reg. 33(1-1) SOLAS. According-ly, the following subjects concerning this State are discussed consecutively: (i) Responsibility to achieve disembarkation, and (ii) Temporal component of disembarkation. The prior provides for the norm, whereas the latter tempers it.

(i) Responsibility to Achieve Disembarkation

Arguably, the discussion on the responsibility of the State responsible for the SRR is the most contentious one, raising questions of interpretation in literature on the ordinary meaning to be given to the text of Reg. 33(1-1) SOLAS. Hence, the nature of the responsibility is examined 131

in the following paragraphs.

Pursuant to the second sentence of Reg. 33(1-1) SOLAS, the State responsible for the SRR region bears a legal duty to provide SAR services for a determined area. This obligation 132

derives from the SAR Convention and its endeavour to cover all maritime areas with search and rescue regions (SRRs) in order to ensure the safety of life at sea. After the establishment of a 133

SRR by multi-/bilateral agreement, the IMO shall be informed of such agreement. Subsequent134

-ly, the Organization communicates the establishment of the SRR to other States. In order to 135

provide for adequate co-ordination of SAR services within the determined area, States are re-quired to establish rescue co-ordination centres (RCCs). Noteworthy for this paper is the cre136

-ation of rescue co-ordin-ation centres (RCCs) in the central Mediterranean by Italy, Malta and

E.g. Proelss 2017, pp 726-727; See also Papanicolopulu 2016; Barnes, ICLQ 2004; Coppens &

131

Somers, IJMCL 2010.

Annex SAR Convention, Chapter 2.

132

Ibid, Chapter 1 and para 2.1.3.

133

Ibid, 2.1.4.

134

Ibid, 2.1.6; see IMO’s Global SAR Plan.

135

Ibid, 2.3.1.

(27)

Libya. Each of these States have one SRR. Thus, depending on the SRR in which the re137 138

-covery activity took place, there is to be one State responsible for co-ordinating the SAR activi-ties. 139

The text of Reg. 33(1-1) SOLAS entails a legally binding obligation upon the State re-sponsible for the SRR to exercise primary responsibility to assure SAR co-ordination and co-op-eration. However, questions have been raised in literature regarding the tangibility of the duty to exercise responsibility. It has been argued that the responsibility to ensure ordination and co-operation is merely an obligation of conduct, not of result. Yet, it is essential to stress that the 140

second sentence of Reg. 33(1-1) SOLAS should not be seen without duly noting the other sen-tences of this regulation, because these three parts are heavily intertwined and constitute the carefully drafted compromise text the Member States refer to in the drafting history. The fol141

-lowing illustrates why the responsibility upon the State responsible for the SRR should be deemed as a duty of result, rather than of conduct.

The citation “ensuring such co-ordination and co-operation occurs” of the second sen-tence of Reg. 33(1-1) SOLAS refers to the first sensen-tence in which coastal States are obliged to co-ordinate and co-operate. However, “such” does not merely alludes to the conduct, but just as well to the outcome of this co-ordination and co-operation. Similar to what has been said earlier, the co-ordination and co-operation should lead to the release of the master - i.e. a duty of result incumbent upon coastal States. Accordingly, in line with the wording of the text, one cannot sep-arate the conduct (co-ordination and co-operation) and the result (master’s relief).Given the fact that liberation of the shipmaster can only realistically take place through disembarkation of the

Ibid, 2.1.6; see IMO’s Global SAR Plan.

137

Ibid.

138

Remark that Italy and Malta are ‘merely’ coastal States when recovery action takes place within

139

Libyan SRR. Accordingly, the duties as described under heading 2.3.2. are then applicable to Italy and Malta.

140

MSC 78/26, para 16.49.

(28)

survivors , the coastal States possess a shared obligation to allow disembarkation. Even 142 143

though the second sentence of Reg. 33(1-1) SOLAS contains in principle an obligation of con-duct , the inevitable connection with the previous first sentence - through the word “such” - 144

establishes an obligation of result upon the State responsible for the SRR. I.e. this State bears responsibility for achieving disembarkation of the survivors to a place of safety. Responsibility of that State to that end is twofold; allowing for disembarkation itself or ensuring other States allow for disembarkation. This interpretation is further espoused by the Member States. More145

-over, a similar conclusion follows from the words “so that”, which link the responsibility of the State responsible for the SRR on co-ordination and co-operation with the achievement of the survivors being disembarked to a place of safety.

What has furthermore been contended in literature concerning the obligation of the State responsible for the SRR, is the difference between primary and a more exclusive form of respon-sibility. Because the State responsible for the SRR has the primary (and not the exclusive, ab146

-solute or final) responsibility according to the literal wording of the text, the State responsible for the SRR will simply point at another coastal State when being faced with a request to allow for disembarkation. Although one is likely to draw the foregoing conclusion when scrutinising the 147

wording of the text, the drafting history and the Guidelines seem to suggest a relatively more ex-clusive form of responsibility upon the State responsible for the SRR.

Hypothetically, the master would be released from his duty when the survivors die. But that ‘result’ is

142

not very likely to happen at the hands of the coastal States, which makes that scenario unfeasible. See under heading 2.3.2.

143

144

IMO Res MSC.167(78), under ‘REALIZING FURTHER’.

145

146

During the drafting process, Norway stressed this point emphatically. See MSC 77/10/2.

(29)

While considering the Guidelines, Malta notably opposed the draft version as prepared in the sub-committee. According to the Mediterranean State, the Guidelines would put the final 148

responsibility on the State responsible for the SRR to accept the persons rescued at sea in its ter-ritory, if no other State would be willing to provide a place of safety. That would encourage the trafficking of illegal migrants, since the vessels carrying them would simply have to enter the closest neighbouring SRR and call for assistance. Subsequently, the State responsible for the SRR would then have the duty to render assistance and provide them with a place of safety. 149

Malta did not welcome these consequences. Instead, it proposed amendments to the initial draft, aiming at a clearly stated shared responsibility of States to provide for disembarkation to a place of safety. 150

Yet, Malta’s proposal was merely backed by a minority of the present delegations. The majority acknowledged the drafted text of the Guidelines and regarded it as a carefully drafted compromise. Hence, they gave preference to considerations of striking a fair balance between 151

SAR concerns and sovereignty concerns, respectively:

“The fundamental difference between the COMSAR 8 text and the proposed amendment by Mal-ta was that the latter would delete the assurance given to the master that a place of safety will be provided by Contracting Governments for the persons in distress at sea. In the view of those sup-porting the text prepared by COMSAR, this deletion would endanger the well-balanced compro-mise text.” 152

Even though Malta’s interpretation of the responsibility involved - i.e. responsibility is (nearly) exclusive - does not per se reflect the intended or ordinary explication by all the Member States

Guidelines.

148

This example refers to situation in which the initial ‘refugee vessel’ makes it all the way to the SRR,

149

yet one could imagine a similar dislike on Malta’s side when refugees are assisted by a NGO vessel and the NGO vessel request the State responsible for the SRR to allow disembarkation.

MSC 78/26, paras 16.46-16.48. 150 Guidelines. 151 MSC 78/26, para 16.49. 152

(30)

to be given to the text, it exemplifies that the meaning of the text is not unambiguous. Further-more, the lack of proponents for its proposal, made Malta reserve its position to the Guidelines as well as to the SAR and SOLAS amendments. 153

Another reason for searching beyond the limits of the text, as well as for the reservation of Malta to the amendments, is the debated text of the resolutions adopting the amendments to the SOLAS and SAR conventions. After the Committee conducted a vote, the majority of the 154

Member States were in favour of the preambular paragraph which literally declares: “It is further intended that the responsibility to provide a place of safety, or to ensure that a place of safety is provided falls on the [State] responsible for the SAR region in which the survivors were recov-ered”. The adopted resolutions express the interpretation of the Member States and prescribe a 155

clear obligation upon the State responsible for the SRR, viz. clearer and more exclusive than stated in the wording of Reg. 33(1-1) SOLAS. 156

Based on the latter interpretation, the State responsible for the SRR has the duty to achieve disembarkation. Thus, despite the difficulty of pointing out one particular coastal State due to the lack of an exclusive obligation , one might find solace in the concentrated duty of 157

the State responsible for the SRR to achieve disembarkation. Yet, in the Central Mediterranean there are two important complicating factors. First of all, Libya cannot provide a place of safe-ty , which implies that its dusafe-ty to achieve disembarkation reaches as far as ensuring another 158

State permits disembarkation. Secondly, Malta has reserved its position on Reg. 33(1-1) SOLAS

MSC 78/26, paras 16.52-16.54.

153

IMO Res MSC.155(78) on SAR Convention; IMO Res MSC.153(78) on SOLAS Convention.

154

Ibid, preambular paragraph starting with ‘REALIZING FURTHER’. See for discussion and voting:

155

MSC 78/26 paras 3.63-3.67, 3.72-3.75.

Note in this respect Malta’s comment (MSC 78/26, para 3.67), which asserted the Committee did not

156

have the right to include the interpretative preambular paragraph as there was no consensus reached. Proelss 2017, p 729.

157

See fn 92.

(31)

and the corresponding Guidelines. To that end, Malta is not obliged to exercise primary re159

-sponsibility to ensure disembarkation of the recovered persons within its SRR, as it merely has to abide to the SAR regime subsequent to the 2004 amendments - which does not entail such re-sponsibility. 160

Besides, even if the 2004 amendments have been accepted and there are no issues of pro-viding a place of safety, one could still debate on the practical value of the concentrated duty to achieve disembarkation upon the State responsible for the SRR. For instance, the question is whether States are able to render diplomatic pressure when the State responsible for the SRR fails to comply with its obligation to permit disembarkation, in case other coastal States refuse as well. During the drafting process, delegations thought it was “reasonably easy” to do so. Yet, 161

in order to properly answer the last question it is essential to understand what constitutes to a breach of the obligation to achieve disembarkation, pursuant to the interpretation of Reg. 33(1-1) SOLAS in the light of their object and purpose. To that end, the question is what specific con162

-duct is to be expected from the State responsible for the SRR, particularly in the (likely) event other States reject disembarkation of the recovered seaborne refugees. How long can the State 163

responsible for the SRR reasonably seek for multilateral co-operation before it is obliged to per-mit disembarkation itself?

Following the wording of the article, the time of judgement depends on the circumstances of the case and what relevant IMO guidelines prescribe. I.e. what is reasonable to expect from the State responsible for the SRR at one moment, differentiates situationally. The principle of reasonableness in the context of disembarkation has been made explicit by the last sentence of

See fn 63. 159 COMSAR 6/22, para 8.74. 160 COMSAR 7/23, para 8.18 161 Crawford 2002, p 125-130. 162

Damrosch & Murphy 2014, p 501.

(32)

Reg. 33(1-1) SOLAS (“as soon as reasonably practicable”) and is therefore touched upon under-neath within the discussion on the temporal component of disembarkation.

(ii) Temporal Component of Disembarkation

In the foregoing paragraphs, the most prominent duties upon the State responsible for the SRR have been discussed. However, alongside the determination of a responsibility to achieve this result incumbent upon the State responsible for the SRR, practice exhibits another constituent worth mentioning: the temporal component of disembarkation. Particularly cases which involve a high number of rescued refugees, turn out to be difficult to resolve within a reasonable time frame. Also NGO rescue vessels have been often subject to a long delay before disembarka164

-tion took place. Therefore the ques-tion is raised what is a reasonable time before the State re165

-sponsible for the SRR is obliged to permit disembarkation itself. Or, to approach to issue from an anthropocentric angle: how long could refugees reasonably be held on a rescue vessel until their disembarkation shall be operationalised?

First of all, the last sentence of Reg. 33(1-1) SOLAS refers to the principle of reason-ableness. This sentence is both demanding and restrictive on what is expected from the coastal States. On the one hand it holds the mandatory nature (“shall”), which stipulates one more time the obligation to allow disembarkation upon coastal States. On the other hand the text implies that the temporal component of disembarkation is relative to what could reasonably be expected from the coastal States, given the capacities and facilities available in underlying case.

The reasonableness of the practicability thus depends on the circumstances of the case, as referred to in the second sentence of the amended provision’s text. This casuistic approach is likely to provide space for striking a reasonable balance between the rescue duties upon the shipmaster and the coastal States’ duty to allow disembarkation to a place of safety,

MSC 77/10/2, para 5.

164

Alarmphone analysis, 4 July 2019.

(33)

respectively. However, the premise of this case-to-case analysis is as indicative as it is hollow. 166

Even though the construction avoids rigidity and leaves room for reasonableness, it can also be subject to an application where state sovereignty-sentiments are likely to become dominant.

For instance, one could argue there is in principle little to expect from the State responsi-ble for the SRR, because the practicability should also be seen in the light of the States’ ability to reach an agreement with regards to the relocation of the rescued refugees, as that could either reduce or increase the reception capacity of refugees. Given the gridlock within the European Union concerning relocation of refugees (see introduction), one cannot reasonably presume a swift agreement ensuring relocation of the survivors to other EU Member States after they have disembarked.

Besides the reasonableness and the casuistic nature, the citation “with minimum further deviation from the ships’ intended voyage” from the second sentence of Reg. 33(1-1) SOLAS provides for a condition of disembarkation. However, it does not dwell upon the time-related is-sue, as it only makes mention of deviation in terms of navigation. Since this quote lacks a mandatory time-frame, it allows coastal States to block assisting vessels from their ports and leave the often heavily crowded ships floating around the sea in anticipation for their release.

Apart from the literal wording of the text, the Guidelines also take a prudent approach on the temporal component of disembarkation. They state a humanitarian safeguard by expressing that coastal States and the State responsible for the SRR should make every effort to minimise the time of survivors aboard an assisting vessel. Yet, this is not a concrete duty upon the ad167

-dressees, as it leaves much room for a discretionary judgement and it does not encompass an obligation of result. Furthermore, even though the State responsible for the SRR should hasten

166

Duty to Allow Disembarkation of Rescued Refugees: Floating Between Sovereignty and Humanitarianism