Disaggregating Jordan’s Syrian refugee response: The ‘Many Hands’ of the Jordanian state

ABSTRACT This is a disaggregated study of different factors which shaped Jordan’s Syrian refugee response. It considers the response’s internal workings and how hosting a large displaced population from the Mediterranean state of Syria is distributed across different public institutions with the involvement of international actors. The argument is that an agenda intent on securing the status quo influences the response, but that it is not always coherently implemented by the many hands of the Jordanian state. The main aims are to resist the permanence of Syrians so as not to undermine the demographic balance that favours Trans-Jordanians; to secure income for hosting Syrians; and to limit the possibilities for formal Syrian economic competition with Jordanians. At the same time, and related to these aims, there are initiatives to render Syrians legible, and these legibility initiatives serve different goals depending on which hand of the state is enacting them. The paper is based on qualitative fieldwork conducted from December 2016 to March 2017 in Jordan. It features interviews with Jordanian officials from national and municipal institutions, and with staff from international organizations.


Introduction
Azraq is a highly securitized environment in the desert, and Syrians' actions are heavily managed by SRAD and humanitarians (Gatter, 2021;Hoffman, 2017). The Colonel added: 'If the Syrians were in the cities, how would people know how much water they were using? . . . We don't know them, their background, and any state would want to control what is happening on its territory'. Elsewhere in the Jordanian state, the atmosphere differed. The Sharia Courts, part of the government of Jordan with legal powers, offer low-cost assistance to Muslim families in Jordan, including Syrians. A Directorate Investigator there explained: . . . we don't want to harm families with high expenses . . . we meet the basic needs of the refugees [in documenting]: marriages, births, deaths, divorces. But it would cause more problems in the long term if the issues were not addressed in this way by the Sharia Court.
Concerns for Syrian families combined with those about the risk of creating generations of undocumented, potentially stateless, Syrians in the country. Near Syria, I met with municipal officials with cross-border kinship ties with the Syrian arrivals in their municipalities, and who spoke of solidarities of culture, faith and tribe, that motivated their efforts to support new arrivals. The Umm Al-Jimal Al-Jideeda municipality organized volunteers to conduct a count and needs assessment of Syrian arrivals and secured central government assistance with that information. As I will show, the 'many hands' of the Jordanian state approached Syrians in different ways, and within a hierarchy of state institutions. While there was a broad national agenda intent on securing the political status quo through interrelated security and legibility initiatives, other significant motivations were evident.
The violent repression of the Syrian uprising, and the ensuing proxy wars and destruction, produced the Mediterranean region's largest displacements. Since 2011, Syrians have sought refuge in neighbouring Lebanon and Turkey (Betts et al., 2020;Tsourapas, 2019), in Greece, Italy, Spain (Saatçioğlu, 2020) and in Egypt, Algeria, and Morocco (Buehler et al., 2020;Norman, 2020). But almost 12 per cent of all Syrian refugees reside in Jordan, increasing its population by at least 7 per cent in just a few years (United Nations High Commissioner for Refugees [UNHCR], 2021;World Bank Public Data, 2020), and making Jordan a relevant case to explore as part of a crisis that has affected the entire Mediterranean region. This article asks what has shaped Jordan's policies towards Syrian refugees, highlighting the internal workings of the Syrian refugee response. It shows that the work of hosting a large, displaced population is distributed across different public institutions. It finds that legibility initiatives combine with an agenda to secure the status quo which is perceived to underpin Hashemite rule. And it notes that international organizations are involved in these processes. Existing literature on Jordan as a host of Syrian refugees focuses on foreign policy (Tsourapas, 2019), the Jordan Compact and Syrians in the labour market (Lenner & Turner, 2018), the effects of refugee integration on migrant labour in Jordan (Hartnett, 2019), encampment policies (Turner, 2015) and the securitization of encamped Syrians by state and humanitarian actors (Gatter, 2021;Hoffman, 2017). Recent work by Lenner (2020) and Lenner and Turner (2018) offers disaggregation at national and international levels, considering institutional interactions and settlements between international organizations, donor states, and various Jordanian authorities. Scholars of the responses in Lebanon and Turkey have undertaken disaggregation that includes the subnational level (Betts et al., 2020;Mourad, 2017), but scholarship that includes subnational level disaggregation of Jordan's Syrian refugee response (Betts et al., 2017) remains rare.
The paper disaggregates knowledge of the response by considering the roles of national and municipal state institutions. At the national-level it examines the SRAD which is part of the Ministry of Interior (MOI); the Ministry of Planning and International Cooperation (MOPIC) -and the Ministry of Labour (MOL). Additionally, it includes the role of the government's Sharia Courts which are hitherto absent from scholarship on Jordan's Syrian refugee response. Local-level disaggregation occurs via municipal officials' perspectives from Um Al-Jimal Al-Jideeda and Al-Za'tari & Al-Munshiya municipalities, both of which experienced major population increases after Syrians' arrival, and had to contend with Za'tari camp's impacts on their land. I use Morgan and Orloff's (2017) metaphor of the 'many hands of the state' to guide the paper's disaggregation of Jordan's Syrian refugee response. Disaggregating states into their national and sub-national level institutions offers a granular understanding of a state's policies that is absent when analysing states as unitary actors (Hechter, 2000;Pacewicz, 2020). The state is heterogenous and sometimes incoherent (Janoski et al., 2020, p. 6). State action can often be contradictory and unsuccessful (Morgan & Orloff, 2017, p. 18). The state has 'many functions, organizations, and purposes while complicating the initially sharp boundaries drawn between public and private, state and society.' (Morgan & Orloff, 2017, pp. 7-8). States can shape social relations, defining, classifying, and measuring the world around them in ways that enable officials some control over it (Morgan & Orloff, 2017, p. 11). Legibility -rendering legible the incomprehensible and unpredictable elements of everyday life -has been at the core of the modern state (Hussin, 2017;Scott, 1998, p. 183). But, says Hussin, rendering legibility has not only been a top-down endeavour, but also a project undertaken from below which can 'confer benefits for those positioned to receive them' (Hussin, 2017, p. 349). The state is not a thing hovering above society: ideological and cultural factors shape the lines between state and nonstate, and citizens' perceptions of them (Mitchell, 1991;Morgan & Orloff, 2017, p. 10). Disaggregating Jordan's refugee response into the different national and sub-national institutions active in its design and implementation helps to avoid reification and simplification. It reveals that while a security agenda -understood holistically in terms of objectives and practices (Kaldor & Selchow, 2015) -guided Jordan's national response, there are aspects of this agenda which have not been coherent or well-coordinated; and occasionally contradictory. Legibility initiatives have featured in an agenda that aims to secure specific elements of the status quo that Jordanian ruling elites believe are essential to securing Hashemite rule. However, different state institutions pursued legibility initiatives to serve different goals.
A key argument is that the Government of Jordan's (GOJ) response is shaped by a security agenda -but not one understood narrowly in relation to violence and armed conflict. In discussing security, I refer to different objectives and practices (Kaldor & Selchow, 2015). Security objectives relate to the referent object that is being secured. Security as a practice consists of the interplay of different security apparatus. This includes the military, police, and intelligence institutions and their material equipment -and their strategies and tactics which are expressions of political authority (Kaldor & Selchow, 2015). At the national level, the objective is to secure the Hashemite monarchy's rule in a number of ways. One is by securing external income from international donors. A number of national legibility and visibility initiatives are geared to this end by different ministries, including the Interior and Labour ministries, to produce an agreed-upon 'count' of Syrians (Lenner, 2020), both in highly visible refugee camps, and in less visible urban spaces, with an additional layer of legibility through issuing work permits in the labour market. Municipalities conducted their own legibility initiatives, often preceding national counts, to secure funding from central authorities and international organizations to support strained municipal budgets.
Another security objective is to prevent the permanence of the Syrian presence which, in the GoJ's eyes, is seen as a potential destabilizer of Jordan's demographic balance and therefore of Hashemite rule. This relates to the divide in Jordan between 'Trans-Jordanian' citizens, and its Palestinians, many of whom are 'refugee-citizens'. Trans-Jordanians receive preferential treatment in public sector jobs, in the army, police, and with political positions more generally because their loyalty to the Hashemite monarchy is assumed to be certain. While many Palestinians have citizenship and represent an economic success story, discrimination against them is common, particularly towards those who do not have citizenship -such as Gazans -and thus are effectively stateless, unable to own property, or open bank accounts (Minority Rights Group International [MRGI], 2020). As a politically active native-Arabic-speaking population, Syrians represent a potential challenge to Trans-Jordanians. Syrians are seen as potential vectors of anti-authoritarian politics; as a security risk by the SRAD and the MOI (Hoffman, 2017;Lenner, 2020); and potentially as an economically competitive population that could further 'crowd-out' Trans-Jordanians (Lenner, 2020) in an economy with chronically high unemployment. The success of Palestinians, an Arab refugee cohort, some of whom became permanent 'refugee-citizens' and key to state-building in the nascent Jordan, is ignored (Lenner, 2020). What matters more is the 'very real fear' among Trans-Jordanians of creating even 'a semi-permanent residence for another large Arab cohort' (Lenner, 2020, p. 291;Stevens, 2013, p. 19). Before the 2016 Compact, the GoJ used an approach it previously applied to Iraqis of creating 'incentives to transit': letting them in, but depriving them of formal status and livelihoods options thereby 'encouraging' them to move on (Chatelard, 2002). From 2016, Syrians' formal participation in the labour market remains restricted by GoJ to sectors dominated by Egyptian labour (Hartnett, 2019) in order to secure revenues within the terms of the Jordan Compact.
With the aims of resisting permanence and securing international funds, senior officials from GoJ, the King included, emphasize the burden of hosting Syrians in the Jordanian media, and at international donor conferences, without acknowledging the beneficial and generative aspects of Syrians' presence (Abdullah, 2019;Alrai Newspaper, 2016). This narrative was reinforced during interviews with officials focused on the pressures on infrastructure and services. A further aim of the burden narrative has been to divert blame for Jordan's economic and infrastructural ills from its government to Syrians to limit citizen unrest (Baylouny, 2020). However, that strategy backfired because Jordan's citizens understood that Syrians exacerbated pre-existing problems and instead blamed GoJ for not addressing them (Baylouny, 2020).
In the following sections, I explain my methods and then the different roles of state institutions in the refugee response, starting with a discussion of the hierarchy of decision makers in Jordan and a discussion of national ministries. Finally, I will share perspectives from institutions seldom considered in scholarship of Jordan's refugee response: Jordan's Sharia Courts, and the municipalities on whose lands Za'tari refugee camp lies -Umm Al-Jimal Al-Jideeda and Al-Za'tari & Al-Munshiya.

Methods
The research is qualitative and inductive, drawing upon field-notes and interviews with officials from the aforementioned institutions, and staff from international organizations, conducted in Jordan from January 2017 to March 2017. I sought a disaggregated understanding of Jordan's refugee response by including national and sub-national institutions, and the international organizations which influenced them. I interviewed officials from ministries and international organizations known for their involvement in the Syrian response. I identified relevant ministries through a combination of desk research and making contacts in the UK. I met helpful individuals at a London School of Economics Middle East Centre symposium, in June 2016, where representatives from host governments, international organizations, and NGOs working on the Syrian refugee response discussed its long-term challenges. I was also fortunate that there were participants from Jordan, working on the Syrian response, at the summer school of the Refugee Studies Centre, University of Oxford, for whom I conducted the research. Helpful interactions in the UK led to meetings with GoJ officials who I also asked to connect me with others working on the response; at times they volunteered this information. In Jordan, I was fortunate also to attend events at the International Labour Organization, the UNCHR, and Institut française du Proche-Orient which enabled further instructive interactions. Additionally, I drew upon the academic network established through my research, ongoing since 2008, into the dynamics of conflict and displacement in Iraq, Syria, and neighbouring countries, to connect with knowledgeable individuals from humanitarian, international I/NGO, civil society, and academic organizations, and with independent researchers. The majority of interviews with government officials were conducted in Arabic, with a handful in English, while the opposite was the case with international organizations. In Jordan I emailed and telephoned ministries and municipalities using information available on their websites. In most cases, this did not lead to an interview taking place. It was where I was able to use a personal email address or phone number acquired through a meeting or interview with someone known to them that an interview occurred. Fieldwork was limited to three months for this one-year long research project which also included fieldwork in Lebanon. Therefore, it was not possible to meet with officials multiple times, nor to seek meetings with multiple officials from each institution. I chose to capture perspectives from a wide range of institutions over delving deeply into a specific ministry's workings. I sensed there was a specific person that researchers were guided to in ministries. Tables listing participants are available at the end of the document and in the online appendix.

Jordan's refugees
Jordan is a major refugee-receiving country. Over 2.1 million persons of Palestinian descent arrived and settled in different stages associated with the Arab-Israeli conflict, and with varied levels of rights to residency, documentation, and services (De Bel-Air, 2016). There are 662,000 Syrians registered with UNHCR today, peaking at 673,000 in November 2018 (UNHCR, 2021). Jordan has not signed the 1951 Refugee Convention but Syrians with valid passports enjoyed an open-door policy (although without employment rights) until restrictions began in 2013 (De Bel-Air, 2016). Before then, Syrians entered Jordan through the framework of Jordan's Law of Residency and Foreigners' Affairs and were recognized as prima facie refugees through a Jordan-UNHCR Memorandum of Understanding signed in 1988 (De Bel-Air, 2016). Syrians were crossing into Jordan at a dramatically increasing rate, and mostly through informal border crossings where the Jordanian military, the Ministry of Interior, and UNHCR were registering entries. From June 2013, informal border crossings began closing, culminating in full closure in June 2016 after terrorists attacked Jordanian soldiers. By that time, UNHCR had registered over 650,000 Syrians. They are dealt with in different ways by Jordan's national and sub-national state institutions, with the involvement of different international humanitarian and developmental organizations, as well as with the involvement of donor governments and their agencies.

The 'many hands' of the Jordanian state
In Jordan it is not only security institutions like the SRAD police that deal with Syrian refugees. During interviews I learned that a hierarchy of different state institutions and security apparatus handle the refugee response. The King sits atop this hierarchy and appoints the heads of the country's security institutions, ministries, and the Prime Minister himself, who is also Minister of Defence. A pro-palace elite circulates between ministries (Valbjørn, 2013). A senior international agency official, working in Jordan since Syrians began arriving in 2011, noted that in his experience, a small group makes the major decisions regarding the Syrian refugee response. The King, the Chief of Staff, the PM, and often the Interior Minister, deal with security issues, and border openings and closures. The MOI, the General Intelligence Directorate -Jordan's secret police -(GID) -and the Royal Court deal with deportations. The King, PM, and MOPIC's minister who (at the time of fieldwork) was Imad Fakhoury, dealt with donors and international assistance. The ministries of Labour, Education, and of Local Administration (MOLA) (overseeing municipalities) were not, he said, very influential and were typically reined in by MOPIC or the PM. MOPIC decides how much external aid goes to MOLA's budget for municipalities. The senior humanitarian I interviewed called these uninfluential ministries 'line' ministries. I assumed this was an informal term for service ministries, which -as relayed to me by Jordanian officials I interviewed -were less powerful than sovereign ministries. A Ministry of Labour official explained that it was a service ministry that did not transgress the authority of sovereign ministries such as the MOI.
Consequently, he was unwilling to discuss specific MOI policies. The sovereign ministries are those of the Interior, Defence, Finance, and Foreign Affairs (Aburuman, 2008), a categorization present in other Arab countries including Iraq and Lebanon. Those ministers will meet with the King during national emergencies. Their significance in the refugee response is unsurprising, considering their financial, diplomatic and security powers. But what is also interesting is MOPIC's prominence, as a service ministry. Imad Fakhoury (minister from March 2015 -June 2018) was former Chief of Staff to King Abdullah II for Planning & Programming at the Royal Hashemite Court, a Governor of Jordan at the World Bank Group, and additional high-profile international appointments on his CV. His role in the response, particularly in creating the 2016 Jordan Compact, was significant. MOPIC's prominence may be explained by its proactive minister. Additionally, it may be because, as an official explained: 'MOPIC does not take money from donors, it is there to regulate, to prevent misspending and abuse by donors, lenders, and implementers.' With external aid being so significant in the refugee response, a regulator of that funding is understandably important.
The many hands of the Jordanian state did not always work coherently in the Syrian response. Officials I interviewed all agreed that the best outcome for Jordan and for Syrians was for Syria to 'stabilize' in a way that would allow Syrians' safe return. News of Jordan establishing a Military Operations Centre (MOC) in 2013 to support and coordinate rebel groups in southern Syria, with US and Saudi involvement, suggests that for some parts of the Jordanian state, stabilizing Syria was not a priority. MOC funded and armed different rebel groups, with US and Saudi participation, but never enough to defeat Assad's forces and allies (Maayeh & Sands, 2014). MOC angered Syrian rebels with its poor-quality support; this allowed extremist factions, like the Al-Qaida-affiliated Al-Nusra Front, to defeat them (Hubbard, 2014;Maayeh & Sands, 2014). As US interest in Syria's uprising declined, MOC worked with Russia, weakening the opposition to Assad in ways that favoured the rise of extremist groups that Jordan had earlier worried about empowering (Sadakai, 2016). Such was GoJ's fear of Al-Nusra becoming powerful near Jordan's territory, that once it seized the Nasib border crossing in April 2015, GoJ closed it. While denying Al-Nusra customs revenue from the 300 trucks that used Nasib daily, it cost Jordan between 2-3.5 USDbn in annual revenue (Turkmani et al., 2015, p. 18). While it can also be seen as an effort to establish influence over conflict dynamics in southern Syria, there is a contradictory element to prolonging the fighting which caused the Syrian refugee crisis.

Securitization, legibility and visibility
The ministry delegated with the largest Syrian legibility initiative was the MOI and its SRAD. The MOI oversees policing and security in Jordan, and the SRAD is a branch of the police (Directorate of Public Security) overseeing security in the camps. A cabinet decision in April 2014 established the SRAD (Alrai Newspaper, 2014 . Even Syrians not registered with UNHCR were expected to acquire these. Among them were Syrians I met who had lived in Jordan since fleeing Iraq in 2003, where they had been living as refugees before the Anglo-American occupation. The UVE's aim was to produce a verifiable count of Syrians. GoJ needed to regain credibility after it grossly exaggerated the number of Iraqi refugees in Jordan, diverting funds it raised to Jordanian developmental projects (Arar, 2017;Chatelard, 2010;Peteet, 2007;Seeley, 2010).

Azraq and Za'tari camps: Emblems of burden, tools of control
Encamped refugees, as visible 'spatially legible' populations, (Peteet, 2011, p. 18) can help secure aid (Harrell-Bond, 1998). Iraqis sought refuge in Jordan because of coercion associated with the Anglo-American occupation (Ali, 2020a, 2020b). They settled in cities, and Jordan struggled to acquire international aid for this 'invisible' population (Turner, 2015). Consequently, Jordan encamped many Syrians. They are only 20 per cent of the 660,000 registered with UNHCR, but the camps warrant attention because they encapsulate parts of the broader national agenda focused on security, control, and revenue generation. They are emblems of burden: beacons to attract international funding. I visited both camps briefly in early 2017 and spoke to key informants. I was fortunate to meet the SRAD Colonel in charge of Azraq camp who answered frankly when I asked why Jordan built the camps. To recap, he said: If we hadn't built the camps, then the world would not understand that we were going through a crisis. . . . The UN makes daily life better for refugees. We provide security.
As SRAD exerted greater control over the camps, MOI permits became obligatory for access. I received mine relatively quickly, in three weeks, but was only allowed one visit to each camp, minded by a SRAD officer throughout. Attached to my paper permit was a copy of the email request I was instructed to send, outlining my research. I declared the Swiss Federal Department for Foreign Affairs funding while emphasizing the academic independence of the research. The funder's name was re-formatted on my permit into bold, indicating the importance to the government of assisting someone perceived to be connected to potential donors, whilst carefully managing the visit.

Za'tari
Za'tari is a highly visible emblem of burden used to secure international assistance. The camp's early growth was shaped by Syrians' rapid self-organized practices that UNHCR could not control (Dalal, 2015). Consequently, Za'tari was seen as a planning failure, a case of 'losing control' (Dalal, Darweesh, Steigemann, & Misselwitz, 2018). Similarly, in Za'tari's nascent days, Jordanian authorities struggled to maintain their own notion of order. A senior humanitarian official I interviewed in Amman said that Syrian protests about their conditions were frequent and that the Free Syrian Army (FSA) was present in Za'tari: In 2013 the police presence was almost zero, they were afraid of the Syrians. Police responded to the dozens of protests that year only in full riot gear, 50 of them, firing tear gas.
The FSA in Jordan were not armed. But Jordanian elites feared a repeat of the Palestinian uprising, when a politically active Arab population, displaced by active conflict in a neighbouring country, established itself in Jordan. Palestinian political movements espoused radical politics that challenged the Hashemite monarchy which was firmly in the 'moderate' camp in the Arab Cold War. The PLO established a 'Guerilla Sanctuary' in Jordan, building proto-state institutions, openly rivalling the monarchy (Sayigh, 1999). Jordan's military expelled Palestinian fighters at great cost in the 1970-71 civil war (Sayigh, 1999). The potential parallels with the Syrian population, connected to the active conflict in Syria, and influenced by the Arab uprisings against authoritarianism, were surely not lost on Jordan's rulers. The MOI and SRAD saw Syrians mainly through a lens of security and threat. Tellingly, MOI resisted calls to include camp residents in the work permit scheme, singling them out as a 'security risk' until mid-2017 when the poor uptake of permits compelled a policy shift (Lenner, 2020, p. 296).

Azraq
The MOI and SRAD desire to control Syrians manifested itself in the securitized construction of Azraq. Officials from both GoJ and UNHCR wanted to learn from the 'mistakes' of Za'tari (Oddone, 2014). For UNHCR, the stated lessons related to improvements in safety and services -but Azraq's design made refugee-led organization impossible (Dalal et al., 2018) and ensured police and humanitarian control, to obstruct the possibility for Syrians' collective political expression (Dalal et al., 2018;Gatter, 2021;Hoffman, 2017). Humanitarian agencies assist the SRAD in the surveillance and control of Syrians: geographic information systems collect data about Syrians' behaviour (Hoffman, 2017). The camp's extreme securitization creates feelings of imprisonment and criminalization among refugees (Gatter, 2021). Azraq is isolated and also functions as a punitive tool for minor transgressions, such as for working without a permit before the Jordan Compact. For major transgressions, and accusations of terrorism, Syrians are deported to Syria (Human Rights Watch [HRW], 2017). Parts of Azraq were prisons within a prison. Residential villages 5 and 2 were fenced off in the beginning, apparently hosting Syrians from tribes that had sworn allegiance to the so-called 'Islamic State', explained the SRAD Public Relations officer accompanying me to UNHCR's Azraq office. UNHCR persuaded SRAD to open up Village 2, explained the senior humanitarian I interviewed, but at the time of research, Village 5 remained off limits and he did not expect that to change.

The Jordan compact: MOPIC and The Ministry of Labour
The 2016 Jordan Compact reduced restrictions on Syrian participation in Jordan's formal labour market. Bilateral agreements signed by Jordan and Syria allowed their citizens mutual visa-free movement and residency, but not complete freedom to work legally in both countries. Jordan's labour market faced numerous challenges before the Syrian crisis, including high unemployment and dependency on low-waged foreign labour (International Labor Organization [ILO], 2015, p. 3). Jordan protects sections of its labour force from competition: Engineering, medical, and legal posts remain closed to non-Jordanians through professional Guilds. Its commitment through the Compact to provide 200,000 formalized work opportunities for Syrians was widely praised. In exchange, the EU will relax trading rules for products manufactured in Jordanian factories that hire a proportion of Syrian employees. It is a legibility exercise devised by MOPIC and international funders, implemented by MOL. Syrians are rendered legible by and for security institutions like SRAD through the issuance of biometric ID cards. The work permits constitute an additional layer of legibility to conform to the needs of European Union and World Bank funders.
MOPIC was behind the initiative to formalize 200,000 work opportunities and not MOL, another reason behind MOPIC's major role in the response despite it being a service ministry. MOPIC's then Minister, Imad Fakhoury, played a key role in creating and negotiating the Compact, drawing upon his significant experience working for international financial institutions like the World Bank in addition to Jordanian government agencies. An official from MOPIC's Project Management Unit explained that MOPIC can block projects that conflict with GoJ's aims. MOL implements part of the Compact through directives (ta'leemaat) within existing labour legislation.
Before the Compact, acquiring work permits was very expensive; only 5,700 (less than 2 per cent) of the 324,410 permits issued were to Syrians (ILO, 2015, pp. 4, 10). Employers had to pay between 170JD and 800JD for a permit, depending on the sector (ILO, 2015, pp. 9-10). After the Compact, GoJ reduced administrative obstacles for Syrians to formalize, slashing fees to 10JD in April 2016. By January 2017, 37,900 were issued. As a further incentive, 'relaxed' labour inspections applied to Syrians and not to other workers, according to a MOL official. Employers received two warnings before facing court fines. Contrastingly, the official said, if the workers are not Syrian, there is one warning, then a fine, and possibly the courts terminating the business. Egyptian, not Jordanian, workers, have most to lose from the Compact, demonstrated by the nearly 5,000 deportations of Egyptians in 2016 (Abaza, 2016) who were working without permits. A high-level diplomatic intervention from Egypt early in 2017 gave workers a grace period of several months to acquire permits before deportation. But 'replacing migrant workers with Syrians will be tricky', explained Patrick Daru, ILO Country Coordinator in Jordan at the time, as: . . . employers don't always understand what the Syrians want. Most migrant workers are single and working 10-12 hours per day to send remittances home while Syrians have families here and manage high living costs in Jordan. They prefer multiple piecemeal jobs to factory work. This is particularly true of the workforce in the Qualifying Industrial Zones (QIZ) and Special Economic Zones (SEZ) where south Asian labourers reside on site in low-cost dormitories.
While innovative, the Compact will not upset the status quo by creating Syrian competition with Trans-Jordanians for employment. MOPIC understands that the Syrian presence can be used to secure revenues without creating competition. The closed professions will remain closed, and Jordanians will likely avoid the arduous and low-paid jobs that the architects of the Compact hope Syrians will take. Unemployment in Jordan is high. Authorities are sensitive to antagonizing a Jordanian labour force that is ready and willing to protest. A UNHCR Livelihoods Officer explained that the MOL was not actively advertising jobs for Syrians for fear of antagonizing Jordanians. Instead, UNHCR was doing most of the outreach work connecting potential employers with Syrian job-seekers, effectively acting as a recruitment agency. Furthermore, while donors were expecting 200,000 individual Syrians to work each year of the Compact, MOL, and Jordan's government more broadly, had in mind 200,000 work opportunities, measured by the number of work permits issued. As one MOL official clarified: 'If a Syrian works in one place, and then moves to another, then that is two employment opportunities.' By January 2020, MOL reported issuing, in total, 179,000 permits since 2016 (Ministry of Labour [MOL], 2020).

Jordan's Sharia Courts: Legibility, preventing statelessness
Jordan's religious courts (al-mahaakim addeeniya) are a government institution that do not see Syrians as a security risk but as families needing affordable documentation. They are not part of a security apparatus like Jordan's Ministry of Justice's courts dealing with crime and law enforcement. And they differ hugely from the SRAD. But their services do serve the broader security agenda in that the referent object they are securing is the demographic balance. By helping Syrians to document births and marriages, they lower the risk of a potentially stateless and permanent Syrian presence. The Sharia Courts (SC) belong to Jordan's national legal infrastructure and service institutions. They provide affordable family documentation and dispute resolution to Jordanians and more recently to Syrians. Islamophobia creates misgivings about the role of SC. Sharia simply means the rules and regulations pertaining to personal and family conduct in Islam, varying between different schools' (and governments') interpretations.
I interviewed Dr. Ashraf Al Omari, an Investigator at the Directorate of Reform, Mediation and Reconciliation of the Supreme Judge Department (SJD) (Qaadhi al-Qudhaat). SJD provides technical support and expertise to the SC, which are part of governmental religious courts. Legally trained professionals provide services -like divorces -subsidized by the state at half the cost of the Ministry of Justice courts (al-Qadhaa' al-Nidhami). Services such as family conflict advice, and mediation, are free for Muslims. Jordan has equivalent religious courts for Christians.
Dr. Ashraf explained the SJD's and SC's functions and their adaptations to support Syrians. The SC registers births and deaths, and mediates early-stage custody and divorce disputes without costly lawyers. 'We don't want to exhaust families with high legal costs for essential procedures' he explained. The SJD's expertise supports the SC's services for Jordanians. Despite overwhelming demand, Syrians can use its services, thanks to accommodating procedural innovations. In 2014, the SC introduced provisions for non-Jordanians to use Jordanian law, primally benefitting Syrians. SC established branches in Za'tari and Azraq camps, the first such courts in any refugee camp in Jordan, to register marriages, births and deaths. The SC made concessionary gestures where Syrian practices have transgressed Jordanian laws. Not registering a marriage in Jordan is a crime with punitive fines.
The fines were an obstacle to documentation. . . . The courts introduced grace periods in 2013 and 2014; no fines imposed as long as people registered their marriages.
The SC are under pressure as Syrians need documentation services. In Mafraq, a new office opened to cope with demand. Dr. Ashraf emphasized the importance of affordable services for Jordanians and Syrians alike, despite the extra cost to the government. Despite the administrative pressures the courts are under, Dr. Ashraf understood 'it would cause more problems in the long term' if the SC did not help. He added: Syrians may not be able to go to the Syrian embassy to get papers, they may be wanted by the regime. We . . . will not force a Syrian to go to the embassy if he feels he is in danger. This is in stark contrast to the security forces' practices of detaining Syrians in Azraq camp before deporting them to Syria.
The Sharia Courts represent a different aspect of the Jordanian government's refugee response. The court's religious element offers some equality as a place where a shared faith-based ethical framework underpins its work (albeit under another state's authority). The SC and SJD adapted to Syrians' needs and are not security institutions dealing with threats. Their aim is to assist families, but the long-and short-term harms of statelessness and of life without documentation are well known. Documentation means that avenues for Syrians' onward movement, however slim the chances, are not completely closed. The services therefore help to secure a referent object of the status quo: Jordan's demographic balance. A generation of undocumented births, from a Syrian population already marginalized by Syrian state authorities, risks creating a permanent undocumented and potentially stateless population on Jordan's territory. Although not expressed explicitly, the political importance of documentation support for Syrians in Jordan surely is not lost on Jordan's leadership.
The SC and SJD's efforts to document Syrians highlights the sometimes contradictory actions of the many hands of the Jordanian state when contrasted with Jordan's border authorities that confiscated 219,000 documents from Syrians (NRC & IHRC, 2016, p. 14). Many such documents were stored at Rab'a al-Sarhan registration centre, near the Syrian border, in coded paper envelopes. An accidental fire at the facility could have created a sizable undocumented and potentially stateless population. Fortunately, the practice of confiscating documents ceased in 2014. Thanks to a push from UNHCR, 180,000 documents were sorted, digitized, and returned in autumn 2016 (Alsalem, 2013;NRC & IHRC, 2016, pp. 14, 42). Until then, some 76,000 Syrian families were without their documents.

Municipalities: Different perspectives
Municipalities were chronically underfunded before Syrians arrived (Clark, 2012). At two near the Syrian border, local solidarities of tribe and kinship combined with the need to secure assistance from central government authorities and from international donors. The approaches of the officials I spoke with differed from national institutions as a result, I believe, of these social ties. Municipalities are the least powerful among the institutions considered here, and these are only two of Jordan's one hundred municipalities. Nevertheless, they demonstrate how heterogenous and sometimes incoherent states can be (Janoski et al., 2020), and also that legibility can come from below (Hussin, 2017). Additionally, they show how state institutions do not always hover above a society, but are influenced by cultural factors that blur the assumed lines between them (Mitchell, 1991;Morgan & Orloff, 2017).

Al-Za'tari and Al-Munshiya municipality (ZAM)
Al-Za'tari & Al-Munshiya municipality (ZAM), also known as Za'atari Village, is just a kilometre west from Za'tari camp. The roads were badly maintained, and ZAM receives little attention, globally and nationally, unlike Za'tari camp. In March 2017, I interviewed the head of its Local Development Unit, Shakir Al-Khalidi, ZAM's only remaining engineer. Others had left for better pay in the private sector. At the time, around 7,000-8,000 of ZAM's total population of 12,000 were Syrians who had arrived after 2011 and preferred living in ZAM to Za'tari camp. He explained that most of the Syrians were from the Khalidi tribe of Homs in Syria. They settled in ZAM because many locals are from Jordan's branch of the Al-Khalidis. Shakir, himself a Khalidi, estimated that 60 per cent of them lived in rented accommodation and the rest in tents scattered around the village.

Pressures on services
The large and rapid population increase tested municipal resources.
Our clinic previously dealt with 40 cases daily, and now we have two centres dealing with 70-80 cases. There are 60 students in the school classroom compared with 30-40 in the past.
The roads need extra maintenance, a duty of the municipality. Traffic has increased but not asphalt supplies. Locals are known for digging up roads to run electricity or water lines to their homes without municipal permits and, complained Khalidi, 'some will run another line or pipe to a tent, and the road will need additional repairs.' Additionally, ZAM deals with 5,000 tonnes per year of solid waste, up from 2,000 in 2011.
Za'tari camp caused ZAM problems and demonstrated the state hierarchy in the refugee response. The camp is on ZAM lands, but is run by MOI and SRAD with support from UNHCR and partners. Shakir explained that, in addition to litter landing in the municipality, it leaked sewage into the adjacent Za'tari valley and polluted nearby Jordanian-owned farmland. SRAD officers turned Shakir away when he arrived at Za'tari camp in a municipal vehicle. But he knew that shops operated inside the camp without municipal licences. Despite the camp's impacts on ZAM, it remained MOI's exclusive domain.

Donors and 'the organizations'
Donors imposed demanding legibility conditions on their aid, and the actions of others exasperated and confused ZAM staff. They secured a World Bank grant of 712,000JD to improve roads, and build 300,000 m2 of new ones; to buy 4WD vehicles, waste collection trucks, a digger, and a compressor; and 250,000JD to build a sports centre. But it comes with time-consuming accounting and legibility duties. When I met him, Shakir was completing a report detailing every municipal project's impact the previous three years.
. . . the conditions of the World Bank [pressure us]: creating a database, risk analysis, a census. . . . I work until the evening although my formal hours are 8am-2pm. I volunteer to do this, . . . it is my duty.
Conducting a census of non-Jordanian residents is not a municipal responsibility, but ZAM, a hand among many of the Jordanian state, lacked the resources to support the trebling of its population.
Donors helped ZAM establish a small garment workshop employing Syrians and Jordanians, and provide training courses in tailoring and I.T. ZAM owns a women's salon, offering training courses for Syrian and Jordanian women. ZAM also built a fruit and vegetable market with World Bank support.
World Bank aid was appreciated but not, he said, adequate. Follow ups, and proper understanding of context, were often absent. The World Bank funded tree planting, but not the necessary maintenance.
The man who runs the water well won't give us water without cash. The compressor machines need maintenance, people to operate them. An organisation may declare that they are going to support us to refurbish a workplace . . . and tell us to buy new furniture. But they don't appreciate that the law does not allow us to buy the furniture. . . . for fear of corruption. . . . . We are following the law; they don't understand it.
Certain humanitarian organizations and donors often bypassed the municipality, unconcerned with the impacts on the local population, perturbing Shakir. One donor distributed 1400 aid packages to Syrians, ignoring Jordanians in ZAM needing support. 'This has a negative effect on social views between people' he said. It was a private donor who ignored the municipality and went straight to local charities. Development or humanitarian projects require MOPIC approval, but some -most commonly private donors -evade regulations. But Shakir said that aid agencies engaged in similar actions without consulting ZAM. Mercy Corps appointed a group of 'Local Leaders' and ignored the municipality, even with large construction projects. 'We built a sports hall. Mercy Corps built a similar and more expensive one nearby without consulting us. Now we have two -in the middle of the area.' Shakir noted the added resentment Mercy Corps created by only working on one school in the area, a girls' school, while ignoring a boys' school also needing assistance.

Involving Syrians
The municipality involved Syrians in decision making processes, in consultations about new projects in the municipality such as the design of a new park, and included Syrian youth in sports matches in the stadium. ZAM also invited associations working with Syrians to participate in sessions and to share their views. 1 Litter from Za'tari camp landed in ZAM; Syrian and Jordanian volunteers participated in public clean up campaigns organized by ZAM. Cultural matters are important. ZAM arranged for Syrians and Jordanians to pray together on occasions of religious significance at the stadium, and distributed sweets during Eid to promote social interaction. Shakir was frustrated that the important social duties that he believed municipalities have was not reflected in formal legislation, nor in terms of adequate funding for those roles. His strong sense of duty towards Syrians in ZAM was, in my interpretation, partly related to tribal kinship. The boundaries between public and private, state and society, can be blurred by disaggregating the state (Morgan and Orloff, 2017, pp. 7-8). In ZAM, the supposedly sharp borders between host-state, host-community, and refugees, were blurred by tribal solidarity, faith-based hospitality, and a diligent municipal official. They combined to produce a civic duty that went above and beyond what could be expected of a chronically underfunded municipality.

Um Al-Jimal Al-Jideeda municipality (UJM)
Five miles from the entrance to Za'tari camp is the Um Al-Jimal Al-Jideeda municipality (UJM) building. In the mayor Hassan Araheeba's office, guests are greeted with baklawa sweets, and once seated can see on display the awards the mayor and the municipality have received. He is keen to emphasize the municipality's achievements and its credibility. He has been invited to speak at international conferences for mayors and local governance, and is eager to continue developing links to international organizations and local government networks.
Part of the Governorate of Mafraq, it comprises of 13 villages, and its territory shares a 36 km border with Syria and a 9 km border with Za'tari camp. UJM's population is 35,000 including 10,000 Syrians. There were 25,000 Syrians previously, but many left for Europe and elsewhere.
Araheeba was deputy mayor from 2003-2007, and has headed UJM since 2013, witnessing the Syrians' impact on UJM. They began arriving in 2011 at the beginning of the uprising, with significant anti-Assad protests happening in Der'a close to the border with Jordan. 'The Syrians are related to us in many ways which have led to our solidarity with them: tribally, culturally, by faith, and by Arab nationalism.' He explained that the Syrians in UJM were from branches of the same tribe in Syria, including the Masa'eed, the Sharafat, Adhamaat, and tribes from Zubaid. 'We wanted to help them, we have done everything we can, coordinating with state institutions, and with international organizations to help our Syrian brothers who have been persecuted.' Araheeba says 87 per cent of the camp is located in UJM and was named after the Za'tari valley, and not Za'tari village nearby. The camp is under the jurisdiction of MOI. Like ZAM, UJM was not consulted about building a camp there, and it feels the effects of both the camp and the Syrian presence.

Legibility: A local initiative
It became clear that Syrians would not quickly return to their homes. At this point UJM, acknowledging that they lacked the staff and resources to deal with the crisis alone, organized volunteers from the 13 villages into 11 Lujaan Mahaliyya -Local Committees formed of charities and individual volunteers. The Lujaan connected the municipality and the Syrians, and the municipality met the Lujaan weekly. With UJM, they counted them, and made their own needs assessments for Syrians and of the effects of their presence on UJM, independently of national state institutions. UJM made sure that no informal tents or shelters could be installed without the permission from the landowner: 'this helps us to know where the refugees are, who they are, and on whose land they are living'. It was an exercise in rendering the Syrian population legible, characteristic of a state institution.

Pressures
UJM had a special team conducting studies of the crisis' impact on the municipality. Heavy trucks supplying water to Za'tari camp's 81,000 residents damaged roads. Solid waste increased. When he was deputy mayor during 2003-2007, UJM collected 11 tonnes daily, rising to 15 tonnes in 2011. By 2017 they collected up to 45 tonnes per day. 'Another problem with the camp is the loose garbage that ends up in UJM. We distributed bags to people to collect this, but it costs us.' The refugee crisis exacerbated Jordan's pre-existing water scarcity. In UJM, as in most of Jordan, there are no pipes pumping clean water into homes. Municipal trucks deliver water, filling tanks kept on buildings' roof tops. 'We charge 10JD per transport of water -this is nothing, it is very cheap.' This is in addition to the pressures on the labour market. 'We have 1,300 Syrians working in agriculture. They displaced Jordanian labourers because they work for less. That is a big number in an impoverished area.' Perhaps this is why the Mayor was unconcerned when I mentioned that Syrians in Azraq camp were located far from work opportunities. 'Refugees don't need to work, in Azraq the UNHCR pays for everything.'

Assistance from the central state and international organizations
The municipal building faces Roman basalt ruins, regionally among the bestpreserved Byzantine sites, a heritage of which the mayor is proud. He lobbied for infrastructural support in a meeting with King Abdullah II. After he explained UJM's challenges, he had the attention of the ministries of Planning, Public Works, Tourism, Local Affairs, and the Directorate of Archaeology. The basalt ruins were restored and the roads leading to them and the municipality were resurfaced in a community regeneration project. UNDP provided direct support for cleaning and waste collection trucks, USAID provided almost 1 USD million, and UNHCR gave support worth around 800,000 USD. While grateful, the mayor noted it was not enough to cover costs.
Jordanian municipalities are not obliged to conduct a census of non-Jordanian populations. But UJM rendered Syrians legible for the central state, as well as for itself, 'from below' (Hussin, 2017) that conferred benefits from the King. I do not know how Araheeba secured a meeting with King Abdullah II, and as I visited ZAM before UJM, I did not think to ask Al-Khalidi if he had contacted the King. What is apparent is that Araheeba secured benefits using information collected by the 11 volunteer committees which UJM organized to render Syrians legible. The ability of UJM to organize such an endeavour with local volunteers also shows how the boundaries between state and society are blurred, and between refugee and host, and that cultural factors -like tribal, faith-based, and Arab-nationalist -shaped the lines between the three groups rather than there being a state that hovered above the host society and refugees.

Conclusion
Jordan, like most Mediterranean states, has faced the challenges of hosting displaced Syrians. It is a case study of how an authoritarian monarchy has managed the presence of a large displaced population without challenging the status quo upon which its power rests. Ruling elites aimed to preserve Hashemite rule by preventing Syrians from becoming a permanent presence and competing with Jordanian citizens for formal employment, while maintaining international assistance flows to Jordan. The aims have overlapped with the need to make Syrians legible to funders and to Jordanian authorities. The agenda of securing the status quo was not always coherent. A disaggregation of Jordan's refugee response has shown how different state institutions engaged in legibility with different intentions. Broadly speaking, they had commonalities, such securing support from international sources for local projects. However, I pointed to some incoherent and occasionally contradictory elements. Perhaps the most interesting of them was at the municipalities where, in particular geographies near Syria, I did not sense that resisting a permanent future presence of Syrians was a priority. The municipal officials were as keen to draw in material support from the central government and international organizations present in the country as was the central government to draw international funds into Jordan's coiffeurs. But the affinities they spoke of, and the obligations to members of shared social networks, undermined the dichotomies of citizen and foreigner, refugee and host society, and even host state and refugee at the municipal level. We ought not to assume an analytical separation between these categories when assessing refugee responses elsewhere. Those municipalities contrasted starkly with the national security perspective of the SRAD. The MOI appeared to be the most powerful of the state institutions that Syrians encountered directly, and one which often viewed them distinctly as a potential foreign security risk. Yet other significant institutions, like the Religious and Sharia Courts, appeared to view Syrians primarily as co-religionist families needing documentation. What shaped the many hands of the Jordanian state's refugee response was a combination of perspectives relating to finance, security, fears of history repeating itself, but also to affinities and care based on faith, family, and kinship. Furthermore, what was also evident was the active role of international actors at national and local levels of the Jordanian state, at times constructively engaged in supporting the refugee response, and at others complicit in the policing and control of a population already subjected to injustice by brutal counter-revolutionary forces in Syria. Scholars researching how other Mediterranean states hosted recent Syrian arrivals may find disaggregation useful. Considering the many hands of other states could offer interesting findings about the relationships between different state institutions, possibly revealing that the assumed lines between state and society, and refugees and hosts, are often blurred. Note 1. Other municipalities I visited, Sahaab and Zarqa, also consulted and involved Syrians in municipal activities to different extents.

Acknowledgments
I wish to acknowledge the generous funding from the Human Security Division of the Swiss Federal Ministry of Foreign Affairs, VolkswagenStiftung, Riksbankens Jubileumsfond, and Wellcome Trust, while noting the independence of the research. Additional thanks to Alex Betts for initiating The Politics of the Syrian Refugee Crisis project and for thoughtful discussions that helped to distil fieldwork. I'm grateful to Dawn Chatty for organising the symposium at NYUAD which spurred the writing of the paper; and to LSE's IGA for supporting its completion. Dawn Chatty and Cory Rodgers provided helpful comments on earlier versions. Many thanks to Matt Buehler and three anonymous reviewers for helpful feedback. I'm very grateful to all of the participants who helped in Jordan, and for a convivial residence at the British Institute, Amman.

Disclosure statement
Due to issues of confidentiality the data underlying this research cannot be made publicly available.

Funding
This work was supported by the Swiss Federal Department of Foreign Affairs; Riksbankens Jubileumsfond; Volkswagen Foundation; Wellcome Trust.