O ﬀ shore Processing Arrangements: E ﬀ ect on Treaty Ratiﬁcations of Receiving States

: This article examines o ﬀ shore processing arrangements of four di ﬀ erent time-periods and geo-political regions—the Safe Havens of the United States with Jamaica and the Turks and Caicos Islands; the 2001 and 2012 Paciﬁc Solutions of Australia with Nauru and Papua New Guinea; and the EU–Turkey deal. In examining these arrangements, the article attempts to ascertain whether each of these arrangements had an impact on the ratiﬁcation of refugee and human rights-related treaties by the states receiving the asylum seekers and refugees for processing and / or settlement. It does so by ﬁrst assessing the contents of the o ﬀ shore processing agreements for refugee and human rights clauses and obligations. The article then looks at the general patterns of treaty ratiﬁcation of each receiving state, prior to its entering into o ﬀ shore processing arrangements. After the general patterns of treaty ratiﬁcations of each state are established, the article goes on to investigate whether o ﬀ shore processing arrangements had any e ﬀ ect on these patterns. This is based on the analysis of the contents of the agreements, together with an examination of the timing of the refugee and the human rights treaty ratiﬁcations of the receiving state, at the time of the arrangements. The article ﬁnds that the e ﬀ ect, although minimal, is quite nuanced.


Introduction
Increasingly, wealthy developed states are engaging in international deterrence policies to stem the inflow of asylum seekers. According to Tan, deterrence policies "encompass both migration control, aimed at preventing asylum seekers ever arriving at the destination state, and asylum processing or refugee protection, focused on shifting the location of asylum procedures or refugee protection to third states" (Tan 2018, p. 35). While states can and do engage in deterrence policies unilaterally, more commonly they engage in 'cooperative deterrence' by signing bilateral and multilateral agreements with other states and institutions, thereby externalising border control and asylum procedures. These externalisation policies include interceptions (also called interdictions) at sea, third country (also called offshore) processing and people exchange arrangements, as well as third country protection (through settlement arrangements) (FitzGerald 2019; Tan 2018, pp. 35-46;Hathaway and Gammeltoft-Hansen 2015;Gammeltoft-Hansen and Tan 2017, pp. 32-40;Den Heijer 2012, pp. 167-87, 209-83).
Often, transfer (also called disembarkation) arrangements encompass the combination of these cooperative externalisation policies. Globally, four versions of such arrangements, in which offshore including a returns arrangement, 5 a people swap arrangement similar to the one attempted by Australia with Malaysia in 2010 (Tan 2018, pp. 43-44) 6 and, for some asylum seekers, an offshore processing arrangement (Tunaboylu and Alpes 2017, pp. 85-86). 7 It is in the form of the last type of (offshore processing) arrangement that the EU-Turkey deal is retained in this article. As a result of the difficulty in comparing such disparate case studies and the unlikelihood that such a comparison could be a predictor for any future arrangements, the article approaches the case studies from the perspective of an independent descriptive (rather than comparative) analysis.
Thus, in analysing the case studies, the article first examines the contents of the offshore processing agreements (where they are publicly available). This analysis investigates what international, refugee and human rights requirements were incorporated in the agreements for the receiving countries to adhere to (Section 2). The analysis also touches upon the power dynamics of the parties to the agreements, the financial incentives for the receiving states, as well as the strong focus of the agreements on the element of border control. The analysis helps to ascertain how (if at all) the agreements contributed to the treaty ratification events 8 of the receiving states, either through the requirements in the agreements or through any other of the abovementioned elements.
After exploring the contents of offshore processing agreements, Section 3 looks at whether these agreements, and the offshore processing arrangements they were part of, affected the treaty ratifications of the receiving states. It does so by first identifying the patterns of treaty ratification of key refugee and/or human rights treaties that the receiving states adhered to prior to the offshore processing arrangements. This identification relies on the categories and patterns supplied by the field of Political Science. 9 The United Nations Treaty Collection (UNTC) database is utilised to check whether any of the key refugee and/or human rights treaties were ratified at the time of the offshore processing arrangements, taking into account the contents of the agreements in each of the case studies. While this in no way proves causation, certain conclusions can still be drawn, especially when evaluating secondary supporting documents, such as the Reports of the Universal Periodic Review (UPR).
Section 3 concludes by complementing the main analysis with an evaluation of treaty ratification patterns of treaties on statelessness and border control. While these are peripheral to the main focus of the article-being the development of the asylum system of the receiving states through refugee and human rights treaty ratification-this analysis brings depth to the discussion and conclusions. Finally, the article concludes the independent descriptive analysis of the case studies by providing a nuanced answer to the research question (Section 4).
Before embarking on the analysis, it is necessary to clarify what is meant by 'offshore processing arrangements', the 'asylum system' of the receiving country, and 'treaty ratification'. To assist with 5 It was argued that the EU-Turkey statement was part of an extension to the Readmissions Agreement negotiations (Idriz 2017, p. 5). According to Kaya (1995, pp. 129-30),The EU-Turkey [readmissions agreement] RA is intended to expel irregular migrants who do not fulfil the entry and residence conditions of the country concerned. . . . However, there is no mention of asylum seekers and refugees in the RA. The EU-Turkey Statement raises the question of whether asylum seekers and refugees may be readmitted to Turkey based on the implicit assumption that Turkey is a safe third country or first country of asylum in accordance with the Asylum Procedures Directive. (emphasis added)According to the Asylum Procedures Directive (2013), a 'safe third country' is one where a person is not at risk of serious harm or refoulement, and which provides access to an assessment of refugee status (Article 38). A 'first country of asylum' is a country that already recognised the person as a refugee or otherwise provides sufficient protection, including from refoulement (Article 35). 6 In response to political pressure to address the increase in irregular boat arrivals between 2008 and 2010, the Australian government entered into a 'people swap' arrangement with Malaysia in 2011. As per the Memorandum of Understanding (MOU), Australia would send 'back' to Malaysia 800 asylum seekers that had arrived irregularly by boat and in return would resettle 4000 people that were determined to be refugees by the UNHCR in Malaysia. This deal was challenged in the Australian High Court and ruled that same year as incompatible with the Australian Migration Act 1958 as it stood at that time. The Malaysia Solution is not discussed further in this article, as, beyond honouring the resettlement of refugees promised by Australia, the arrangement was abandoned by the Australian government (see 2011 Australia-Malaysia Agreement;Plaintiff M70/2011 v Minister for Immigration and Citizenship;andPlaintiff M106 of 2011 v Minister for Immigration andCitizenship [2011] HCA 32; Gillard 2011). were resettled out of Nauru in early 2008 when the Australian government announced the end of the Pacific Solution arrangements (Phillips 2012, pp. 12-13).
The closure of the processing centres coincided with another increase of irregular movements, with irregular boat arrivals to Australia increasing from 179 in 2008 to 6850 in 2010 (Houston et al. 2012, pp. 23, 70). This led to a political debate that resulted in the 2012 Pacific Solution. While publicly dismissing the option of reopening the Pacific Solution in 2010, the Australian government, with little success, began canvassing neighbouring countries, including East Timor, to process refugees (Gillard 2010). By 2012, an expert panel was tasked to make recommendations to address the influx of irregular boat arrivals. On 13 August 2012, it issued its report, recommending the reopening of the Pacific Solution (recommendations 8 and 9, see Houston et al. 2012, pp. 9, 16). Consequently, on 29 August 2012 and 8 September 2012, Australia signed MOUs with Nauru and PNG, respectively, to reopen the processing centres (2012 . 15 While in many respects a repetition of the 2001 Pacific Solution, the 2012 iteration introduced the notions of refugee status processing being undertaken by the receiving countries of Nauru and PNG. A renewal of the agreements in 2013 introduced a further amendment, in which recognised refugees were to be settled in Nauru and PNG (or another third country) but never in Australia.
The EU-Turkey deal of 18 March 2016 (EU-Turkey Statement) 16 was similarly reached in response to a large increase in irregular movement by sea of migrants, asylum seekers and refugees from Turkey to European states. The number of irregular arrivals peaked in 2015. 17 One of the key elements of the deal was the 'people swap' arrangement, which provided that for every return of a Syrian person from the Greek Islands to Turkey, another Syrian is to be resettled from Turkey to the EU (EU-Turkey Statement, point 2). Based on this arrangement, an accelerated form of RSD processing would take place in Greek 'hotspots' by Greek authorities (assisted by the European Asylum Support Office). Those who had not lodged or had withdrawn their asylum application, opted for assisted return, or had their application rejected or dismissed as inadmissible on the ground that Turkey is a 'safe first country of asylum' were returned to Turkey (Tunaboylu and Alpes 2017, p. 84). This resulted in some people who were still in need of having their asylum claims assessed in Turkey, being returned (European Commission 2016c, pp. 5-6;Van Liempt et al. 2017, pp. 15-16;Alpes et al. 2017, pp. 4-6).
As can be seen from the case studies, offshore processing arrangements are generally proposed and implemented by developed states in times of an influx of irregularly arriving asylum seekers and migrants (see generally FitzGerald 2019). These arrangements essentially entail the transfer of asylum seekers and/or irregular migrants who reached the territory of their country of destination or were intercepted at sea on the way to their country of destination. The transfer is made to a specified locale in a country that was not the destination country of the asylum seekers undertaking such a journey, that is, the arrangements are between a destination (also called the transferring state) and a third state (also called a receiving state) that may or may not be a transit state.
In the third state, the asylum seekers are held in processing centres 18 for assessment of their refugee status. The assessment of the refugee status is performed either by the transferring state, the receiving state, or by an international organisation such as the United Nations High Commissioner for Refugees (UNHCR). Although, the latter usually holds a monitoring role. Those who are found to not be refugees are deported to their countries of origin, while those assessed to be in need of asylum are to be provided with a durable solution through voluntary return, local integration or resettlement. 15 The 2012 Australia-PNG MOU superseded one signed a year earlier on 19 August 2011 (see discussion in Section 2.2 below). 16 While it is widely known as the 'EU-Turkey deal', it is in fact a statement of an arrangement finalised at the meeting of 18 March 2016 between individual "Heads of State or Government of the European Union" and their "Turkish counterpart" (CJEU, Case T-192/16, NF v European Council, Order of 28 February 2017, ECLI:EU:T:2017:128, paragraph 68). 17 Over a million migrants, refugees and asylum seekers arrived in the EU territories by irregular means and this inflow resulted in nearly 4000 deaths (Clayton et al. 2016). 18 Often processing centres are indistinguishable from detention centres (see generally, Provera 2013).
As such arrangements focus primarily on deterrence rather than providing refugees and asylum seekers with durable solutions, they are heavily criticized for putting people at risk of refoulement, arbitrary detention, and torture (see for instance , Koh 1994;Taylor 2005;Legomsky 2006;Gleeson 2016;HRC 2017;Kaya 2017). Depending on whether the transferring state or the receiving state is undertaking the assessment of refugee claims, that country's asylum system needs to be robust enough to ensure that, among other things, fair and efficient processing takes place, as well as timely durable solutions are provided. The term asylum system of a state, while commonly used (for example, Common European Asylum System), is often difficult to define. The term is used inconsistently by UNHCR, especially in its discourse regarding offshore processing. UNHCR used asylum system synonymously with 'asylum space' (UNHCR 2010a, para. 48), 'protection space' (UNHCR 2013e, para. 3(iv); see also UNHCR 2010a) and 'protection capacity' (UNHCR 2010a, para. 45). Legomsky, in his 2003 paper on return of refugees to third countries, attempts to define the term 'effective protection', while arguing that it is the basis of a domestic asylum system (Legomsky 2003). According to UNHCR and Legomsky, one of the key elements of what makes for effective protection and thus an effective asylum system is the ratification of the 1951 Convention Relating to the Status of Refugees (Refugee Convention) (UNGA 1951) and its 1967 Protocol Relating to the Status of Refugees (Refugee Protocol) (UNGA 1967b), as well as the ratification of and respect for other rights, 19 particularly, the nine core human rights treaties (Legomsky 2003, pp. 633, 639-45, 658;UNHCR 2017, pp. 23-26). 20 These are:

•
International Convention on the Elimination of All Forms of Racial Discrimination (CERD) (UNGA 1965 It is important to highlight that ratification of the Refugee Convention, its Protocol and the core human rights treaties is only one part of a country's asylum system. Although not discussed in this article, an asylum system also entails, amongst other elements, the implementation of the refugee and other human rights treaties into the state's domestic legislation (Legomsky 2003, pp. 633, 639-45, 19  658); the provision of a fair and efficient refugee status determination (RSD) process (Legomsky 2003, p. 654; 21 the provision of durable solutions for recognised refugees (Legomsky 2003, p. 662); as well as the effectiveness of that asylum system, by ensuring that the state has the capacity (that is, resources, staff and experience) to execute the above requirements (UNHCR 2017, pp. 61, 127, 154, 164). Thus, although only one part of a country's asylum system, treaty ratification 22 serves as a basis for ensuring that asylum seekers and refugees receive effective protection. Treaty ratification research focuses on the interrelated questions of why states ratify human rights treaties (Hathaway 2007) and whether ratification makes a difference to the human rights systems of that country (Hathaway 2002). While the latter question relates to domestic implementation of treaties, the former question is more pertinent to the analysis of this article. It is important, however, to understand what the general patterns of treaty ratification are, before considering whether countries taking part in offshore processing arrangements deviated from those patterns at the time of the arrangements. While theories of treaty ratification such as rationalism, constructivism and liberalism form the basis of the literature in the field of Political Science, 23 it is their observations regarding patterns of treaty ratification that are relied on during the analysis. 24 These patterns relate to states acting in self-interest in pursuit of both tangible (financial) and intangible (reputational) rewards, 25 states following the ratification patterns of their regional communities, 26 as well as the tendency of newly-independent (democratic) states to ratify human rights treaties but not their enforcement protocols. 27 Before beginning the analysis, a few other treaties should be introduced, as they are pertinent to both an effective asylum system and the nature of offshore processing arrangements. These include the treaties on statelessness and smuggling. UNHCR strongly advocates the ratification of the two conventions on statelessness as being an important element of an effective asylum system (UNHCR 2019b, p. 1;UNHCR 2015b, pp. 9-11;UNHCR 2015c, pp. 11-12;UNHCR 2014a, pp. 5-7). The two conventions are the 1954 Convention relating to the Status of Stateless Persons (1954 Statelessness Convention) (UNGA 1954) and 1961 Convention on the Reduction of Statelessness (1961 21 It must be noted that the RSD process itself involves multifaceted requirements, including reception and registration of asylum seekers and refugees, legal representation, provision of interpreters, interviews, notification of decisions, and access to review and appeal of negative decisions (UNHCR 2003), not to mention the consideration of the refugee definition itself (UNHCR 2019a). 22 The countries discussed in the case studies became parties to the various conventions through ratification, accession and succession (see, for instance, UNTC, Chapter IV: Human Rights n.d.; UNTC, Chapter V: Refugees and Stateless Persons n.d.). Throughout this article, however, 'ratification' is used as a generic term to refer to the action taken to become a party to any of the treaties. Additionally, where the action under discussion is signing up to a treaty, or making or removing of a reservation or declaration, for the purposes of this article, this action is referred to generally as a 'ratification event'. 23 While (neo)realism is touched upon, especially with regards to the significant power imbalances between the regional hegemons of the US and Australia and their arrangement partners of Jamaica, TCI, Nauru and PNG, respectively, these power imbalances relate more to the negotiations and signing of the MOUs. According to the realist theory "[w]eaker governments 'accept international obligations because they are compelled to do so by great powers.'" (Helfer 2002(Helfer , p. 1842). However, it has been shown that, while regional hegemons do at times attempt to influence less powerful states, as in the case of the Caribbean in the late 1990s, they do so unsuccessfully (Helfer 2002(Helfer , p. 1892). On the whole, "powerful countries are rarely consistent in their application of human rights standards to their foreign policy, and . . . rarely employ sanctions-political, economic, military, or otherwise-to coerce other countries into improving their human rights record" (Neumayer 2005, p. 926;Hathaway 2002Hathaway , pp. 1944. 24 Although utilising the theories found in the field of Political Science, this article sits firmly in the field of International Law (for examples of research in International Law that draws on explanations of treaty ratification from the field of Political Science, see Schloenhardt and Macdonald 2017;Koh 1997; see also Lentner 2019). 25 These are then balanced, as part of rational self-interest, against financial (through sanctions) and reputational costs of non-ratification and non-compliance (see generally, Hathaway 2002Hathaway , pp. 1942Cole 2005, pp. 475-76, 490-91;Spence 2014). 26 This is so, especially if those communities share cultural norms. The acceptance or otherwise of human rights treaties is not limited to norms shared by geographic neighbours and can generally be expanded to a sense of belonging to the international community. In particular, where a state participates in international forums (for example, the Universal Periodic Review), it is more likely to ratify human rights treaties (see generally, Helfer 2002Helfer , pp. 1845Helfer -48, 1898; Hathaway 2002, pp. 1960-62 discussing Koh 1997 (transnational legal process model); Cole 2005, pp. 476-78, 491 (world polity institutionalism); Hathaway 2007, pp. 611, 613;Spence 2014 (constructivism)). 27 This pattern, together with placing a focus on the type and nature of a state's domestic actors and processes, belongs to the liberalist theory of treaty ratification (see Hathaway 2002Hathaway , p. 1954Hathaway 2007, pp. 611, 613).
Statelessness Convention) (UNGA 1961). These treaties allow those determined as stateless, certain minimal standards, such as socio-economic rights (including education, employment and housing), identification and travel documents (for example, 1954 Statelessness Convention, Articles 17-22 and 26-28), as well as the access to a grant of nationality (and by that step, a durable solution) (for example, 1961 Statelessness Convention, Article 1). While the focus of this article is on the treaty ratification of international refugee and human rights instruments, the heavy focus of the offshore processing arrangements on border control and anti-smuggling operations means that the effect of offshore processing arrangements on ratification patterns of the receiving states with regards to border control treaties should not be ignored. The article thus briefly looks at the ratification by receiving states of the United Nations Convention against Transnational Organized Crime (Organized Crime Convention) (UNGA 2000c) and its three supplementing Protocols (UNGA 2000d;UNGA 2000e;UNGA 2001), and in particular, the Protocol against the Smuggling of Migrants by Land, Sea and Air (Smuggling Protocol) (UNGA 2000e).

Content of Offshore Processing Agreements
Scholarship on externalisation policies (of which offshore processing arrangements is one of the measures) advocated for the inclusion of human rights and protection clauses in the agreements 28 relating to such measures. For instance, Giuffre (2013) in her investigation of readmission agreements between EU states and safe third countries or first countries of asylum, argues that non-affection 29 and human rights clauses should be included in the agreements, in order to avoid potential treaty conflict between human rights treaties and the readmission agreements (Giuffre 2013, pp. 103-6).
However, a 2018 study by Donno and Neureiter (2018) on whether human rights clauses in trade agreements reduce repression, showed that even when compliance with human rights clauses is legally bound to greater financial benefit, a country's human rights record is unlikely to improve. They found that punishment for human rights infringement through aid sanctions (on foreign aid but not on trade) does promote compliance amongst developing nations more than rewards for that compliance (Donno and Neureiter 2018, p. 344; see also Spence 2014, p. 428). However, donor states use these measures in rare situations, such as serious breaches of basic political rights (that is, coups and flawed elections), rather than other types of human rights breaches (Donno and Neureiter 2018, pp. 346-47). Even so, this only works for countries where there is a strong dependence on foreign aid (rather than on trade), and where there is a strong power imbalance between the donor and receiving country (Donno and Neureiter 2018, p. 348).
The Donno and Neureiter results are relevant to the case studies of this article, because all agreements contain within them some form of financial consideration for the receiving states, combined with financial benefits ancillary to the arrangements. Although, as will be seen (with the exception of Turkey) they are more linked to general aid, as well as support for combating smuggling and trafficking operations and border control, rather than supporting host communities or providing human rights assurances for asylum seekers and refugees. Furthermore, in all but the EU-Turkey deal, there is a significant power imbalance and aid relationship between the partnering states.
Another way that human rights clauses in offshore processing agreements can be argued to be ineffective is the very limited enforceability of these agreements. The form of agreement most often used in the case studies is a Memorandum of Understanding between the partnering states. This form of agreement is acknowledged to not be legally binding (Roberts 2017, para. 33.20-33.24). The EU-Turkey deal is even more ambiguous. The discussions on the EU-Turkey agreement centre 28 The term 'agreement' refers to the written document, irrespective of the form it takes, between the parties entering into an offshore processing arrangement. Thus, an 'arrangement' is the general term that encompasses not only the agreements(s) between the parties, but also their negotiation and implementation. 29 A non-affection clause requires the contracting parties to an agreement to comply with rights and duties under other refugee and human rights conventions (Giuffre 2013, p. 93). on whether it is a 'Statement' issued as part of a press release makes it a legal document, before even considering its legal enforceability (for a summary of the ongoing arguments, see Kaya 2017, pp. 139-42). Nonetheless, human rights clauses, irrespective of their potential enforceability, reveal more about the intentions and common values of the partnering states (Giuffre 2013, p. 96). It is therefore important to look at their existence or otherwise in the agreements, as part of the full picture of the effect such agreements and their implementation, have on the asylum systems of the receiving countries, and in particular, on the treaty ratifications of those countries.

US Safe Havens
The MOU between the US and Jamaica was signed on 2 June 1994 (US-Jamaica MOU). As outlined in the MOU, the US committed to covering the cost of the arrangements, reimbursing Jamaica for any costs it incurred "subject to the availability of appropriated funds" (US-Jamaica MOU, Articles 4(E), 4(F)), as well as paying any claims arising out of the arrangements (US-Jamaica MOU, Article 9(B)). Whether Jamaica received any financial aid or incentives for its assistance to the US government is unknown. Local media reported, following a press briefing with the then Minister of Foreign Affairs and Foreign Trade, that there would be "no financial benefit" (The Gleaner 1994a). 30 Some indirect financial benefits were anticipated from the US military officials being based at the sole hotel in the small fishing community of Port Royal for the duration of the processing exercise. However, these were in fact limited; the expected boost to the economy was not forthcoming, as the US military had travelled with their own supplies. 31 In addition, a desire for closer ties to the US government might also explain Jamaica's willingness to enter into the arrangement. During this period, Jamaica was seeking accession to the North American Free Trade Association/NAFTA. 32 While explaining the terms of the MOU during a press briefing, the then Jamaican Minister of Foreign Affairs and Foreign Trade noted that "it is in our national interest to have close co-operation with the United States . . . a closer relationship will redound to our benefit". This indirect benefit no doubt influenced the decision in the absence of direct financial gains to the government (The Gleaner 1994a). A final consideration of Jamaica's involvement is its concern for the humanitarian situation in Haiti (see comments made by representatives of the government of Jamaica, The Gleaner 1994a). This is, in fact, reflected in the Preamble of the US-Jamaica MOU, which emphasized the human rights context in Haiti, and the deteriorating political situation, causing people to flee their country. 33 It is interesting here to contrast the arrangement with Jamaica to the (comparatively) 30 However, a 1993 State Department paper evaluating the various Caribbean nations as potential sites for offshore processing of Haitian refugees contains estimates for financial incentives that Jamaica may expect. For Jamaica, these include: "−$30 million in [Economic Support Funds] for each fiscal year in which the facility operates. −$10 million for [Foreign Military Financing] during same time period" (NSC 2020, p. 473). 31 This comment from The Sun, a Florida-based paper, challenges an earlier story printed in the local press, where US officials were reported as saying that several services, including construction, catering and garbage disposal, would be undertaken by Jamaican companies (see Thompson 1994;cf. San Martin 1994). 32 NAFTA is a free trade agreement between the United States, Canada and Mexico. Jamaica was not successful in its bid to join NAFTA. It did, however, receive enhanced benefits under the extant Caribbean Basin Initiative, a bilateral trading regime instituted by the United States with select Caribbean states (Wint 1994 protracted negotiations with the TCI. 34 This is in part due to its status as a United Kingdom (UK) dependency, and thus the need for UK, as well as local legislative approval (USA Today 1994, available in Magaziner 2006. The involvement of the UK created an interesting dynamic. The TCI was negotiating with the US as a regional hegemon. As a dependency of the UK, the TCI likely benefitted from the UK's involvement in the negotiations of the arrangements. 35 The proposed location of the facility on land, in a small archipelagic state constrained by resources to implement border control policies, would have also likely influenced the concessions granted to the TCI government. Consequently, in the first set of meetings (held between officials of the US, TCI and UK in Washington), the TCI might have made specific demands of the US (Goss 1994, p. 59;Smith 1994, p. 96;McKinley 1994, p. 106), 36 including $12 million in direct and indirect aid over a six months period (USA Today 1994, available in Magaziner 2006. Thus, while the US-TCI MOU contained similar clauses to those in the US-Jamaica agreement regarding covering costs, conditional reimbursement and commitment to the paying of compensation claims (Articles II(1)(vii), II(1)(viii) and V(2)), it also contained the intention "that it should contribute to the development of the infrastructure of the islands" and utilise local labour and contracts in support of the processing facility (US-TCI MOU, Preamble, Article II (1)(viii)). An upgrade of the local infrastructure, such as a former US military base and facilities for Haitians did take place, despite the migrant processing centre never becoming operational (Reynolds 2003, p. 39). On the other hand, the US-TCI MOU was signed despite local resistance to the use of the island as a site for offshore processing, with concerns of negative implications for the tourism sector (Greenhouse 1994a;Greenhouse 1994b), as well as management of the number of Haitians in the island. 37 The US Safe Havens agreements between the US and Jamaica and the US and the TCI were part of the US's broader border control and interdiction operations (FitzGerald 2019, pp. 71-95;Ghezelbash 2018, pp. 74-81). The preambles of the MOUs distinguish the national (and regional) priorities in the Jamaican and TCI responses to the Haitian crisis. Thus, potentially based on the abovementioned humanitarian grounds, the US-Jamaica MOU does not reference border control in its clauses. 38 The US-TCI MOU, however, does suggest that the processing facility on the TCI territory "should contribute . . . to the solution of the migration problem facing the [TCI]" (US-TCI MOU, Preamble). To that end, the TCI was to permit the US the right to enter and transit its territorial and internal waters "for the purposes of interdicting and rescuing Haitian nationals" (US-TCI MOU, Articles II(1)(vii) and II(2)(viii)). 39 Furthermore, the US had the right to "patrol" the waters between Haiti and the TCI for six months following the end of the arrangement (US-TCI MOU, Article X(iii)).
Any protections offered to the transferred asylum seekers in the agreements with Jamaica and the TCI were linked to the responsibilities of the US both the transferring and processing state, with 34 While the MOU with Jamaica was signed on 2 June 1994, the MOU with the TCI was signed more than a fortnight later, on 18 June 1994. 35 While no sources were found that detail the exact nature of the negotiations, the UK's involvement in those negotiations can be seen in its extended powers to terminate the arrangements ( 39 Of interest here is the involvement of the UK. In particular, the UK could not only monitor the processing of RSD claims, but it could also ensure that the interdiction (though not the RSD processing) took place in accordance with the UK's (and thus the TCI's) international obligations (US-TCI MOU, Articles II(1)(ii) and II (3)). The UK was even entitled to terminate the agreement should those obligations be breached (US-TCI MOU, Article VIII (3)).
Jamaica and the TCI undertaking a merely "facilitative role" (Francis 2008, p. 293). The US-Jamaica MOU clearly stated that Jamaica was to permit the US to conduct an operation in which the officials of the latter country would conduct RSD processing on the territorial sea and internal waters of Jamaica (US-Jamaica MOU, Preamble, Articles 1-3). It further stated that no obligation was to be imposed on Jamaica to "grant or consider granting of refugee status . . . for the purposes of the operation" (US-Jamaica MOU, Article 12). Finally, the protections that were required of the US in the US-Jamaica agreement were merely cursory. 40 While the UNHCR was invited to observe and provide guidance during RSD processing (US-Jamaica MOU, Article 7), it was seen as a way of creating a public perception of adhering to international protection standards, rather than doing so in practice (see Francis 2008, pp. 296-97; see also McKinley 1994, pp. 96-99;McKinley 1995, pp. 206, 211;and cf. O'Neill 1994, pp. 45-48).
While the processing of refugee claims in the US-TCI agreement was also to be conducted by the US (US-TCI MOU, Article II(1)(i) and (iii)), the protections were potentially more robust. 41 In addition to the monitoring and advisory role outlined for the UNHCR in the US-Jamaica MOU, the US-TCI MOU tasked the UNHCR to advise on the application of international law in RSD processing, as well as "to provide technical assistance and training to the Government of Turks and Caicos Islands" and the government of the US (emphasis added; US-TCI MOU, Preamble, Articles II(1)(xi), III(1)(i) and (ii)).
Conversely, the agreement between the US and TCI actually limited the rights of refugees by requiring legislation to be passed that not only offered immunity to US personnel associated with the processing centre, but also "oust[ed] the jurisdiction of the courts in the host country to hear claims against any officials acting in relation to an [extra-territorial processing centre]" (US-TCI MOU, Article II(2)(xi)(a); Francis 2008, pp. 306-7). This requirement effectively precluded access to the European Court of Human Rights (ECtHR) under the Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR) (CoE 1950) to which the TCI was (and still is) a party based on the UK's territorial application. 42

2001 Pacific Solution
At the outset, it must be noted that the 2001 Pacific Solution MOUs between Australia and Nauru and PNG are not publicly available. Nonetheless, they were subject to commentary by secondary sources, which are relied on in this article to discuss some of the elements of the 2001 Pacific Solution agreements. These sources confirm that processing in the respective agreements 43 was undertaken by the transferring state, that is, Australia (Francis 2008, pp. 293-95 referring to the 2005 Australia-Nauru MOU, paragraph 3). As with Jamaica and the TCI, Nauru and (presumably) PNG played a "mainly facilitative" role (Francis 2008, pp. 293, 985 referring to 2005 Australia-Nauru MOU, Schedule A). 40 Only the Refugee Protocol (not the Refugee Convention) was relied on with regards to international protections due to the fact that the US is only a signatory to the Protocol. Otherwise the law of the US was applied (US-Jamaica MOU, Preamble, Article 3(A)). Any further reference to either international law or Jamaican law related to the issues of health and sanitation, damage to the environment and immunities (US-Jamaica MOU, Articles 4(D), 4(G), 8 and 10). For further analysis, see Francis (2008). 41 The stronger protections were evidenced by the MOU's specifying that the processing facility should be operated in line with international standards (US-TCI MOU, Preamble). Unlike the US-Jamaica MOU that only referenced the Refugee Protocol, the US-TCI MOU undertook that processing was to be done in accordance with US law, in alignment with the standards of the Refugee Convention and its Protocol (US-TCI MOU, Article II(1)(a)(i)). In other ways, the US-TCI MOU was equivalent to that of the US-Jamaica one, specifically when equating the relevant rules of international law to those relating to immunities (US-TCI MOU, Articles IV(i) and VI). Other clauses explicitly limited the rights of asylum seekers. In particular, the processing of each refugee claim was only to have taken seven days (US-TCI MOU, Article II(1)(i)), the refugees were to be detained under guard at the processing centres, and the TCI was to pass legislation to legalise such actions (US-TCI MOU, Article II(1)(vi) and II(2)(xi)(a)). For further analysis, see Francis (2008). 42 As a UK dependency, TCI is a party to the ECHR and those on its territory, therefore, have access to the ECtHR (CoE n.d.). 43 See Francis (2008)  Nonetheless, as far as it is possible to ascertain, the nature, benefits and state power dynamics of these agreements are instructive. Nauru and PNG, being strongly dependent on Australia for development aid, received significant financial benefits from participating in the 2001 Pacific Solution. Nauru, in exchange for hosting two processing detention centres, was promised desperately needed development aid to the amount of 26.5 million Australian dollars for power and water generation, education and health (Senate Select Committee 2002, para. 10.27-10.38). While development aid was not explicitly part of the MOU with PNG, aid to infrastructure and the PNG Defence Force was nonetheless given or accelerated in conjunction with the agreement (Senate Select Committee 2002, para. 10.54-10.55). A meeting between Australia and PNG, held 8 October 2001, also discussed "the first tranche of Australian assistance for the reform of the Papua New Guinea Defence Force, totalling $20 million" (Senate Select Committee 2002, para. 10.46). The agreement itself further "committed Australia to support PNG, through advice and technical and financial assistance, in its management of nationals from third countries who are illegally entering the country" (Senate Select Committee 2002, para. 10.56). 44 Despite the 2001 MOUs specifying that all MOU-related activities were to be conducted in Nauru and PNG "in accordance with their own constitutions and relevant laws" (Senate Select Committee 2002, para. 10.62), it is likely that, as was the case in the US Safe Havens arrangements, this referred to all activities linked to the establishment and running of the processing centres, rather than the RSD processing of asylum seekers. Commentary on the 2001 MOUs does not explicitly mention human rights clauses (see Senate Select Committee 2002; Francis 2008). Under the MOU with Nauru, Australia was to respect its responsibilities under the Refugee Convention until a durable solution for the recognised refugees was found (Francis 2008, p. 292 referring to the 2002 Australia-Nauru MOU, paragraph 22). Additionally, while the original MOU with Nauru did require Nauru to not refoule any person transferred by Australia under the arrangement, a later version of the MOU omitted this requirement (Francis 2008, pp. 298-90 referring to the 2001 Australia-Nauru MOU, paragraph 30, and the 2004 Australia-Nauru MOU, paragraph 24, and omitted in the 2005 Australia-Nauru MOU).
As argued by Francis (2008)-similar to the US agreements with Jamaica and the TCI-the asylum seekers and refugees were under the "principal or exclusive control" of Australia, despite being on the territory of either Nauru or PNG, with the receiving countries taking on merely a facilitative role (Francis 2008, p. 295). These circumstances, together with the jurisdictional challenges of appealing either the detention or the refugee status decisions made by the transferring states, effectively eliminated any protections that could or should have been offered by the receiving countries (see Francis 2008, pp. 304-9).

2012 Pacific Solution
As a precursor to the 2012 Pacific Solution, an MOU was signed between PNG and Australia on 19 August 2011. This MOU was almost identical to the 2012 Australia-PNG MOU, with the notable exception of the clauses committing PNG to RSD processing. The absence of these clauses suggests that in 2011, a repeat of the 2001 Pacific Solution was envisaged, where an assessment centre was to be set up on Manus Island but RSD was to be conducted by Australia. As a result of these similarities (apart from the one major difference of RSD processing), the 2011 MOU between Australia and PNG is discussed alongside its 2012 version, which, together with the 2012 MOU between Australia and Nauru marks the start of the 2012 Pacific Solution. The 2012 MOUs signed between Australia and Nauru (29 August) and Australia and PNG (8 September) were themselves superseded by the 2013 MOUs signed on 3 August with Nauru and on 6 August with PNG. The MOU with PNG complements a Regional Resettlement Arrangement (RRA) between the two countries signed on 19 July 2013 . 44 The "national from third countries" refers to West Papuan refugees (see Glazebrook 2014;Taylor 2010). There are, in fact, several clauses promoting human rights and other protections of refugees in these MOUs. 45 For both Nauru and PNG, Australia commits to covering all costs under the respective MOUs. 46 While absent from the 2012 MOU with Nauru, aid to Nauru is not precluded. The 2013 MOUs further contain a general "broader benefit" to the communities in which the refugees are to settle in both countries. 47 The PNG MOUs contain a commitment to an additional "package of assistance" in line with the PNG-Australia Partnership for Development. 48 Outside of the financial benefits, all agreements are heavily focused on border control, as well as cooperation on irregular migration and disrupting people smuggling operations as their objectives. 49 As part of these objectives, Australia confirmed its 2001 pledge to PNG to support it "in its management of nationals from third countries who are illegally entering [PNG]". 50 The prioritisation of border control measures in the Preambles and Objectives sections can be seen in the ambivalence towards protections for the transferees. The Preambles merely state that "[t]he Participants are State parties to the 1951 Convention Relating to the Status of Refugees and its 1967 Protocol". 51 Similarly, the Preambles recognise "the need to take account of the protection needs of persons who have moved irregularly who may be seeking asylum", at the same time as "the need to ensure . . . that no benefit is gained through circumventing regular migration arrangements". 52 More certainty is further seen in the MOUs, where the obligatory 'will' features in the clauses on human rights. Specifically, "[t]he Participants will ensure the Transferees will be treated with dignity and respect and that relevant human rights standards are met" (emphasis added); 53 and "[s]pecial arrangements will be developed and agreed to by the Participants for vulnerable cases including unaccompanied minors". 54 The MOUs define a 'Refugee' as per the Refugee Convention definition. 55 With the exception of the 2011 Australia-PNG MOU, they also each contain a clause in which the respective governments assure the Australian government that they will "make an assessment, or will permit an assessment to be made" on whether a transferee meets the definition of refugee as per the Refugee Convention and its Protocol. 56 They also make assurances that they will not refoule a transferee whose life or freedom is threatened for convention reasons, 57 or send them to another country where there is a real risk that the transferee would be "subjected to torture, cruel, inhuman or degrading treatment or punishment, arbitrary deprivation of life or the imposition of the death penalty". 58 According to the MOUs, "the Participants will establish a Joint Committee with responsibility for the oversight of practical arrangements required to implement this MOU". 59 The Joint Committee is to meet regularly (changed from once a month in the 2011 Australia-PNG and the 2012 MOUs) and "may include relevant non-government organizations and service providers where appropriate". 60 This is an important clause not only because it envisages a form of oversight but also because it includes non-government organisations as part of that oversight. While little information is available on the workings of the Joint Committee, the UNHCR did take up an oversight role conducting monitoring visits (UNHCR 2012, para. 50;UNHCR 2013a, para. 106;UNHCR 2013b, para. 116;UNHCR 2013c, para. 82;UNHCR 2013d, para. 129-32).
The 2011 and 2012 MOUs with PNG contain two extra points of protection-in addition to those offering a guarantee of non-refoulement discussed above-that are excluded from the 2012 Australia-Nauru and 2013 MOUs. While all MOUs state that all activities that relate to the MOUs are to be conducted in accordance with the respective Constitutions and laws of Australia and PNG/Nauru, 61 the 2011 and 2012 PNG MOUs provide that "[a]ll activities undertaken in relation to this MOU will be conducted in accordance with international law and the international obligations of both Participants". 62 The 2011 and 2012 MOUs with PNG also contain a definition of an 'Asylum Seeker' as someone who is seeking international protection. 63 The omission of these two additional points of protection from the Nauruan and 2013 MOUs, as well as the RRA with PNG, appears to intentionally exclude the requirement for Nauru and PNG to provide international protection in line with international law and obligations, beyond those of non-refoulement and those found in the Refugee Convention. 64 In addition to the clauses mentioned above, the 2013 MOUs incorporated the potential durable solution of settlement in Nauru and PNG. Both Nauru and PNG committed, with Australia's assistance, 65 to provide people found to be in need of international protection with (re)settlement opportunities, 66 by enabling their "lawful" stay in each respective country. 67  which relates to the lawful stay of not only refugees but also of asylum seekers, was missing in the 2012 MOUs. Additionally, limited that commitment to "an agreed number" of people. 68 Australia was also to assist the remainder of refugees to resettle in another willing third country and return those who had their asylum claim rejected. 69 Although the MOUs of the 2012 Pacific Solution could be more specific with regards to responsibilities under the human rights conventions, for instance the CAT, they (especially the 2011 and 2012 PNG MOUs) are quite comprehensive in ensuring protections for people being transferred. Nonetheless, while the focus of this article is on the effect of these agreements on the treaty ratification patterns of the receiving states, and, as mentioned earlier, these agreements are neither legally binding nor enforceable, it must be noted that monitoring by the UNHCR showed that the human rights requirements of the MOUs were not complied with in practice (see UNHCR monitoring observations: UNHCR 2012; UNHCR 2013a; UNHCR 2013b; UNHCR 2013c; UNHCR 2013d; UNHCR 2014b; UNHCR 2015a; UNHCR 2016).

EU-Turkey Deal
The EU-Turkey deal is an outcome of a number of statements based on decisions reached in earlier discussions between the EU states and Turkey. While not formal parts of the agreement, a number of documents, issued during the negotiation phase, contain supporting information (these include While the EU-Turkey deal is primarily a people swap arrangement that does not strictly contain offshore processing requirements, the agreement itself deals with themes similar to that of the other case studies. The agreement is very heavily focused on border control and anti-smuggling operations. This is particularly the case in the Joint Action Plan, in which the second of two parts is dedicated to "[s]trengthening cooperation to prevent irregular migration" (EU-Turkey Action Plan, Part II).
With respect to the allocation of financial and other benefits not associated with transfer costs, hosting, and enhanced border control operations, the EU-Turkey agreement differs somewhat from the agreements in other case studies, which mainly benefitted by additional allocation of development aid. The agreement does offer financial support to Turkey, focusing not only on covering the costs of the transfer of persons from Greece, but also financial support of two instalments of 3 billion Euros under the Facility for Refugees. These funds are in support of "Syrians under temporary protection and their host communities in Turkey" (EU-Turkey Action Plan) in the areas of Humanitarian assistance; provision of legal, administrative and psychological support; support for community centres; the enhancement of self-sufficiency and participation in economy and their social inclusion during their stay in Turkey; improved access to education at all levels; but also actions supporting host communities in areas such as infrastructures and services. (EU-Turkey Action Plan, Part I EU point 1) This difference in the allocation of funds is likely due to the relative equality of Turkey as a negotiation partner. While Turkey is seen by the EU as a transit state along the Eastern Mediterranean Route, it is primarily a destination state in its own right. By 2015, Turkey was host to the largest number of refugees in the world, which further strained the limited capacity and quality of protection provided to asylum seekers (İçduygu andŞimşek 2016, pp. 62-69). In the EU countries, right wing and populist the 2013 Australia-PNG MOU (paragraph 14) complements PNG's commitment to the limited withdrawal of social and economic reservations to the Refugee Convention (2013 Australia-PNG RRA, paragraph 7) by committing to treat all those who settle in PNG in accordance with the Refugee Convention. 68 2013 Australia-Nauru MOU, Preamble, paragraph 12. 69 2013 Australia-Nauru MOU, paragraphs 13-14;2013 Australia-PNG RRA, paragraphs 5, 8. parties were rising in the political spectrum, creating a 'threat' understanding of human mobility (Hameleers 2019;Onraet et al. 2019;Bigo and Guild 2019, pp. 10-11). For European states, therefore, the deal was a measure to externalise the borders and reduce the number of incoming migrants and asylum seekers. For Turkey, the deal was an effective way of bringing the 'burden sharing' discourse into a tangible reality (Çorabatır 2016, p. 17; see also EU-Turkey Action Plan). Consequently, both the EU and Turkey had immense domestic pressures to better manage increased displacement caused by the Syrian and other conflicts. The EU-Turkey deal therefore was an opportunity in easing the political tensions for both sides, with both sides taking an active role in the summit negotiation. Thus, instead of development aid, Turkey and the EU "re-energised" Turkey's accession process to the EU, the Visa Liberalisation Roadmap, 70 as well as the cooperation on the Customs Union and Energy (EU-Turkey Statement). 71 The few human rights mentioned in the EU-Turkey agreement are linked to the 'one-for-one' people swap scheme. According to the agreement, Turkey would readmit all irregular migrants going to Europe (EU-Turkey Statement, point 1) and "[f]or every Syrian being returned to Turkey from Greek islands, another Syrian will be resettled from Turkey to the EU" (EU-Turkey Statement, point 2). While this arrangement differs from the offshore processing requirements in the other agreements, the EU-Turkey agreement shares the sparseness and non-enforceability of human rights 'clauses' of the other case studies. In discussing the return of irregular migrants as part of the people swap arrangement, the agreement includes only a passing reference to international (human rights) law in its very first point: All new irregular migrants crossing from Turkey into Greek islands as from 20 March 2016 will be returned to Turkey. This will take place in full accordance with EU and international law, thus excluding any kind of collective expulsion. All migrants will be protected in accordance with the relevant international standards and in respect of the principle of non-refoulement. . . . (emphasis added) (EU-Turkey Statement, point 1) Interestingly, this point appears to offer human rights protection to the 'migrants' being removed from the EU, thus referring to the already existing international and EU law protection obligations of EU states towards the irregular arrivals affected by this arrangement. It is not clear if the third sentence applies solely to the EU states or to both EU states and Turkey, as the 'clause' then returns to the obligations of the Greek state in processing asylum applications. If it does, in fact, apply to both EU states and Turkey, it is interesting that the less onerous term of 'international standards' is used compared to the 'international law' of the previous sentence that applies solely to the EU states. 72 Thus, the Statement of 18 March 2016 contains no explicit protection obligations to be put on Turkey.
Similarly, while the Joint Action Plan does identify some of Turkey's intentions towards the displaced people within its territory, no explicit mention of human rights is present. Turkey's intentions include continuing and further enhancing "the effective implementation of the law on foreigners and international protection"; registering migrants and providing them with documentation; through legislation, opening up access for Syrians under temporary protection to education, health services and access to the economy; identifying and taking care of vulnerable people (EU-Turkey Action Plan, Part I Turkey points 1-4); as well as quickly completing any initiated asylum procedures (EU-Turkey Action 70 However, the liberalised visa regime has not so far been put into action. This is mainly because there were still seven benchmarks to be finalised by September 2016. Following the 15 July coup attempt in Turkey, the government introduced a prolonged State of Emergency, which affected the adoption of these benchmarks, including some sensitive topics, such as revising legislation and practices on terrorism in line with European standards (European Commission 2016a). 71 See also the CoEU (2015) for the establishment of the Strategic Energy Cooperation and High Level Energy Dialogue. 72 While 'international standards' are broader in nature, and may encompass the full spectrum of the human rights applicable to refugees, "they cannot be considered part of international treaties" and are therefore not enforceable (Rüdiger 2010). "Respecting international law", on the other hand, can be directly linked to treaties ratified by the state in question and therefore has the potential to be enforceable through decisions of an international, regional or national court or body (depending on the legal system of the said state) (see generally, Shaw 2014).
Plan, Part II Turkey point 4). The lack of explicit mention of human rights requirements for Turkey in the EU-Turkey agreement is a significant omission, as the condition and fate of transferees from the EU, some of whom end up being asylum seekers, remains largely unknown. Reports, however, suggest that returnees are unable to apply for asylum and are subjected to prolonged detention and refoulement ). Finally, missing from the EU-Turkey agreement is the requirement for Turkey to remove the geographic limitation to the Refugee Convention. Turkey still retains the geographical limitation to the Refugee Convention, meaning that only refugees from European states are granted international protection. Other asylum seekers are provided temporary protection and labelled as 'conditional refugees'-a status that provides little hope of a durable solution in Turkey, unless refugees are able to return to their countries of origin or be resettled (Kaya 2017, pp. 120, 160). The Joint Plan of Action actually does explicitly state that Turkey is a party to the Refugee Convention "with geographic limitations". However, this is stated only in a footnote and no further discussion or even mention can be found in the agreement or its supporting documents.
The omission of the requirement of Turkey to remove its geographic limitation differentiates the EU-Turkey deal from other EU-Turkey agreements that affect asylum seekers and refugees, namely the accession negotiations and visa liberalisation roadmap. Since Turkey's candidacy to the EU first began to be discussed in the late 1990s, the condition that Turkey lifts the geographical limitation to the Refugee Convention remained a requirement (European Commission 1998, p. 44). However, since the introduction of the Foreigners and International Protection Law 2013 (which came into effect in 2014), EU's progress reports on Turkey have ceased to focus on requesting that this requirement be met. The reports merely state that Turkey maintains its geographical limitation and that the protections offered to 'conditional refugees' are not as strong as would be offered to those to whom the limitation does not apply (that is, refugees from Europe), without further elaboration on the subject (see, for example, European Commission 2014a, p. 64; European Commission 2019, p. 47). 73 It is tempting to argue that the omission of the requirement from the EU-Turkey deal is in line with both Turkey's legislative developments and EU's policy change towards Turkey; however, the situation is somewhat more complex. While Turkey's legislative amendments were significant, they were not, in fact, sufficient to designate it as a safe country, nor to make the removal of the geographical limitation an insignificant step (see, for example, Roman et al. 2016;Alpes et al. 2017;Gkliati 2017;Kaya 2017, pp. 132-35). When linked to the accession dialogue, the easing of EU's position on this requirement is not only linked to the legislative amendments (and negotiations for the Readmissions Agreement) but also to the fact that Turkey always maintained that it would remove the geographical limitation, but only upon being accepted to the EU and not before (Derviş et al. 2004, p. 34;Kaya 2017, p. 120).
This argument cannot and does not apply to the EU-Turkey deal, which directly affects asylum seekers and refugees. The presence of the requirement to remove the geographical limitation in the accession and visa liberalisation agreements means that, while the EU can argue that the requirement is met in practice while it believes it (or finds it politically expedient) to be so, it can always insist on the requirement again, should the situation change. The omission of this requirement together with other human rights protections from the EU-Turkey agreement means that the asylum seekers and refugees impacted by this deal are left unprotected on the Turkish side. With the nature of the deal being a response to the influx of displaced persons in need of protection, and since this protection does not entail a possibility of a durable solution in Turkey, the exclusion is telling. 73 Interestingly, the removal of the geographical limitation was also a requirement of the visa liberalisation roadmap (European Commission 2013, p. 10). However, the first report suggested that the new legislation made the removal merely "symbolic" and future reports did not mention the limitation at all (European Commission 2014b, pp. 16-17).

Contents of Offshore Processing Agreements in Summary
Before looking into treaty ratification patterns of the receiving states in offshore processing arrangements, it is worth extracting the key points from the contents of the agreements of each case study. The offshore processing agreements (with the notable exception of the US-Jamaica MOU) are heavily focused on border control and anti-smuggling operations, for which the transferring state often offers financial and technical support, such as capacity building.
With regards to the power dynamics between the partnering states, it appears to influence the nature and amount of assistance negotiated, rather than the contents of the agreements, much less the strength of the human rights clauses within. With the US and Australia being the regional hegemons, Turkey stands out as the country that received the most financial and tangential benefits for its refugee and host communities. Although, the bigger play for EU accession and/or visa liberalisation in the end fell through. Even within the Caribbean example, the TCI, with UK's backing, could be seen to reap better benefits, despite the processing never taking place, than did Jamaica, whose primary (though intangible) benefit appears to be the potential for closer ties with the US.
The most instrumental distinguishing factor with regards to the contents of the agreements is, thus, whether the receiving or transferring country conducts the RSD processing. 74 The best examples of this are the MOUs between Australia and Nauru during the 2001 Pacific Solution and the 2012 Pacific Solution. The early agreements of the former arrangement only required Nauru to respect the principle of non-refoulement. The agreements of the latter arrangements required Nauru to not only respect the principle of non-refoulement for convention reasons and where people might be "subjected to torture, cruel, inhuman or degrading treatment or punishment, arbitrary deprivation of life or the imposition of the death penalty", 75 but also to treat the transferees with dignity, respect and to meet the relevant human rights standards. An even clearer example is seen in the requirement to exclude the jurisdiction of national and regional courts in the US-TCI agreement. Additionally, while some transferring states require of themselves (in the best-case scenario) to respect international law, the receiving states are only asked to meet relevant human rights or international standards. 76 As a final observation, in none of the case studies were any treaty ratification requirements included as part of the contents of the agreements. There is merely a statement that the receiving countries are state parties to the Refugee Convention and its Protocol, irrespective of the significant reservations registered, for instance, by Turkey (geographic limitation) and PNG (omitting socio-economic rights). With regards to the EU-Turkey deal in particular, omitting the request to remove the geographic limitation can be argued to be a step backwards from previous deals concerning refugees and asylum seekers, as the requirement was part of other EU-Turkey arrangements.
In light of the foregoing, any effect on treaty ratification as a result of the said agreements is unlikely. However, while general patterns emerge, there are often nuances that need to be identified. These are discussed below by first looking at general patterns of treaty ratification (Section 3.1). This is followed by a discussion of offshore processing arrangements (including their agreements) and their influence on the ratification of the Refugee Convention, its Protocol, and the core human rights treaties by the receiving states (Section 3.2), as well as on the ratification of other relevant treaties (Section 3.3).

Ratification of Refugee and Human Rights Treaties
As discussed earlier, ratification of the 1951 Refugee Convention and its 1967 Protocol, as well as the core human rights treaties (ICERD, ICCPR CPED) is one of the elements of an effective asylum system. This is because human rights treaties offer additional protections to the asylum seekers and refugees (UNHCR 2017, pp. 23-26). Apart from the Refugee Convention and its Protocol, the nine core human rights treaties are discussed holistically in this article, rather than by assessing the ratification of each individual treaty in detail. As was seen in the discussion in the previous section, the disparities in the time-periods and the geo-political development of each of the arrangements in the four case studies make both comparisons, and conclusions based on these comparisons, quite difficult. A comparison is therefore not attempted. What follows is a descriptive analysis of each of the case studies individually, with conclusions drawn about the receiving country's ratification events for a period of up to a year prior to the respective arrangements, as well as for the duration of those arrangements. A further factor is the complexity of the broader topic of general patterns of treaty ratification by states. These patterns are addressed first.

General Patterns of Treaty Ratification
Countries choose to ratify (and not to ratify) human rights treaties for a multitude of reasons. To a large part, this choice depends on a state's characteristics. These include the extent of democracy (including whether it is a newly democratic/independent state), normative framework and respect for human rights, level of development and/or dependence on foreign aid, as well as the ratification rate (high or low) of the region to which the state belongs. A different combination of these characteristics was shown to result in different patterns of treaty ratifications and compliance.
Observations of the general patterns of treaty ratifications indicated that democratic states (that is, liberal states with strong domestic institutions) with good human rights records are more likely to ratify more human rights treaties (Cole 2005, pp. 490-91;Neumayer 2005, p. 950), while democratic states with poor human rights records are less likely to ratify them (Hathaway 2007, p. 613). Ratification of treaties by democratic states tends to improve their human rights records, as they relate to that specific treaty (Neumayer 2005, p. 950). However, the outcomes are somewhat reversed for non-democratic states. Non-democratic states (that is, autocratic states with poor domestic institutions) with poor human rights records are more likely to ratify human rights treaties, especially when they are more dependent on foreign aid. They do this to garner a better international reputation through ratification, rather than adherence (Spence 2014, p. 429). As a result of this 'subterfuge', ratification of treaties by non-democratic states tends to worsen their human rights records, as they relate to that specific treaty (Neumayer 2005, p. 950).
Other patterns of treaty ratification might reflect whether a state is a newly independent state and the geographic region it belongs to. For instance, newly independent (democratic) regimes are more likely to ratify human rights treaties to commit future governments to the human rights norms (Cole 2005, pp. 490-91;Hathaway 2007, p. 613;Spence 2014, p. 426). However, not yet having established the enforcement mechanisms for a complaints procedure, they are not likely to ratify human rights enforcement treaties, usually in the form of Optional Protocols or inquiry procedure articles in the treaty itself (Cole 2005, pp. 490-91;Hathaway 2007, p. 613). Finally, countries belonging to a geographic (normative) region are likely to have similar ratification patterns, meaning that countries in the geographic region with higher ratification rates are more likely to ratify human rights treaties and vice versa (Hathaway 2007, p. 598).
The last two points are particularly pertinent to the receiving countries of the case studies, as all of them with the exception of Turkey, gained independence in the decolonisation era. Jamaica became independent from the United Kingdom in 1962, while Nauru and PNG gained independence from the Australian Trusteeship in 1968 (UNGA 1967a) and 1975(UNGA 1974, respectively. The TCI is an exception, as it remains a dependency of the UK. While the available ratification data for the TCI is limited, as a dependency of the UK, territorial application of the UK's ratifications applies to all of TCI's ratifications of the six core human rights treaties (being the ICCPR and ICESCR in 1976, the CEDAW in 1986, the CAT in 1988and the CRC in September 1994, and as well as the ECHR in 1964) (see UNTC, Chapter IV: Human Rights n.d.; CoE n.d.). As a result, the TCI is difficult to classify with regards to its ratification events. This is because it is both a territory of the UK and a member of the Caribbean region. It, thus, can be argued to straddle the normative framework of both the Caribbean and the UK (see Clegg 2005).
Jamaica's position is clearer. Since its independence, Jamaica ratified the majority of the core international human rights treaties, 77 with the notable exception of the CAT. The early ratification of these treaties by the Jamaican government in its post-independence era is emblematic of its foreign policy, which is supportive of regional and international normative agendas for human rights (Helfer 2002). 78 Jamaica ratified those treaties that were in keeping with its national interests, and which largely reflected a culture for protection of the vulnerable (for example, women and children). Conversely, not ratifying the CAT and denouncing the ICCPR Optional Protocol reflected a lack of synergy with domestic culture/practice (for example, support for the death penalty) (Heyns and Viljoen 2001, generally and p. 496). This was in line with the argument that Jamaica, since its independence, followed the pattern of a region with relatively high ratification rates of the core human rights treaties. 79 This pattern is further confirmed by the fact that the majority of the Optional Protocols and inquiry procedure articles 80 are not ratified. Small and newly independent states often lack capacity (human, financial and technical resources) to implement and enforce the ratified treaties (Heyns and Viljoen 2001; see also Dietrich Jones (forthcoming)).
It could be argued that Nauru and PNG, having gained independence in the post-colonial era, have similar characteristics to Jamaica, as they relate to treaty ratification patterns. However, they differ in one very important element. They belong to a region with, on average, a very low rate of ratification of the core human rights treaties. 81 The low ratification rate by Pacific states was attributed to factors that include a lack of resources (financial, human and institutional) to implement the treaties, as well as comply with their reporting obligations (Baird 2011, pp. 20-22;cf. Olowu 2006). Additionally, states are cautious of a potential mismatch between the human rights requirements and cultural realities, seeing human rights norms as belonging to Western values and potentially amounting to an infringement on sovereignty (Baird 2011, pp. 15-19;cf. Olowu 2006). This position is, however, shifting, with a comparatively significant increase in human rights treaty ratifications seen across the region in the 2010s (see OHCHR n.d.).
Turkey is adhering to a pattern of treaty ratification linked to its traditional westernisation and modernisation policy, defined at the start of its republican era in the 1920s and 1930s (Derviş et al. 2004, pp. 6-7, 10-12). Even so, until the late 1990s, Turkey only ratified three of the core human rights treaties (CEDAW in 1985, CAT in 1988and CRC in 1995 and the regional ECHR (in 1954) 82 (UNTC, Chapter IV: Human Rights n.d.; CoE (2020)). However, it was observed that while Turkey ratified 77 In 1966 Jamaica signed the ICCPR (ratified 1975), the CERD (ratified 1971), and the ICESCR (ratified 1975). It later ratified CEDAW in 1984 andCRC in 1991. Of the more recent treaties, Jamaica ratified the CRPD in 2007 and ICMW in 2008, omitting only the CPED (UNTC, Chapter IV: Human Rights n.d.). 78 In fact, Jamaica "became the main broker of progress in UN human rights diplomacy from 1962 to 1968 . . . [and] had a profound and long-lasting effect on international human rights work" (Jensen 2016, p. 71). 79 Of the six core human rights treaties in force prior to the 1994 US-Safe Havens arrangement, the average ratification rate of the Caribbean region was 4.2 treaties, with Jamaica ratifying five of them (CERD, ICCPR, ICESCR, CEDAW and CRC), the exception being the CAT. The current ratification average in the region is 6.3 of the nine core human rights treaties with Jamaica having ratified an additional two (ICMW and CRPD) with the exception of the CPED (OHCHR n.d.). 80 With the exception of the ICCPR Optional Protocol (ratified in 1975 denounced in 1997 in response to the multitude of communications relating to the protocol, UNTC, Chapter IV: Human Rights n.d.; Heyns and Viljoen 2001, p. 514), and the two CRC Optional Protocols, which were ratified in 2002 and 2011 respectively. 81 Of the six core human rights treaties in force prior to the 2001 Pacific Solution arrangement, the average ratification rate of the Pacific region was 1.7 treaties (significantly lower than that of the Caribbean), with Nauru ratifying one treaty (CRC) and PNG ratifying three (CERD, CEDAW and CRC). The current ratification average in the region is 4.3 of the nine core human rights treaties, with both Nauru and PNG ratifying a further three each. Nauru catching up on CEDAW and CAT and ratifying CRPD and PNG catching up on ICCPR and ICESR and ratifying CRPD (OHCHR n.d.). 82 Turkey accepted the right to individual application to the ECtHR in 1987 and recognized the compulsory jurisdiction of the ECtHR in 1990 (Republic of Turkey 2020).
human rights treaties, it followed a pattern of doing so without intention or capacity to implement them into domestic law (Dutton 2012, p. 56;Alkan-Olsson and Alkan-Olsson 2012, pp. 15-17). It is only since the beginning of Turkey's candidacy to the EU in the late 1990s that Turkey's ratification both of international and regional human rights treaties (and to varying degrees their domestic implementation) spiked in the 2000s (UNTC, Chapter IV: Human Rights n.d.; CoE 2020;Derviş et al. 2004, p. 21). It is worth noting that Turkey's domestic drive coincides with the EUs 'conditionality' tactics, which uses the incentives of accession and of policy (offering visa liberalisation schemes) to "transform the governing structures, economy and civil society of the candidate countries" (Derviş et al. 2004, p. 16;Kaya 2017, pp. 123-24). It, therefore, appears that Turkey's ratification events are driven by international and regional reputation and are based on incentives of strong domestic benefit.
Ratification of the 1951 Refugee Convention and its 1967 Protocol, in particular, follows the general pattern of treaty ratification of each receiving state. Turkey ratified the Refugee Convention and its Protocol early, in 1962 and1968, respectively. Jamaica ratified the Refugee Convention in 1964 and its Protocol in 1980. PNG ratified the Refugee Convention and its Protocol simultaneously in 1986, as did Nauru in 2011. The TCI was the last to ratify sometime in 2014 (the exact date being difficult to ascertain). 83 While no direct connection can be drawn, this ratification event coincides with the region-wide adoption of the 2014 Brazil Declaration and Plan of Action (Cantor 2018, pp. 9, 64). 84 With the exception of Nauru, all receiving states registered reservations. 85 The most significant, and thus relevant to the discussion of this article, are the geographic reservation made by Turkey and the socio-economic reservations made by PNG. The Turkish geographic reservation technically limits the protections of the Refugee Convention to only those people who are seeking refuge out of Europe. 86 The PNG socio-economic reservations exclude recognised refugees from having access to protections relating to wage-earning employment, housing, public education, freedom of movement, non-punishment of unlawful entry, expulsion and naturalisation. 87 Having established the general patterns of treaty ratification followed by the receiving states, all of whom ratified at least one of the core human rights treaties prior to the offshore processing arrangements taking place, it is now possible to take a look at whether these patterns were affected by the offshore processing arrangements and, if so, in what way(s).

Treaty Ratification and Offshore Processing Arrangements
This section is not able to demonstrate a direct link between each of the offshore processing arrangements and ratification events. Instead, the negotiation period of the offshore processing arrangements-summarised to a year before the actual agreement-together with the period of duration of the arrangements is examined to determine whether any treaty ratifications took place in that time period. Where there were ratification events, secondary sources were consulted, while referring to and taking into account the contents of the agreements signed in each of the case studies, to ascertain whether there is in fact a stated link between the arrangements and the ratification event. 83 No formal communication with regards to territorial application has been deposited by the UK with the UN, nor is there a record of the TCI making such a communication independently. Although it may be possible that it has not as yet been recorded (see UNTC, Chapter V: Refugees and Stateless Persons n.d. as at 6 July 2020). Nonetheless, no mention of the TCI in relation to the Refugee Convention is made in UN documents in (UNGA 2013. However, from 2014, the TCI is being referred to as a party to the Refugee Convention and its Protocol (United Nations UN; UNGA 2014, para. 42; UNGA 2020, para. 46). 84 The Brazil Declaration and Plan of Action is non-binding agreement among Latin American and Caribbean states, which commits states to protect the rights of refugees, the displaced and stateless persons. 85 No information is available on whether the TCI has registered reservations. 86 These reservations are, however, somewhat alleviated by the development of the Turkish national asylum system to provide refugees from all regions with temporary protection (see above discussion in Section 2.4). 87  Discussion of the case studies is divided based on whether the transferring state or the receiving state conducts RSD processing (see summary in Section 2.5 above). Thus, the US Safe Havens and 2001 Pacific Solution are grouped together for analysis, based on the nature of their arrangements as having RSD processing undertaken by the transferring states. The 2012 Pacific Solution is then evaluated individually, as the only arrangement that envisages RSD processing by the receiving state, with the EU-Turkey deal concluding the analysis.

US Safe Havens and 2001 Pacific Solution
With respect to the 1994 US Safe Havens offshore processing arrangements, in addition to the fact that treaty ratification was not a requirement of the US-Jamaica MOU, Jamaica already ratified the majority of the core human rights treaties post-independence and prior to the signing of the offshore processing agreement. 88 Therefore, treaty ratification was not a corollary of the arrangement between the two countries to facilitate the offshore processing arrangement. The same is true of the US-TCI MOU, where only the US is requested to adhere to the Refugee Convention and its Protocol. TCI's ratification of the Refugee Convention and its Protocol in 2014 is thus also an unrelated event.
With respect to the 2001 Pacific Solution, in the absence of the MOUs, a similar conjecture can be made to the US Safe Havens arrangements, due to the large gap that exists in ratification of core human rights treaties by PNG between 1995 (CEDAW) and 2008 (ICCPR and ICESCR) (UNTC, Chapter IV: Human Rights n.d.). Nauru, however, signed a number of the core human rights treaties and their protocols on 12 November 2001 (the CERD and ICCPR are not yet ratified; the CAT was ratified in 2012) (UNTC, Chapter IV: Human Rights n.d.). 89 This is two months after the signing of an Agreement with Australia (2001 Australia-Nauru Agreement) and a month before signing of the MOU (2001 Australia-Nauru MOU).
Even so, no evidence found so far sheds light on the reason for these signatures and whether they have any relationship to the Pacific Solution, or were instead an expression of the Nauruan government's pursuit of other national, regional or global goals. While it is impossible to state one way or the other, in the absence of supporting documentation, the fact that no such signatures were made by PNG, together with the fact that Nauru neither signed nor ratified the Refugee Convention during the 2001 Pacific Solution, suggests that it is unlikely that the signatures related to the arrangements that were made around the same time.
However, some doubt remains. It was already shown in the previous section that although not publicly available, it is unlikely that the MOUs pertaining to the 2001 Pacific Solution contained any clauses relating to human rights or the necessity of ratification of any treaties. The only (insignificant) exception existed in the early versions of the MOU between Australia and Nauru asking for an assurance of non-refoulement. This assurance might, in fact, be the only 'argument' towards the 2001 Pacific Solution arrangement influencing Nauru's signatures (see, however, the further discussion in Section 3.3 below).
The US Safe Havens and 2001 Pacific Solution case studies share the assumption that no RSD processing was to be done by the receiving state, namely Jamaica, TCI, Nauru and PNG, respectively. With RSD processing being the purview of the transferring states and the lack of any such requirements in the contents of the agreements, it is not surprising that the ratification of the Refugee Convention, its Protocol, the core human rights treaties, or the development of any other elements of an asylum system in the receiving states did not take place. 88 It must be noted that no core human rights treaties entered into force from 1990 (CRC) until 2003 (CPMW). As Jamaica ratified the core human rights treaties (apart from the CAT and the CPED) soon upon their entering into force, there were consequently no human rights treaty ratifications for Jamaica between 1992 and 2002. 89 Interestingly, while Nauru merely signed the human rights treaties (and the ICCPR and CAT Optional Protocols), on the same day, it ratified the Rome Statute of the International Criminal Court, as well as quite a number of treaties on disarmament and environmental protection (UNTC, Nauru n.d.; see also UNTC, Chapter XVIII: Penal Matters ; UNTC, Chapter XXVI: Disarmament n.d.; UNTC, Chapter XXVII: Environment n.d.).

2012 Pacific Solution
The 2012 Pacific Solution case study creates an interesting basis for analysis for the relationship between treaty ratification events and offshore processing arrangements. With the receiving states (Nauru and PNG) being entrusted with the processing of asylum claims, and later (in 2013) locally settling the refugee population, it can be argued that the offshore processing arrangements can affect the treaty ratification events of the receiving states. The clearest example of this is PNG, which had maintained significant reservations against the socio-economic rights upon the ratification of the Refugee Convention in 1986. Those reservations were maintained during the first year of the offshore processing arrangements and were only removed after the signing of the settlement agreement on 19 July 2013 and the superseding MOU on 3 August 2013. The language of the reservation removal clearly connects it to the offshore processing arrangements: On 20 August 2013, the Government of the Independent State of Papua New Guinea notified the Secretary-General, in accordance with article 42 (2) of the Convention, of its decision to partially withdraw its reservation made upon accession: " . . . In accordance with article 42, paragraph 2 of the Convention, I wish to communicate to you that Papua New Guinea withdraws its reservation with respect to the provisions contained in articles 17 (1) While this is indeed a ratification event that is linked to the offshore processing arrangements, it is particularly connected to the settlement of recognised refugees in PNG, rather than merely to the processing of their claims, which was the focus of the 2012 agreement. The reservation removal, however, does not apply to any other spontaneously arriving asylum or refugee populations in PNG, such as West Papuan refugees from the Indonesian province of Papua. It thus could not be argued to be truly a 'development' in the PNGs asylum system.
The connection of Nauru's ratification of the Refugee Convention and its Protocol on 28 June 2011 (UNTC, Chapter V: Refugees and Stateless Persons n.d.) to the 2012 offshore processing arrangements is less obvious. However, a line of argument can still be drawn connecting the ratification to the impending offshore processing agreement signed the following year. The rise in boat arrivals seen in 2010 coincided with an election year in Australia and the then opposition leader was already campaigning to restart the Pacific Solution and, in particular, to reopen the processing centre on Nauru (Abbott 2010a, 2010b). During the same campaign, the then Prime Minister maintained that restarting the Pacific Solution was out of the question, while at the same time approaching East Timor to host a regional processing centre as well as New Zealand and UNHCR for assistance and support (Gillard 2010).
The following year saw more decisive action. The Malaysia Solution was announced on 7 May 2011, the MOU signed on 25 July 2011, and the adverse High Court decision handed down on 31 August 2011 invalidating that MOU. 90 During the same time-period, a preliminary MOU was signed with PNG on 19 August 2011 to reopen the Manus Island processing centre. This MOU was not acted upon until it was superseded by the 2012 version. Nauru's ratification of the Refugee Convention and its Protocol falls within this four-month period. Australia's active attempts to find regional partners 90 One of the primary objections of the decision was the fact that the agreement with Malaysia was incompatible with requirements of the Migration Act 1958 as it stood at the time. This was because Malaysia was not a party to the Refugee Convention, did not have the domestic legislative protections necessary for the protection of asylum seekers, nor provided those protections in practice ( for offshore processing allows for the possibility that Nauru was previously approached and asked to ratify the Refugee Convention and its Protocol in preparation for a potential reopening of the processing centre. Further support for this connection can be found in the National Report submitted by Nauru to the UN Human Rights Council (HCR) Working Group on the UPR on 5 November 2010. The National Report contained the following paragraph: The possibility of reopening the Centre has been raised recently by the Australian government and the Australian opposition, and Nauru has indicated that it would be open to that possibility because of Nauru's desire to assist Australia and because of the economic benefits that the Centre would bring to Nauru. (HRC 2010, para. 90) The National Report also confirmed that Nauru was receiving advice from unspecified "officials" regarding the ratification of the Refugee Convention, which was "likely to be considered by the government of Nauru within the next few months" (HRC 2010, para. 36). Support for Nauru's ratification of the Refugee Convention can be found throughout the UPR documents. It is seen especially in connection with Nauru's involvement with Australia in hosting the processing centre (HRC 2011a, para. 63, 64), with several countries-with the conspicuous omission of Australia-recommending Nauru's ratification of the Refugee Convention and its Protocol (HRC 2011a, para. 79).
The final confirmation of the relationship between Nauru's ratification of the Refugee Convention can be found in the submission by UNHCR. It not only confirmed that Nauru had expressed an interest in reopening the processing centre as early as November 2008, but also that Nauru had considered ratifying the Refugee Convention in mid-2006 [W]hich undoubtedly related to the then Australian Government's desire to enact further legislation to strengthen the offshore processing arrangements. However, after the abandonment of the legislation, the Government of Nauru displayed no further interest and has not established a formal policy regarding asylum-seekers or refugees outside of the context of the Pacific Strategy. (UNHCR 2010b) In considering the timing and Nauru's willingness to ratify the Refugee Convention and its Protocol (the ratification took place in 2011 and the MOU was signed a year later in 2012), it must be concluded that it was not the contents of the agreement, nor Nauru's alignment with the normative content of the Refugee Convention and its Protocol that prompted Nauru to ratify. This is especially, as Nauru is not a country that is impacted by spontaneous arrival of asylum seekers or refugees (UNHCR 2015b, p. 1). Instead, Nauru is highly dependent on the boost to its meagre economy that the offshore processing of asylum seekers brings and it is that specific national interest that likely motivated its ratification events more than any normative or even persuasive arguments of the more powerful Australia (HRC 2010, para. 90;see additionally Senate Select Committee 2002, para. 10.40-10.44;Maley 2003, p. 195;Connell 2006, pp. 55-59;Fraenkel 2016, p. 283;Neilson 2017).
Finally, Nauru's ratification of other human rights treaties in the following years should not be omitted. Until that point, of the nine core human rights treaties, it had only ratified the CRC. However, in contrast to the Refugee Convention, the documentation of the UPR shows no support for the connection between ratification of the core nine human rights treaties and the offshore processing arrangements. Nauru had expressed its reservations on ratifying treaties based on the onerous reporting duties that these ratifications entail (HRC 2011a, para. 20). However, it had "recently" established a Working Group on Treaties, which was reviewing the UPR recommendations on ratifications and acting accordingly (HRC 2011a, para. 21). Consequently, between 2011, Nauru also ratified the CEDAW in 2011, the CAT and CRPD in 2012, and the CAT Optional Protocol in 2013 There is, however, no evidence that these ratifications are connected to the offshore processing arrangements. They follow the regional trend by Pacific countries, which saw an increase of ratifications of human rights treaties in the 2010s. Similarly, the ratifications follow a further pattern of treaty ratification that suggests that states participating in human rights conferences (or reviews such as the UPR) are more likely to ratify human rights treaties (Cole 2005, p. 491). This does appear to apply to Nauru, as this was indeed the first cycle of the UPR process. The UPR was introduced by the UN in 2008 and Nauru's first review was not until early 2011. This lack of connection to offshore processing arrangements can also be seen in PNG's single ratification, in 2013, of the CRPD. 91

EU-Turkey Deal
Before going on to the ratification of other relevant treaties, the case of EU-Turkey is worth discussing. This is because Turkey's treaty ratification trajectory also appears to remain un-impacted by the EU-Turkey arrangement, with the last core human rights treaty (the CRPD) ratified in 2009 (UNTC: Chapter IV: Human Rights). Instead, Turkey's treaty ratifications are seen as path-dependent on meeting the accession and visa liberalisation requirements. However, as discussed above, the requirement for removing the geographical limitation was eased (if not removed) in both sets of dialogues. Since any human rights obligation or even the removal of the geographical limitation were conspicuously omitted from the contents of the EU-Turkey deal, much less attached to it through incentives or benefits, it is unsurprising that the limitation was not removed nor that other human rights treaties were ratified by Turkey in connection with the deal.

Ratification of Other Relevant Treaties
This section looks at ratification by the receiving states of other treaties, specifically, the two statelessness treaties and the Smuggling Protocol of the Organized Crime Convention. These provide a fuller picture of the receiving states' development of their asylum systems, as well alignment with the border control focus of the offshore processing arrangements.
Both conventions on statelessness have comparatively low ratification rates. 92 The 1954 Statelessness Convention offers socio-economic rights on par with citizens or other non-nationals, while the 1961 Statelessness Convention contains obligations such as providing children who would otherwise be stateless with nationality at birth (1954 Statelessness Convention, Articles 17-22 and 26-28;1961 Statelessness Convention, Article 1). Both Jamaica and the TCI are parties to the 1961 Statelessness Convention. Jamaica ratified the 1961 Statelessness Convention in 2013 and the TCI is a party by territorial application of the UK since 1966 (UNTC, Chapter V: Refugees and Stateless Persons n.d.). Based on the timing of Jamaica's ratification and the territorial application to the TCI, both countries can, therefore, be seen to be pursuing their own domestic agenda outside of the US Safe Havens arrangements.
The only receiving state to ratify the 1954 Statelessness Convention was Turkey on 26 March 2015. Interestingly, Turkey did not ratify the 1961 Statelessness Convention (UNTC, Chapter V: Refugees and Stateless Persons n.d.). It is likely that the ratification of the 1954 Statelessness Convention was for the purposes of implementing the greater socio-economic protections gradually extended to the Syrian refugees. What is not clear is whether the ratification was in connection with the EU-Turkey deal, or the general development of the Turkish national asylum system, which has been developing since at least 2014 when the Foreigners and International Protection Law 2013 came into effect. 93 The UNHCR submission on Turkey, during the 2020 UPR cycle, supports Turkey's ratification as being linked to its own national development. The submission states that Turkey had promised to ratify the 1961 Statelessness Convention in 2011, prior to the EU-Turkey deal (UNHCR 2019b, p. 1; see also The Grand National Assembly of Turkey 2020).
Finally, a discussion of the effects of offshore processing arrangements would be incomplete without a discussion of treaties relating to border control and anti-smuggling. This is particularly the case as all such arrangements are entered into in circumstances of increased irregular arrivals, the agreements are heavily focused on "undermin[ing] the People Smuggling industry" and "break[ing] the business model of smugglers", as well as providing financial and capacity support to address this issue. 94 The relevant treaty of focus is the Smuggling Protocol, which supplements the Organized Crime Convention together with two other Protocols. 95 The Organized Crime Convention and its Protocols entered into force between 2003 and 2005; they could, therefore, not be applied to the US Safe Havens arrangements. 96 Additionally, their early ratification by both Jamaica and Turkey shows that neither was influenced by such arrangements. 97 While the Organized Crime Convention has high ratification rates (190 state parties), the ratification of the supplementing Protocols is lower (Trafficking Protocol: 178 state parties, Smuggling Protocol: 149 state parties, and Firearms Protocol: 118 state parties) (see UNTC, Chapter XVIII: Penal Matters as at 6 July 2020). It is particularly low among the Pacific Island states (within the Asia Pacific geographic group) (Schloenhardt and Macdonald 2017, p. 17). According to Schloenhardt and Macdonald's 2017 study on the barriers to ratification of the Smuggling Protocol, the kind of states that ratify the protocol are those that have spontaneous arrivals of asylum seekers and/or migrants through irregular means; are destination states rather than states of origin or transit; and experienced an influx of irregular arrivals (Schloenhardt and Macdonald 2017, p. 23). Critics argue that some states are wary of ratifying the Smuggling Protocol because it is "too Western-centric and is driven by the desire of wealthy industrialized nations to fortify their borders, deter asylum seekers and irregular migrants, and prevent uncontrolled immigration from less developed nations" (Schloenhardt and Macdonald 2017, p. 33).
Schloenhardt and Macdonald explore other barriers to ratification, relying partially on theories of treaty ratification, specifically rationalism, constructivism and liberalism (Schloenhardt and Macdonald 2017, pp. 30-35). They argue that one of the key barriers is a lack of capacity of states to meet the requirements of the Organized Crime Convention and supplementing Smuggling Protocol, as well as a sense of the two treaties infringing on state sovereignty (Schloenhardt and Macdonald 2017, pp. 18-19, 31-32, 35). There is also the normative limitation that the Smuggling Protocol cannot be ratified without first ratifying the Organized Crime Convention (Schloenhardt and Macdonald 2017, p. 32).
Another unexpected barrier for Pacific states was argued to be the Bali Process on People Smuggling, Trafficking in Persons and Related Transnational Crime. This forum was established in 2002, and its members belong to the Asia Pacific region (Schloenhardt and Macdonald 2017, pp. 28-29). According to Schloenhardt and Macdonald, "[t] -1990s. 97 It is, however, worthy of note that in the case of Jamaica, its ratification of the Organized Crime Convention, and its three supplementing Protocols, was due to US influence in the region (see Kempadoo 2007 on the Trafficking Protocol).
view that their involvement in [the Bali Process] makes accession to and ratification of the Protocol unnecessary" (Schloenhardt and Macdonald 2017, p. 28). 98 The PNG in both the 2001 and 2012 Pacific Solution case studies follows the pattern of its region of not ratifying the Organized Crime Convention or any of its supplementing Protocols. While it is difficult to say whether PNG's issue is with the Convention or the Protocol, it nonetheless follows other patterns identified (Schloenhardt and Macdonald 2017, pp. 32-33). These include the fact that while PNG is a host to spontaneously arriving West Papuan refugees from the Indonesian province of Papua, it did not see an influx since the 1980s (Glazebrook 2014, pp. 2-4). Additionally, PNG could be reluctant to ratify the Smuggling Protocol due to its role as a founding member of the Bali Process (IOM n.d.).
The contents of the available 2012 Pacific Solutions agreements showed them to be heavily focused on border control against irregular arrivals and prevention of people smuggling. They all refer to the Bali Process rather than to the ratification of the Organized Crime Convention or its supplementing Smuggling Protocol. 99 It is not clear whether this omission is at the request of Australia or PNG, or is mutually beneficial to both countries. It is, therefore, difficult to conjecture whether the offshore processing arrangements influenced PNG's treaty ratification actions in this case.
The situation is more clear-cut for Nauru. Nauru, belonging to the Pacific Islands region, experienced few, if any, spontaneous arrivals of asylum seekers or refugees (UNHCR 2015b, p. 1). 100 Nor was it subject to an influx of such arrivals. It has no capacity to implement the Organized Crime Convention or its supplementing Smuggling Protocol. As with PNG, Nauru is a founding member of the Bali Process since 2002 (IOM n.d.). Despite all of the reasons justifying a pattern of non-ratification, Nauru signed up to the Organized Crime Convention and all three of its supplementing Protocols in November 2001, and ratified all four treaties in June 2012 (UNTC, Chapter XVIII: Penal Matters ).
Despite the silence on these treaties in the 2012 Pacific Solution agreements with Nauru, 101 the timing is highly suggestive. Looking at past behaviour discussed above, the signature of the Smuggling Protocol in November 2001 can now be added to the simultaneous signature of the other human rights treaties, which were relevant to the principle of non-refoulement requested in the original 2001 Australia-Nauru MOU (specifically the ICCPR and CAT) (Francis 2008, pp. 298-90). 102 While signing up to any of the treaties in November 2001 cannot be proven to be in connection to the 2001 Pacific Solution arrangement, the ratification of the (arguably unnecessary) Smuggling Protocol adds some weight to the argument that the arrangement between the two countries did have an effect on these ratification events.
This argument is further supported by the fact Nauru waited to ratify the Organized Crime Convention and its three supplementing Protocols, including the Smuggling Protocol, until 12 July 2012-just over a month before the signing of the 2012 MOU with Australia on 29 August 2012. A final point in support of the argument that Nauru entered into ratification events in connection with the offshore processing arrangements is Nauru's position with respect to the Refugee Convention, as expressed by UNHCR in Nauru's first UPR report. Specifically (and as per the above discussion), that the ratification of the Refugee Convention was considered by Nauru in 2006, but only in connection with the 2001 Pacific Solution arrangements, and that it was abandoned together with the winding down of the arrangements, until they began to be considered again in 2011 (UNHCR 2010b). All these points stand in favour of Nauru's signature and ratification of the Smuggling Protocol being 98 Please note, this pattern does not hold when compared to the high ratification rates of members of a similar forum, such as the Budapest Process (Schloenhardt and Macdonald 2017, pp. 29-30 linked to the 2001 and 2012 Pacific Solution arrangements, respectively. Although, as discussed above, Nauru was potentially motivated by fiscal benefits, rather than the human rights requirements of the agreements or any other considerations.

Conclusions
In looking back at the research question, it appears that the effect of refugee offshore processing arrangements on the refugee and human rights treaty ratification of a country receiving asylum seekers for processing and/or settlement, while minimal, can nonetheless be identified under certain conditions and is therefore nuanced. The general conclusion to be drawn from the above analysis of the four case studies-US Safe Havens, 2001 Pacific Solution, 2012 Pacific Solution and EU-Turkey deal-is that offshore processing arrangements do not affect the asylum systems of the receiving states as they relate to the ratification of the core human rights treaties. Certainly, where the transferring state is in charge of RSD processing, neither the refugee-related nor the core human rights treaty ratifications of the receiving state are affected by offshore processing arrangements. This is in part due to the omission of treaty ratification and human rights-based requirements in the agreements themselves, even when, as in the case of the EU-Turkey deal, previous arrangements affecting asylum seekers and refugees did contain such requirements.
However, even where human rights-based requirements are included, as in the 2012 Pacific Solution, the effect is minimal. Only PNG's removal of reservations during the 2012 Pacific Solution can truly be argued to be directly linked to the contents of the agreement-specifically to the 2013 amendment requiring recognised refugees to be settled in PNG. The earlier 2012 requirement of only processing RSD claims appears to have had no effect on PNG's intentions to remove its socio-economic reservations to the Refugee Convention. The removal, however, does not go so far as to improve the asylum system of PNG. Specifically, it does not apply to the spontaneously arriving refugees in PNG from Papua or elsewhere, only to the recognised refugees transferred by Australia as part of the arrangement. Consequently, this limitation to the removal of the reservation, limiting the application only to asylum seekers and refugees transferred to PNG by Australia, ensured that PNG did not engage in any ratification events beyond those that were in line with its normative and national interests.
While it is tempting to use Nauru as an exemplar of ratification events being linked to offshore processing arrangements, the reality is more complicated. While Nauru indeed ratified the Refugee Convention and its Protocol, as well as the Smuggling Protocol and maybe even signed human rights treaties (in 2001) in connection with its offshore processing arrangements with Australia, it did so for reasons other than human rights requirements contained in the agreements themselves. Nauru's motivation followed its national interests, which was primarily fiscal in nature; Nauru being highly dependent on the arrangements for shoring up its economy. This is supported by the fact that even if, despite its motivations, Nauru did improve its asylum system through these ratifications, it is the only country in the case studies that does not host spontaneously arriving asylum seekers and is, thus, neither benefitted, nor burdened, by its ratification actions.
This leads to the acknowledgement of the high aid dependence and inequalities between the partner states in the offshore processing arrangements (with the notable exception of Turkey). These inequalities between the transferring and receiving states can be seen to influence the nature (if not amount) of benefits negotiated by the receiving state for the refugees, host communities and the country as a whole. However, the power dynamics identified as part of the negotiation and reward systems appear to play no role in human rights obligations in the contents of the agreements or their effect on the treaty ratification patterns of the receiving states. Thus, even when engaging in arrangements regarding the offshore processing of asylum seekers with regional hegemons, on which the receiving countries are aid dependent, these countries followed established patterns of treaty ratification, especially those in line with and protective of their national interests, regional normative beliefs, and expanding participation in international forums, such as the UPR.
This conclusion confirms previous research that suggests that offshore processing arrangements are utilised by transferring states-despite the rhetoric of stopping people smuggling and of saving lives at sea, and despite being conducted under the guise of international cooperation and responsibility sharing-as a form of deterrence to stop and discourage irregular arrivals. While this article focused on the effect of offshore processing arrangements on the ratification of refugee and human rights treaties by the receiving states, further research is needed to determine the full effect of these arrangements on the development (if any) of a domestic asylum system in each of the receiving states. A developed domestic asylum and human rights system has the ability to provide better protections to asylum seekers and refugees, who are transferred as part of the offshore processing arrangements but, also, who may arrive spontaneously in the receiving state. Such research, expanding on the findings of this article, and depending on the conclusions reached, would complement existing academic literature on the topic, as well as bring forth potentially new areas of examination.
Author Contributions: The first author (M.F.) contributed the original idea, conceptualisation. The authors met throughout the writing process and mutually developed the idea and concept further. Additionally, the first author (M.F.) contributed thematic expertise on the Pacific Solutions case studies, as well as developing the structure and cohesion of the article, together with contributing significantly to the revisions and editing. The second author (N.D.J.) contributed thematic expertise on the US Safe Havens case study, as well as contributing significantly to the revisions and editing of the paper. The third author (Y.C.) contributed thematic expertise on the EU-Turkey deal case study, as well as contributing to the revisions and editing of the paper. All authors have read and agreed to the published version of the manuscript.