Navigating through depoliticisation: international stakeholders and refugee reception in Jordan and Turkey

Abstract The reception of Syrian refugees has dominated negotiations between neighbouring host states, such as Jordan and Turkey, and donors in the Global North. Despite the growing literature on external funding, how international stakeholders navigate the domestic context is understudied. This article employs the lens of depoliticisation and analyses tools employed by international stakeholders manoeuvring domestic contexts in centralised states. We argue that international stakeholders adopt a depoliticised approach that portrays them and the actors they work with as service providers while working with national governments that engage in rent-seeking. Drawing on semi-structured interviews with donors and international non-governmental organisations (INGOs) as international stakeholders in Jordan and Turkey, the contribution of the article is two-fold. First, although the ‘local’ is ever present in policy and research discussions, research and policy often conflate regional organisations, national governments and municipalities. We therefore further disentangle the power relations between the various actors categorised as ‘local’. Second, we offer a conceptual understanding of how international stakeholders navigate the dynamics of local power relations and the domestic politics in centralised states with rent-seeking behaviours.


Introduction
'Because we do not threaten Europe by pushing refugees towards Europe because we think that is [a] responsibility that should be taken care of by us in the area' , said Jordanian King Abdullah in January 2020. This reply to a journalist who claimed that 'Europe wrote a big check to Recep Tayyip Erdoğan' but 'has not done so with Jordan' reveals the omnipresence of 'rent-seeking' (Tsourapas 2019) strategies in the agendas of host states during refugee reception. Previous studies have, in fact, already widely discussed the relationship between international stakeholders 1 and host states, including the prominent examples of Turkey and Jordan, and the various ways host states use refugee displacement in the Global South to attract financial support from the Global North by learning from one another, cooperating ARTICLE HISTORY and emulating one another (Arar 2017;Freier, Micinski, and Tsourapas 2021;Tsourapas 2019). The above statement is thus but one of many examples of such 'rent-seeking' strategies.
Most of these 'rent-seeking' strategies entail financial support of national governments. However, refugees settle at the local level, primarily in urban areas and large cities, which require the navigation of domestic dynamics within host states. Thus, refugee reception targets newcomers' incorporation into a country or a city and their access to rights and services (Memisoglu and Yavcan 2022;Şahin Mencütek 2020). The significance of a broad range of stakeholders has been reflected in the 'localisation of aid' agenda, which has become an omnipresent term among policy circles since the World Humanitarian Summit in 2016 (Fiddian-Qasmiyeh 2018). The main premise of this agenda is to incorporate local actors into aid delivery to ensure effective reception measures and overcome power imbalances underlying the humanitarian system.
However, although the 'local' is ever present in policy and research discussions concerning aid, it remains unclear which actors are included in the category (Roepstorff 2020a). In the context of refugee reception, the 'local' is composed of various actors from different levels that are themselves embedded within domestic power relations as well as with international stakeholders, mainly donors and international non-governmental organisations (INGOs). Thus, what is missing in the literature is an understanding of how international stakeholders interact with local stakeholders other than the national governments of host states with centralised political power (cf. Roepstorff 2020b).
To address this research gap, we aim to analyse the strategies employed by international stakeholders to incorporate a heterogeneous set of 'local' actors as well as the political factors shaping these strategies. Hence, we synthesise findings from the 'localisation of aid' literature and research on 'rent-seeking' through the lens of depoliticisation. In our analysis of international stakeholders' aid delivery in Jordan and Turkey as two cases of refugee-receiving countries, we focus on the domestic arenas in which various actors interact, including national political decision makers, donors, municipalities, INGOs, civil society organisations (CSOs) and community-based organisations (CBOs). This allows for a more nuanced conceptualisation of the 'local' and enables detailing how international stakeholders navigate domestic politics in centralised states. Specifically, while national political decision makers succeeded in shifting power asymmetries in the wake of donors' reliance on their cooperation in and beyond refugee reception, we argue that international stakeholders reproduce domestic power imbalances by depoliticising their own activities as well as the actors opposing national political decision makers. This paper proceeds as follows. First, we discuss existing research on the 'localisation of aid' and 'rent-seeking' . Second, we present the notion of depoliticisation in reference to international stakeholders' willingness and cooperation in refugee reception. Third, we describe the methods and case selection. Fourth, we outline the policy context and challenges underlying refugee reception in Jordan and Turkey and provide empirical evidence. Fifth, we comparatively analyse the two cases. In the sixth and final section, we conclude by detailing context-specific findings, wider implications and possible pathways for future studies.

Localisation of aid, rent-seeking and the role of international stakeholders
To understand how international stakeholders navigate the domestic contexts of centralised host states, we draw on two bodies of literature. Reviewing academic work on the 'localisation of aid' agenda as well as 'rent-seeking' strategies allows us to situate our article in two distinct debates that have taken an interest in the interactions between host states and international stakeholders. Given that both debates focus on questions of power, we suggest putting them into conversation.

Localisation of aid
As just mentioned, the first strand of research of relevance for this article focuses on the 'localisation of aid' agenda (Barakat and Milton 2020;Farah 2020;Kraft and Smith 2019;Pincock, Betts, and Easton-Calabria 2020;Roepstorff 2020a). This agenda is 'often understood as increasing the role of national and local actors in international aid delivery and operations so that projects utilise, and are informed by, local actors' knowledge and capacities to reach those most in need' (Ward 2020, 2). As such, it has been the subject of a growing body of literature, analysing the role of faith-based organisations (Kraft and Smith 2019;Wagner 2018) and refugee-led organisations (Pincock, Betts, and Easton-Calabria 2020), as well as the (diverging) perspectives of staff working for INGOs (van Voorst 2019; Ward 2020).
Yet within academic and policy debates, there is a lack of clarity about what the 'local' encompasses (Roepstorff 2020a). Finding its way into policy and research debates on humanitarian responses, the 'local' can take various forms, referring to regional organisations, national governments, or municipal authorities (Fiddian-Qasmiyeh 2018). Conversely, this understanding of the 'local' often results in a conflation of all these actors into one category that is opposed to the 'international' . For example, Ward (2020) mentions this as a shortcoming in her own ethnographic analysis since she categorises all interviewees who are Jordanian nationals as local. Hence, Roepstorff (2020a) notes that the notion of the 'local' needs further clarification and critical reflection to overcome such a binary understanding of the 'local' and 'international' . Indeed, such binary understandings have been challenged recently, where some scholars have, for instance, pointed to the fact that some actors labelled 'local' are, in fact, embedded in transnational relations (Al-Abdeh and Patel 2019). Although we acknowledge the importance of overcoming such binaries, we are primarily interested in showing that the 'local' encompasses a diverse set of actors, which are also shaped by interactions with actors commonly subsumed under the label 'international' .
Thus, we take Roepstorff's (2020a, 291) work as a point of departure, where she suggests that 'instead of being conceived as a fixed spatial and temporal category, [the local] is then better understood as a highly contextual and relational concept that is a site of ongoing construction and reconstruction' . Specifically, this means that the 'localisation of aid' often takes place in domestic contexts that are highly internationalised yet are also shaped by unequal power relations between domestic actors. The salience of certain (geo)political issues, including issues that potentially exceed the immediate context of the refugee reception, also impacts how, to whom and for which activities aid is allocated. Whereas the 'localisation of aid' agenda builds on the idea that it is necessary to empower local actors to counter unequal power relations between the Global North and South (Roepstorff 2020b), it is therefore important to take into consideration the power imbalances within host states.
As a consequence of the diversity of actors involved in the 'localisation of aid' , it is necessary to understand not only how host states have shifted power imbalances vis-à-vis donors, but also how this affects, and potentially compounds, unequal power relations within domestic contexts. While peacebuilding and development scholarship has been more attentive to these dynamics (Barakat and Milton 2020;Obradovic-Wochnik 2020), only more recently have they been analysed in respect to the 'localisation of aid' agenda (Roepstorff 2020a). This article therefore further aims to advance our understanding of how the involvement of international stakeholders widens power imbalances among 'local' actors. As we further develop in our discussion of the concept of depoliticisation, interventions presented as apolitical play an important role in depoliticising the activities of 'international' and 'local' stakeholders and thereby reproduce power inequalities. First, however, we turn to research on 'rent-seeking' , for questions of power in refugee reception are likewise featured prominently in this body of literature.

Rent-seeking strategies
Another body of literature that has taken a similar, albeit different interest in the centrality of power imbalances in the interactions between host states and international stakeholders is academic work on how what Tsourapas (2019) has termed 'refugee rentier states' use the presence of refugees to gain leverage in international negotiations and attract financial support. This strategy has been particularly prominent in Jordanian refugee policies since the settlement of Iraqi refugees in the 2000s and has been further pursued during the reception of Syrian refugees (El-Abed 2014; Tsourapas 2019) but has also been used by the Turkish and Lebanese governments as leverage against the European Union (EU) in recent years (Tsourapas 2019). Importantly, however, these strategies can be found throughout the Global South (Freier, Micinski, and Tsourapas 2021).
Nonetheless, rent-seeking strategies are not uniform. For example, in his comparative analysis of rent-seeking strategies in Jordan, Lebanon and Turkey, Tsourapas (2019, 465) distinguishes between 'blackmailing' and 'back-scratching' strategies, where the Turkish government uses the former strategy by threatening the EU to open the borders and thereby allow refugees to migrate further if there is no financial help, while the latter is applied by the Jordanian and Lebanese governments, 'promising to maintain refugee populations within its borders, if compensated' . This certainly also resonates with the statement of King Abdullah at the outset of this paper. More recently, Freier, Micinski, and Tsourapas (2021) demonstrate that countries in the Global South are developing ways to further their demands in rent-seeking by learning from one another, cooperating and emulating one another. Yet regardless of the differences in their rent-seeking behaviours, these strategies of national governments in the Global South exemplify their ability to shift power imbalances and gain leverage towards donors.
Although this has resulted in increased funding for the governments of major host states, what has received less attention in the literature are the power imbalances within the domestic contexts of these states. While many international stakeholders acknowledge the importance of working with national NGOs and CSOs, the latter are embedded in domestic power dynamics and have been confronted with a 'shrinking civic space' (Roepstorff 2020b). In this article, we therefore not only understand 'rent-seeking' as a way to increase leverage through financial support but also shed light on its implications for the cooperation between international and local stakeholders. In doing so, we suggest analysing how these 'rent-seeking' strategies not only shift power relations between donors and host states, but also widen the power gap between the national government and other 'local' actors. This shifts attention to how international stakeholders work with such actors operating in the respective contexts, be they municipalities or NGOs and CSOs, and address their need to seek and generate funding.
Hence, an analysis of 'rent-seeking' strategies and their relations with 'localisation of aid' practices needs to account for domestic power dynamics and cannot rely on an understanding of the 'local' that conflates all actors that are not considered 'international' . This would omit the great diversity of 'local' perspectives, and the political conflicts shaping 'local' actors' relations with the governments and political decision makers of host states (Cunningham and Tibbett 2018). The political salience of refugee reception, and other issues of (geo)political importance, in the cooperation between host states and donors, in turn, might also consolidate the political power of the former vis-à-vis domestic critics. Put differently, while the 'localisation of aid' agenda should reduce power imbalances, it strengthens the power of refugee-rentier governments vis-à-vis other 'local' actors.

Sidelining politics
To demonstrate the implications of unequal power relations and depoliticisation, we argue that it is necessary to unpack the relations between the various stakeholders active in host states' domestic contexts, which accommodate both the 'local' and 'international' . At the same time, this focus also requires a better understanding of stakeholders' approach to this heterogeneous, and often competing, set of actors within domestic contexts. Hence, to better understand how international stakeholders circumvent politicised conflicts, we employ the notion of depoliticisation and argue that it marks a key strategy employed by international stakeholders manoeuvring domestic contexts in centralised states.
The concept of depoliticisation has been used in a broad range of scholarly work on the relations between Global North and South stakeholders (Desportes and Moyo-Nyoni 2022;Facon 2022;Ferguson 1994;İşleyen 2015;Norman and Micinski 2022;Zihnioğlu 2019). In his seminal work on development interventions in Lesotho, Ferguson (1994) notes the operations of an 'anti-politics machine' that turns political issues, such as poverty, into technical ones and thereby leaves out structural socio-economic considerations. Depoliticisation therefore includes a 'combination of tactics and instruments that aim to protect the sedimented social sphere from any challenges directed at the foundations of the social order' (İşleyen 2015, 258). Such moves can be found in interventions across geographical contexts as well as policy fields, be it in humanitarian responses to the 2004 tsunami in Sri Lanka (Kleinfeld 2007) and to the drought in Zimbabwe (Desportes and Moyo-Nyoni 2022), in the EU's support of Turkish CSOs (Zihnioğlu 2019) or in European migration and development interventions in Tunisia (Jung 2022).
Humanitarian interventions are, in fact, a particularly telling example of how we can understand depoliticisation, given that ideas of neutrality and impartiality are regarded as fundamental principles (Barnett and Weiss 2008). Apolitical images of victimhood produced by humanitarian actors therefore 'preserve social order' (Carpi 2020, 145), rendering the omission of deeper structural political issues at the heart of contemporary humanitarian responses (Pallister-Wilkins 2022). This raises the question of the rationales behind and implications of such seemingly apolitical interventions in the governance of refugee reception. Depoliticising moves facilitate cooperation in politically sensitive policy areas in and beyond the context of humanitarian interventions (Facon 2022, 7; see also Jung 2022). In doing so, these moves have material implications insofar as short-term and measurable service-based activities have been favoured over rights-based ones (İşleyen 2015;Zihnioğlu 2019). Consequently, depoliticisation is not merely a rhetorical device that enables the cooperation in politically salient areas, but also shapes how and to whom financial support is channelled.
What research has thus revealed is that whereas interventions are presented as if they are apolitical, they are in fact highly political. One important aspect is that interventions, be they humanitarian or development related, unfold in political spaces and have political consequences (Barnett and Weiss 2008, 4). Kleinfeld's (2007) analysis, for example, shows that humanitarian ideals of neutrality have been unsettled in the wake of the 2004 tsunami in Sri Lanka due to competing political rivalries between the government and the Liberation Tigers of Tamil Eelam. However, depoliticisation can also disguise broader political interests. In her recent work, Facon (2022) argues that apolitical humanitarian responses of Global North donors in Lebanon disguise European interests in externalising border policies. Similarly, our analysis of the case of Jordan and Turkey shifts attention to the ways political interests and power dynamics are depoliticised. However, while Facons's study shows how Lebanese actors actively politicise these responses for their own benefit, we take an interest in the further power shifts between national political decision makers and other actors present in host states.
Drawing on this work, we show that depoliticising strategies are applied in two ways. On the one hand, international stakeholders, either as funders or implementers, depict their activities as apolitical and thereby knowingly omit sensitive political issues. In doing so, they aim to secure the ongoing cooperation with national political decision makers in host states. This therefore illustrates the leverage rent-seeking states have gained in the wake of refugee reception. However, international stakeholders not only depoliticise their own activities but also contribute to the depoliticisation of actors other than political decision makers. Hence, on the other hand, being aware of the importance of domestic actors other than the national government and political decision makers in refugee reception, the role of international stakeholders is primarily boiled down to service providers. By adopting such a technical approach, they rely on other local and non-state actors as service providers while they, simultaneously, diffuse their political nature. In short, they become complicit in the reproduction of a 'shrinking civic space' (Roepstorff 2020b).

Methodology and case selection
In offering our empirical analysis, our article adopts a comparative case study design, focussing on Jordan and Turkey. The rationale for the case selection is threefold. First, both Jordan and Turkey are large refugee-receiving states. In 2021, Jordan had 670,000 and Turkey had 3.6 million Syrian refugees registered (UNHCR 2021). Second, in both cases, main international aid donors play a major role in refugee reception programmes and policies. Third, and finally, Jordan and Turkey are centralised and authoritarian states, making the domestic arena increasingly difficult for international and local stakeholders to operate in and carry on their activities related to refugees.
The material consists of 16 interviews with representatives of international stakeholders working in the Jordanian and Turkish context. The term 'international stakeholders' refers mainly to international donors and INGOs as the international actors operating at the local level. While we acknowledge that there might be differences in the action gap, as we focus on their navigation within the politics, we have observed that grouping them under the term 'international stakeholders' allows us to elaborate the focus of this article. The interviewees were selected through the elements of purposive sampling and snowball sampling. They all work as project coordinators within international organisations on the ground in Jordan or Turkey. We conducted seven interviews with respondents working in Turkey and nine interviews with 10 respondents working in Jordan. Each interview lasted approximately an hour. Due to the Covid-19 pandemic, we could not conduct the interviews through fieldwork in Jordan and Turkey. Instead, we had to reach out to the organisations and conduct the interviews over Zoom.
The interviews were voice recorded and transcribed by two of the co-authors. All the interviewees requested to be anonymous and not have their names or affiliations revealed. Therefore, we only provide information about who is funding the different organisations they are working for, such as the United Nations (UN), the EU, or foreign national governments. Each of the international organisations' primary aim is to offer refugee reception programmes. Since the main aim of the article is to analyse how international stakeholders navigate the domestic politics of centralised and authoritarian states in refugee reception, we focus only on these actors.
During the interviews, we asked about the development of projects related to refugee reception in the organisation they work with, how these projects are funded and coordinated, with which actors from which levels, and the content of their cooperation. Furthermore, we asked about the challenges these stakeholders face when operating in a centralised state and how they adapt to overcome these challenges. Finally, we also asked about their level of engagement with the national actors in addition to the local actors to understand which actor from what level has relatively more power when it comes to decision-making in implementing projects related to refugee reception. In addition, the empirical material consists of policy documents and reports, legal documents, and other written material, used to assist in the contextualisation of the analysis.

Depoliticising the 'local'
Employing the notion of depoliticisation to analyse how international stakeholders navigate domestic political dynamics in Turkey and Jordan, we (1) pay attention to the formal political structures governing the reception of Syrians in both countries, and (2) disentangle how donors and INGOs interact with relevant stakeholders in the respective domestic contexts. Specifically, we argue that international stakeholders' approaches vis-à-vis Jordanian and Turkish stakeholders are marked by depoliticisation in which they are primarily seen as service providers.

Navigating geopolitics: the Jordanian case
Since the onset of the Syrian war in 2011, the Jordanian state has worked with various international stakeholders on refugee reception. This includes particularly close cooperation with the EU that resulted in the signing of the Jordan Compact in 2016, which guaranteed Jordan grants and concessional loans in exchange for granting Syrians access to the formal labour market (Şahin Mencütek and Nashwan 2021). More specifically, at the level of policy development related to refugee reception in Jordan, two sets of actors stand out in particular. On the one hand, there are various national actors, most importantly the Royal Hashemite Court (RHC), the King, the Prime Minister, the Ministry of Interior (MOI), and the Ministry of Planning and International Cooperation (MOPIC). Since the arrival of Syrian refugees, they have played significant roles in formulating policies governing refugee reception (Şahin Mencütek 2020, 196). On the other hand, however, the main donors have played a central role in policymaking processes at the national level, most notably the United States (US), the EU, Germany, and the United Kingdom (UK).
Indeed, as Interviewee J4 explains, it is at this level where the policies are designed and where the 'big frame' is defined and 'big politics' play out (Interviewee J4). In particular, the RHC and the king, in close dialogue with the US, remain the key actors, as Interviewee J6 further states: 'At the end of the day, let's be very clear, the King decides. None of us, even ambassadors, [we] don't have that privileged access, except the Americans' . This level of decision-making points to the centralised structure of the Jordanian state (see also Ali 2021). It is, therefore, insufficient to understand the 'localisation of aid' as a process that shifts power from the 'international' to the 'local'; rather, they are steeped into long-standing power inequalities between and among these categories.
A critical factor within this landscape that gives national decision makers leverage and generates financial support is not only Jordan's willingness to host refugees but also its geopolitical role in the region. The privileged relationship between Jordan and the US is an outcome of Jordan's role in the region vis-à-vis Israel. This means that the US government has intervened when their state agencies have touched upon sensitive topics to appease the Jordanian political leadership: So, it's a very sensitive thing, and that's why it's hard to have even the donors on board, because this level of discussion is mainly driven by the US because they are the most important donor out here, and then I think PRM [US Bureau for Population and Migration and Refugees] and USAID [United States Agency for International Development], they tried several times, but then the state department quite regularly, I think, tell them: 'Hey guys. No, Jordan has a deal, we have a peace deal in the region. You don't want to jeopardise any of that by trying to be too pushy on things. Let them act as they want and leave them some space' . (Interviewee J4) This leverage affects the work of actors that provide services to Syrian refugees that touch upon politically sensitive topics, such as 'protection issues that could have a negative perception by the authority' (Interviewee J1). However, the fact that certain topics are more sensitive does not mean that international stakeholders cannot push for their agendas. Indeed, within the broader policy frame defined by the government and its main partners, in the implementation stage, there is some 'margin of manoeuvre' (Interviewee J1), as Interviewee J4 further explains: And then, when they draw those lines, they leave the 'How do you implement?' to all the actors. The problem is more the macro politics and more when you want to get with the details because they are not formulated quite clearly, right? So, you have to get close to those macro limitations, but you never know. When you hit them, you know, because you get a phone call from the MOI that tells you that you might leave the country very soon. So, that's how you know that you hit their upper limit, but it's not clear, right? It's not clear, so there is a grey area, this is our playground, this is where we can try to push a little bit and to get some progress around key things, and protection remains the main topic.
It is within what the interviewee called a 'playground' where policies are implemented, and the various INGOs come into play. An important stakeholder and connecting link between the donors, government, and INGOs is the Jordanian INGO Forum (JIF), which is supported by a range of international donors to represent the collective stances of active INGOs and facilitate the relations between the various actors. Given their widespread involvement in providing services to Jordanians and Syrians, INGOs are therefore central actors. Simultaneously, they also take on an influential role within the localisation of aid agenda, for they act as gatekeepers for access to funding for local actors and have the capacity to navigate administrative requirements.
Consequently, while there is a general agreement among international actors on the advantages of this agenda, in practice, the majority of the funding still goes to international actors: You hear a lot of donors talking about why it [localisation] is important and their commitment and so on and so forth. There are two problems for me in the localisation conversation. Firstly, I agree in principle on it, and I think for me it's perhaps also part of a larger analysis of our sector from a very post-colonial perspective, you know. The notion of privilege, the notion of access, is very much dependent on nationality in the whole overseas development assistance sector, and I don't see any of our donors changing that any day soon. This points to an important limitation within the 'localisation of aid' discourse. Most donors only directly fund partners that are also based in the respective donors' countries. This, interviewees have highlighted, is related to the lack of capacity of Jordanian NGOs to adhere to reporting and compliance standards. In a report by one of the most prominent Jordanian CSOs, Arab Renaissance for Democracy and Development (ARDD), the high compliance standards are also mentioned as one of the main obstacles for local NGOs (ARDD 2020). While local CSOs are therefore praised for their specific contextual knowledge, they receive less funding due to the strict technical and measurable criteria required by international stakeholders. In addition, while international stakeholders highlight the benefits of the localisation of aid agenda, local actors are primarily considered service providers: In principle, we are positive about the whole 'localisation of aid' agenda. We think that in the end, of course, it is the organisations on the ground, the local organisations, who know the situation, who know the people, who speak the language, who are connected on the ground and in this sense, of course, are in the best position to implement the relief measures on the ground. Therefore, we have always promoted that our partners connect [with local partners] and that the actual project implementation is usually done by the local partners. (Interviewee J9) Given their in-depth on-the-ground knowledge, local actors are regarded as important actors for providing services to refugees. Yet they are primarily depicted as apolitical actors. This is due to the fact that international stakeholders who have tried working with local stakeholders are constrained by the centralised structure of the Jordanian state. Explaining that many donors are afraid to touch upon sensitive topics during discussions with the government, Interviewee J4 states: And it's stupid, because actually if you engage with this government and you take a bit of time to show your respect and the fact that: 'No, you're not trying to make a revolution, you don't want to have an Arab Spring in Jordan, that's not the point. That's not what we are trying to do, we are not activists. We are listing some of the challenges that we think could be even a risk for Jordan, and we would like to address it collectively' . And when you speak like that, as long as you don't publish in the media and make a big deal of it, but you really engage with the government, you can talk about even extremely sensitive things, and they will never shut you down.
This points to an important challenge within the 'localisation of aid' discourse in Jordan. While actors acknowledge the importance of CSOs and CBOs in localising aid, the government is afraid of the potential politicisation of these groups. Indeed, as one interviewee remarks, there is 'a shrinking civic space' (Interviewee J7) despite the importance of these actors. However, in asserting that capacity-building activities targeting local NGOs do not aim to politicise them, actors have pushed for the inclusion of NGOs in negotiations. This eventually resulted in the establishment of the Jordanian National NGOs Forum (JONAF) in 2016. At the same time, INGOs often work with a mix of royal NGOs and CBOs and CSOs. While royal NGOs are initiated and led by royal family members and thereby intimately linked with the national leadership, the latter are more independent members of civil society and thereby closely observed by the state.

Manoeuvring party politics: the Turkish case
In the case of Turkey, major leverage of the Turkish government stems from its shared border with the EU that resulted in the signing of the 'EU-Turkey Financial Deal' in 2016. This deal introduced the agreement that 'for every Syrian being returned to Turkey from the Greek Islands, another Syrian will be resettled to the EU' (European Council 2016). In return, an allocation of six billion Euros was planned to be granted to the government to cover refugees' needs if Turkey fulfilled the stated conditions. Although Turkey failed to fulfil these conditions, in March 2018, the EU delivered three billion Euros to Turkey (Pacaci Elitok 2019). In addition, many projects are being funded by the EU and have been allocated to INGOs that are operating in Turkey. For example, as of the 2018-2019 financial period, 4.2 billion Euros has been guaranteed, the majority of which will fund 85 projects. An overwhelming part of these funds is distributed through relevant ministries and these ministries' local representatives.
In 2014, Turkey ratified the Law on Foreigners and International Protection, which introduced the Directorate of General Migration Management (DGMM) under the MOI as the sole institution responsible for foreigners in Turkey and 'Temporary Protection Status' for Syrians. For instance, within DGMM, the 'Migration Policies Board' has been introduced and operates under MOI (DGMM 2019). In July 2018, this section was reformed under the name 'Migration Board' and announced that it had introduced the draft 'Integration Strategy of Turkey and National Action Plan 2018-2023' . This Action Plan delineates all the relevant public offices and municipalities and delegates roles to them in coordination with International Organisations, CSOs and migrants. Proving this point, the interviewed international stakeholders named national-level ministries, DGMM, and other national-level entities as the primary actors they work with. This is related to the centralised structure of Turkey, particularly regarding policies related to the reception of refugees. Despite the significance of the 'localisation of aid' agenda, only a small fraction of 'local' actors is included.
At the same time, as the political parties in charge of municipalities do not always match the ones in power at the national level, international stakeholders must navigate highly politicised contexts, as Interviewee T7 summarises: I want to say something, as my organisation or so, international organisation, deals with ministries, deals with the government and at a local level, the political parties in charge are not always the same as the government. This is an issue. This is an issue, in terms of signing documents, moving forward, we are facing issues here some time, yes, to put in place programmes and getting documents, because there are barriers, there are obstacles. But this is like the big political swings.
[…] Everything is very much politicised here.
[…] I think the Turkish government also is using these foreign refugee migrants as a discussion and bargaining with Europe.
INGOs' strategies of interacting with local stakeholders, particularly municipalities and local NGOs, has been 'very dependent on political views in Turkey' (Interviewee T3). This dependency on political views by the international stakeholders and the distribution of EU funds through ministries and their local counterparts in Turkey is the second element of depoliticising the 'local' as it requires navigating domestic politics. INGOs acknowledged the increased importance of municipalities' resources and the role of INGOs or aid agencies that partner with municipalities in service provision (Baban, Ilcan, and Rygiel 2017;Korkut 2016). According to one interviewee, 'municipalities in Turkey have been at the forefront of the refugee response since the beginning' (Interviewee T6). Similarly, Interviewee T7 highlights: Municipalities are definitely the key because if they are willing to support refugees if we are putting to place a strategic plan so on and so forth, in favour of developing business with Syrians, developing provision of services, working on social cohesions and so on and so forth, municipalities are definitely very key, very key.
Notwithstanding being identified as key actors, Turkish actors other than the central government are faced with financial constraints, as Interviewee T1 emphasises: So, more capacity development, more resources should be provided to local institutions, public institutions, municipalities, and their capacity to deliver services should be developed. We see a lot of money coming to projects, to INGOs, and to World Bank, but still, we see some sort of resistance from European donors especially, that they draw the line between refugee response and development support. And they say some areas receive development support, and Turkey is not in a position to receive them as a developing or developed country. But more funding needs to be mobilised so that local institutions, public institutions, municipalities, can provide the support in a more sustainable manner. Because of project-based support, you can only do so much.
Municipalities have two sources within the financial framework: allocation by the central government, consisting of a small part of the municipality budget, and locally generated revenues. Hence, there is a high reliance on the revenues distributed by the central government according to predetermined criteria, such as total population and a city development index. However, municipalities can also apply for domestic and foreign loans that consist of the latter. Many municipalities prefer the EU-funded projects, which are also subject to the treasury's approval (Callet-Ravat 2016). The central government thus has a direct influence on which actors and activities funds are channelled to. INGOs also indicate the strong presence of the central state as an obstacle when interacting and engaging with local actors. For instance, Interviewee T4 explains: What I experience is that a lot of decisions are taken in municipalities, that there is also a reluctance to take decisions because no one wants to make a wrong decision, that decision-making takes longer, that it's not always so clear who is taking the final decision and why this decision is taken. For me, this is all a bit still opaque.
Thus, a highly centralised state structure obliges the INGOs and donors to familiarise themselves with domestic politics in which they position themselves as service providers. Although there are also acknowledged differences within the INGOs and donors depending on their capacity and organisational structures that influence their relationship with the centralised state in Turkey and vice versa, domestic politics still comes across as one of the most influential dynamics (Kayali 2022;Nimer 2020). Hence, it is fair to argue that familiarising themselves with domestic politics demonstrates how INGOs and donors navigate these dynamics due to the highly centralised state structure of refugee policies. Interviewee T7 highlights: I mean, okay, people still need basic needs, so we still need to keep some humanitarian aspect and protection aspect, it's very important for many, but if you really want to build upon resilience, social economy conclusion and so on, we cannot rely on one-year, even two-year project. States should be much more comfortable in providing much long-term funding. We have to readjust. We need to monitor all of that, and it's not like a blank check for five years.
[…] We cannot ignore the fact that the way we are dealing with refugees here is also connected to the migration policy.
These statements indicate that INGOs are well aware of the domestic politics, and as INGOs, they have to navigate these politics in order to continue to receive funding in which they see themselves as depoliticised service providers as there is a dependency on the central governments' decisions and approvals. These also link back to the part where the interviewees discuss working with municipalities and NGOs as being dependent on the political context. Hence, depending on the project, what INGOs offer is 'guidance provision' (Interviewee T6), 'consultancy' (Interviewee T2), 'technical cooperation' (Interviewee T7), and 'mediation' (Interviewee T5). Therefore, positioning themselves as providing consultancy or technical support, along with other terms they use, INGOs distance themselves from the party politics that might lead them to get involved in local political dynamics that they prefer to avoid. As the government of Turkey does not directly recognise refugees but regards them as 'people under temporary protection' (DGMM 2019), this distancing allows INGOs to continue to work on the projects they have the funding for while delivering services without taking on any role that might be considered advocacy related to refugee rights.

Discussion
Given the politically sensitive nature of the issue of refugee reception and the centralised political system, international stakeholders have to navigate highly politicised contexts.
Analysing the strategies of international stakeholders that enable working with a heterogeneous set of actors in centralised states and the rationales behind these strategies, we argue that international stakeholders navigate domestic contexts by depoliticising themselves as well as the local actors they work with. Our analysis shows that international stakeholders have acknowledged the presence of a strong central state in Jordan and Turkey and pointed to highly internationalised domestic contexts where the national governments and political decision makers have gained notable leverage vis-à-vis donors. While the localisation of aid agenda ought to address power imbalances, it is primarily political decision makers that strengthen their power.
This leverage stems from broader geopolitical concerns that render it necessary to secure the continuous cooperation between international stakeholders and Jordan and Turkey, respectively. Although the role of Jordan in hosting a large number of Syrians has also allowed the government to gain leverage, as the Jordan Compact exemplifies, there is one significant difference compared to the case of Turkey where, as widely discussed, the reception of Syrians has allowed the central government to negotiate the EU-Turkey Financial Deal. Not only does Jordan's role as a major host state matter, but due to the long-standing cooperation between Jordan and the US, Jordan's geopolitical role vis-à-vis Israel is another factor that has given the government bargaining power. Coming back to King Abdullah's statement at the outset of this article, just because the Turkish government has indeed employed a 'blackmailing strategy' (Tsourapas 2019), this does not mean that Jordan has been denied financial support. Instead, both the fact that Jordan has hosted a large number of Syrians and its long-standing partnership with the US have contributed to financial support and the implementation of refugee reception policies that leave domestic power imbalances untouched.
As becomes evident, in both cases, national political decision makers have gained remarkable leverage due to their geopolitical roles. This, in turn, puts a limit on what is considered feasible when carrying out activities. As both Jordan and Turkey use the presence of refugees within their borders as leverage towards donors, international stakeholders are required to navigate domestic power imbalances while being aware of the importance of actors other than the central government. In order to ensure the cooperation of both states, international stakeholders navigate through depoliticisation. This depoliticisation manifests itself in various ways.
Although international stakeholders receive their funding from international actors, they engage with multiple local political actors. This engagement leads international stakeholders to make sense of domestic political dynamics and how different actors operate in a centralised political context, as stated by interviewees in the preceding pages. Therefore, when implementing their refugee reception programmes, international stakeholders adopt the roles of service providers in which they solely deliver what international stakeholders have agreed upon rather than taking an initiative to further offer policy suggestions. The acknowledgment of the highly politicised contexts thus leads international stakeholders to situate themselves as service providers. By doing so they distance themselves from politicised debates within domestic contexts and frame their interventions as merely technical and apolitical.
In Turkey, international stakeholders particularly have to navigate party politics, given that municipalities have an influential role in the reception of Syrians and not all municipalities belong to the same political party as the central government. For instance, Betts, Memisoglu, and Ali (2021) demonstrate that practices in the municipalities differ depending on their political party affiliation, which leads them to adopt more restrictive (municipalities belong to the main opposition party) or more proactive (municipalities belong to the central government's party) refugee policies. As this influences the refugee reception and service provision to refugees, INGOs adopt the role of apolitical service providers to secure the approval of the central government to continue operating in Turkey in the wake of contentious party politics. Therefore, the short-term nature of the funding that municipalities can apply for means that INGOs are primarily seen as service providers that cover the most urgent humanitarian needs rather than actors implementing long-term policies. In Jordan, on the other hand, depoliticising strategies have been employed by INGOs, and by extension donors, when working with NGOs, CSOs and CBOs, given that these actors have faced restrictive government policies. INGOs, as in the case of Turkey, require government approval to implement and channel funding to Jordanian actors and thereby increase the government's control over the ways funds are channelled.
International stakeholders have emphasised that there is some leeway when working with actors other than the national governments to carry out projects. Indeed, the importance of these actors was repeatedly stressed, for example by Interviewee J9, who acknowledges 'that many local NGOs, of course, also demand [direct funding], who rightly says: We have become very strong partners and we actually do not need this intermediate step of the INGOs anymore necessarily' . Nonetheless, it remains unclear to what extent this will change in the future. This is not only due to the high and varying compliance and reporting standards required by donors, but also to a continuous depoliticised understanding of these actors as policy implementers rather than developers. Consequently, international stakeholders risk undermining broader political struggles of these actors and reproducing power imbalances between the latter and national political decision makers.

Conclusion
In this article, we have analysed the ways international stakeholders navigate domestic political arenas in the context of centralised states in Jordan and Turkey. Their major role in hosting Syrians and their geopolitical significance further contribute to the manifestation of their centralised power. International stakeholders, indeed, consider these conditions and acknowledge the centralised political power in both countries. However, to ensure cooperation with actors that are not associated with the central state but are needed for the reception of refugees, international stakeholders depoliticise their own activities as well as those of NGOs, CSOs and municipalities. Given that the role of these actors is boiled down to that of a service provider, these interventions thus avoid challenging broader social and political inequalities.
Hence, considering the growing literature on 'rent-seeking' behaviour of host states and on implications of the 'localisation of aid' in the Global South, this article contributes to our understanding of the role of international stakeholders in refugee reception in two ways. First, the findings contribute to our understanding of the ways international stakeholders carry out the 'localisation of aid' agenda. This attentiveness to how these stakeholders situate themselves as well as 'local' actors as service providers in order to navigate highly centralised state contexts thus demonstrates how these efforts reproduce power imbalances. Second, and relatedly, the article contributes to the growing literature on 'refugee rentier states' , for it unpacks the impact of 'rent-seeking' strategies on domestic politics and power relations. Such strategies therefore not only generate financial support for state actors, but also constrain the room for manoeuvre of 'local' actors.
Consequently, these findings have important implications for the understanding of the 'localisation of aid' agenda, and the impact of 'rent-seeking' strategies. Importantly, while the 'localisation of aid' agenda propagates ideas of local ownership and marks an attempt to address the power relations between the Global North and South, international stakeholders risk reproducing power imbalances within these domestic contexts. Consequently, this heterogeneity of the 'local' calls for a research agenda that pays much more attention to the power imbalances within the domestic arena of host states. This includes, inter alia, a better understanding of the ways actors other than the state respond to this depoliticisation and navigate increasing restrictiveness.
In thinking beyond the case of Jordan and Turkey, there is no reason to believe that our findings are restricted to the two cases investigated in this study. We believe that the depoliticisation of the local by international stakeholders is likely to exist in other, similar political environments, such as Lebanon and Egypt. Even though the distinct conditions matter and we in this study focus on Syrian refugee reception in neighbouring Middle Eastern countries, the results are most likely also applicable to forced migration in the Global South more broadly. Regardless of context, the involvement of international stakeholders might exacerbate existing power dynamics between national political decision makers and other actors. This being said, other relevant cases to explore are settings where the recipient governments are less unified, and where tribal politics -as another 'local' factor -plays a greater role, such as in the case of Dadaab refugee camp in Kenya, on the border with Somalia.
Isabell Schierenbeck, @ischierenbeck, is Professor in Political Science at the School of Global Studies (SGS), University of Gothenburg. Her main research interests are global public policy and administration, within the fields of international development cooperation and migration, as well as Middle Eastern politics. She has a long-standing interest in research ethics and safety and is the co-founder of the Centre on Global Migration (CGM), University of Gothenburg.
Andrea Spehar is Associate Professor in Political Science and Director of the Centre on Global Migration (CGM), University of Gothenburg, Sweden. Her research interest broadly lies in comparative public policy, particularly regarding gender policy and immigrant integration policy in a European context. She is Principal Investigator for the research project 'Refugee Migration and Cities: Social Institutions, Political Governance and Integration in Jordan, Turkey and Sweden' (SIPGI) (2019-2025).