Making Refugees Work? The Politics of Integrating Syrian Refugees into the Labor Market in Jordan

Abstract Refugee response planners no longer frame Syrian refugees merely as objects of humanitarian care. Increasingly they are portrayed as enterprising subjects, whose formal integration into labor markets simultaneously can create self-sufficient actors and cure the economic woes of host countries. However, bringing together humanitarian and economic agendas is not an easy task. This article analyzes the contradictions and frictions that have emerged in the process of implementing the Jordan Compact, a political commitment to integrate Syrian refugees into the formal Jordanian labor market, and which is supposed to showcase such win-win strategies. It argues that the Jordan Compact should be seen as a policy model that has achieved enough consensus and incorporated enough disparate objectives to be labelled a ‘policy success.’ Yet, central actors have neglected core features of Jordan’s political economy and labor market, and/or the lives and survival strategies of refugees, such that their radical blueprints of transformation have been disrupted. Despite the widespread commitment to the scheme, it is thus unlikely that the Jordan Compact will both reinvigorate the Jordanian economy and offer Syrians the prospect of a dignified, self-sufficient life, an important lesson for comparable schemes being rolled out across the globe.

More than five years have passed since Syrians began to escape en masse to neighboring countries to seek refuge from violent conflict. While their presence primarily was dealt with in the first years through 'classic' international interventions that viewed Syrian refugees as objects of humanitarian care and assistance, there has been a notable change in tack since 2014. Western donors as well as international organizations have started pushing for the crisis response to be more development-oriented, aiming to make Syrian refugees more 'self-reliant' or 'resilient,' and thus less dependent on international assistance.
The Jordan Compact represents a milestone in this process. In the Jordan Compact, issued at the conclusion of the London donors conference for Syria and the region in February 2016, the Government of Jordan (GoJ) declared that in the coming years it would allow potentially as many as 200,000 Syrians to obtain work permits in Jordan. Syrians' formal labor market participation rapidly transformed from being a taboo topic into the new benchmark by which the government now would be held accountable by the international community and their donors. Not only that, but government representatives, donors, and humanitarian actors have presented these changes as a 'paradigm shift' in terms of how host states across the world could cope with large and protracted refugee situations.
'Turning the Syrian refugee crisis into a development opportunity' has become the new collective mantra, and Jordan has become the showcase of this projected 'win-win' solution. The possibility of socioeconomic integration for Syrians in Jordan regularly is presented, especially by international advisers, as the means simultaneously to alleviate Syrians' economic woes and to reinvigorate (but also reduce) international support for Jordan by boosting its economy overall, and to reduce the numbers of Syrians opting to make the journey to Europe. Yet as this article shows, this policy model not only is problematic as such but also proven to be much more complicated than its designers' depictions.
In order to demonstrate the significance and pitfalls of these supposedly fundamental policy changes in how refugees are governed, the article draws on academic discussions on humanitarian, refugee and development policies, as well as those on neoliberal reforms in Jordan and their effects on its labor market. In analyzing the main strategies pursued to implement this blueprint, the article teases out the political dynamics that have become apparent in the Jordan Compact's realization, and the main points of friction that have emerged. We argue that in an attempt to build a policy model that elicits enough consensus to be 'implementable,' and to achieve a variety of disparate objectives, long-standing features of political economy in Jordan have been under-appreciated by most of the actors involved in designing the scheme, many of whom lack experience with (Jordanian) labor market issues, and/or with the lives and survival strategies of (Syrian) refugees. Specifically, in its unfolding, the Jordan Compact has run up against three main underlying dynamics that have shaped Jordan's political economy for decades: (a) zonal development strategies, which have been a central but highly problematic feature of growth strategies in the country, (b) a nationally-segmented labor market that builds on specialized, precarious migrant labor, and (c) the relevance of informal labor and the political creation of informality. We argue that these three dynamics disrupt radical blueprints of transformation in ways largely unforeseen by their proponents, and already are starting to highlight the fundamental shortcomings of the initial compromise. Taken together, they will make it extremely difficult to square the circle of giving Syrians in the country the prospect of a dignified, self-sufficient life, while simultaneously making the Jordanian economic desert bloom.
To unfold this argument, this article begins by drawing on ethnographic literatures from the fields of development and humanitarianism to offer an analysis of policy models such as the Jordan Compact as compromises with inbuilt limitations and outlines the main features of the Compact and the Syrian refugee population in Jordan from this perspective. It then gives an overview of the Jordanian labor market, which provides an essential background to understanding how the attempts to implement the Compact have evolved. Subsequently, it examines the different strategies that are being pursued formally to integrate Syrian refugees into the labor market. Firstly, it analyzes attempts to expand investments, and thereby jobs for both Syrians and Jordanians, mainly in export-intensive industries in a number of Special Economic Zones. Secondly, it explores the difficulties experienced with efforts to substitute Syrian refugees for other migrant workers, with a specific focus on a high-profile garment industry pilot. Thirdly, the article turns to the informal labor market, where policy-makers hope for the formalization of existing Syrian labor in sectors such as construction and agriculture. 1 In all three instances, we argue that, despite the high-profile alliances that have pushed to turn policy blueprints into reality, the defining features of the Jordanian labor market and political economy, which have developed over decades, have substantially complicated their realization. Our argument is based on 60 semi-structured interviews and numerous informal conversations conducted jointly or individually with staff of key agencies involved in these attempted policy shifts, as well as with Syrian refugees in Jordan, between June and November 2015, April and August 2016, and April and May 2017, and on an analysis of relevant policy documents and media reporting over the course of 2016/17.

The Jordan Compact: Making of a Global Policy Model and its (Syrian Refugee) Target Group
The Jordan Compact marks not only a shift in globalized thinking about solutions for protracted refugee situations but also a remarkable compromise. It has mobilized a range of actors, agencies, and agendas, and tied them together in a joint project. Such convergences can be created, as authors like David Mosse have argued, by a shared desire to create development policy successes and generalizable, global policy models. Yet these policy models tend to be reliant on their relative detachment from the complex realities upon which they are supposed to act. Different agencies enter into such policies or projects if they can (re)interpret them to fit their specific agenda, yet the different, and often conflicting, interpretations that circulate are either the source of, or point to, unresolved conflicts and frictions. 2 This disjuncture between plans and realities on the ground is an established subject of critical writing in analyses of development policies and projects. Scholars have analyzed the many ways in which communities of experts, (self-)mandated as trustees, misunderstand, explain away or simply ignore the complex realities of the populations and territories they are supposed to 'improve', and the effects these complexities have on the trajectories of such interventions and their target groups. 3 This is also the case for humanitarian workers and humanitarian interventions, which often are characterized by a particularly strong detachment from the displaced populations that are nominally the object of their care. 4 The dynamics that create and maintain the networks of actors and agencies required for such projects of improvement are somewhat less commonly in focus within this literature, but have been highlighted by various authors who ethnographically analyze processes of making and implementing policy in different fields. 5 The Jordan Compact illustrates these very dynamics-both the coming together of divergent actors to form an ostensibly shared set of goals, and the remoteness of their compromise from the realities on the ground. Marking this particular context has been the coming together of two fields of expertise that were not so strongly integrated in the recent past. 6 The search for solutions for protracted refugee populations in countries of first asylum has prompted forced migration experts, humanitarian organizations, NGOs and governmental actors to move their focus (once more) from 'traditional' humanitarian work toward labor market interventions. Simultaneously, economic planners, labor market experts, and businesses have had to integrate refugees into their conceptual and practical horizons. As we explore below, this, on the one hand, has created new relationships and alliances. On the other hand, this coming together of agendas and actors has revealed a number of unresolved inconsistencies, frictions and exclusions that have haunted the project since its inception. These have resulted from the structuring dynamics of the Jordanian labor market and the strategies and priorities of Syrian refugees, as well as their neglect by many of the actors involved.
At first glance the key stakeholders-including the GoJ, the United Nations High Commissioner for Refugees (UNHCR) and its partners, and international donors-have made notable strides in terms of implementing the Compact. There are about 320,000 registered adult Syrian refugees in Jordan, and between the start of 2016 and early May 2017, almost 51,000 annual work permits had been issued to them, 7 although slightly fewer than 40,000 of those permits were valid in May 2017. 8 In part, this progress was due to the GoJ providing a grace period for formalizing the status of Syrian workers, in which there were no work permit fees for employers, no fines for not having previously applied for work permits, and during which Syrians found working informally would not be sent to camps. International donors matched Jordan's commitment; most notably, the World Bank endorsed a plan to facilitate access to up to $1.4 billion of credit at rates typically only available to lower-income countries and signed a $300 million, 35-year loan in the months following the Jordan Compact. 9 Numerous other donors are supporting different schemes intended to increase Syrians' labor market participation and the numbers of work permits issued. levels increased significantly after the Jordan Compact, even though there has been a slump in the funding of the Jordan Response Plan (JRP) in 2017. 10 This progress might not have been expected. As one of the most important Western allies in the region, it often has been sufficient for Jordanian governmental actors to declare their commitment to a variety of donor-sponsored 'development opportunities' in order to get much-needed funding, without necessarily following through with plans that donors envisaged. 11 This is even more noteworthy given the sensitivity of the subject: until the negotiations that culminated in the Jordan Compact, GoJ authorities were loath to imply publicly, through words or actions, that Syrians may be another (semi-)permanent refugee population, in a country in which the divisions between populations of Palestinian and 'East Bank' ancestry are an important political fault line. 12 This included restricting Syrians' access to the formal labor market, attempting to keep camp spaces appearing temporary, and restricting non-governmental organization (NGO) programming on work and livelihoods, such that even 'vocational training' was deemed too politically sensitive a phrase to use. 13 Following through with the Compact's main stipulation to provide Syrians with work permits is thus a notable change. At the same time, a number of inconsistencies become visible as soon as one takes a closer look at the Compact's aims and its practical limitations.
Cracks begin to show when asking even very basic questions, such as who does the target group actually comprise. Of particular note is the vexing question of how many working age Syrians reside in Jordan, and thus how many work permits might plausibly be given or wanted. As of the summer of 2017, according to UNHCR, there are approximately 660,000 registered Syrian refugees in Jordan, a number that has increased only very slightly since the Compact was issued. Of these 660,000, approximately 150,000 are registered camp residents, while the rest are registered as living in Jordan's cities, towns, villages and farmland. In contrast, the Jordanian census of December 2015 reported that there were just over 1.25 million Syrians (refugees and non-refugees) residing in the country. This census number has been treated with strong skepticism in the humanitarian sector, in particular considering the controversy that surrounded the allegedly inflated numbers of Iraqis in Jordan in 2007. 14 10 The Jordan Response Plan (JRP) was established in 2014 as a government-led response plan to the Syrian refugee presence, and as part of the GoJ's ongoing efforts to refocus the response to target Jordanian host communities more strongly. The question of how many Syrian refugees live in Jordan has an obvious bearing on the viability of the 200,000 target. Slightly over half of registered Syrians in Jordan are under 18, and therefore ineligible for work permits; further, in the occupational data that UNHCR holds on approximately 290,000 camp and non-camp Syrians aged between 18 and 65, over 100,000 are housewives, and many thousands may be physically or mentally unable to work. Camp residents, who are more than 20 percent of registered Syrian refugees in the country, also have experienced difficulties in accessing work permits. This is despite initially being, at least implicitly, one of the main targets for the work permits scheme. Different actors involved in governing Syrian refugees, including the Ministries of Interior, Labor, and Planning, have to some extent contrasting agendas, which variously emphasize control over the entry and exit of people and goods into and out of camps, or the economic integration of Syrian refugees. These divergent agendas have led some officials to attempt to disrupt the process of granting camp residents work permits.
Even for those Syrians in Jordan who are willing and able to work, 15 obtaining a work permit is far from straightforward. Characterized as 'the work permit maze,' 16 the process of obtaining a work permit can include reclaiming confiscated documents, applying for a Ministry of Interior (MoI) service card (which in late April 2017 only just over 200,000 non-camp Syrian adults held), 17 a security check, obtaining a rental contract or an address statement from UNHCR, and paying fees for administration and for health certificates, fees that are applied inconsistently across labor directorates. 18 Some Syrians living in urban areas are unable to obtain an MoI card because they left the Za'tari refugee camp after a particular date without the requisite paperwork, making them ineligible for an MoI card, and thus also for a work permit. 19 Another consequence of who was brought to the table and the compromise they made has been the exclusion of non-Syrian refugee populations from the Compact. These include the approximately 63,000 registered Iraqi, 8,000 Yemeni, and 3,500 Sudanese refugees. 20 Given their relatively small numbers in relation to Syrians, it should be relatively easy to facilitate access to the labor market for other registered refugee populations as well. 21 However, the strong prioritization of Syrian refugees in recent donor-Jordan relations, the lack of advocacy 15 An ILO report estimates that in the spring of 2017, around 85,000 Syrian refugees were working formally or informally. ILO (2017 on behalf of non-Syrian refugees in Jordan, and humanitarian agencies' focus on problems that are deemed both pressing and solvable, at the expense of thornier and/or seemingly unimportant questions, prevented the inclusion of non-Syrian refugees. 22 While these factors limit the number of refugees who plausibly could access work permits, this is far from an exhaustive explanation for why attempts to implement the Jordan Compact have faltered. In these attempts, the stakeholders of the Compact have confronted a reality more complex than the one sketched out in their blueprint for transformation, particularly in terms of Jordan's labor market. This process has revealed the divergent interests and agendas that underlay ostensibly shared objectives.

Jordan's Labor Market after 20 Years of Neoliberal Transformation
The inconsistencies, frictions and exclusions that have complicated the implementation of the Jordan Compact must be understood in the context of the broader dynamics that characterize Jordan's political economy and labor market after more than 20 years of neoliberalizations. The Compact resembles these broader strategies of transformation. At the same time, it cannot overcome the contradictions and structuring dynamics of the labor market, which in part are caused by, but also have disrupted, different neoliberalizations. Like the Jordan Compact, many of the (attempted) transformations of Jordan's political economy have relied on policies and rationalities that purport to solve countries' socioeconomic challenges through globally-circulating neoliberal solutions, which seek to increase the 'market orientation' of economies and societies, reinforce free trade and minimize government intervention, while curbing public spending, especially on social issues. Long evident in the country's relationship with the International Monetary Fund (IMF) and its repeated implementation of structural adjustment policies, 23 this approach has been embraced and spurred on by a new generation of technocrats who became increasingly powerful upon Abdullah II's coronation as King of Jordan in 1999, and other major international donors have supported it. These rationalities have fueled the reorganization of space to suit the interests of capital through the creation of development corridors, special economic zones, and other spaces of exception. 24 Zonal development technically has been pursued since the 1970s, when a number of Free Trade Zones were created, although it only really took off with the establishment of Qualifying Industrial Zones in the mid-1990s, followed by the Aqaba Special Economic Zone Authority (ASEZA) in 2000, and a number of Special Economic Zones since 2008. To its promoters, zonal development is a vital element of the vision to turn the country into a regional investment hub, elevate the role of the private sector, and integrate Jordan more firmly into a truly globalized economy. 25 Accompanying such macroeconomic strategies have been various attempts to foster entrepreneurialism and self-reliance, e.g., through microcredit programs targeting different social groups as well as various (community) development 22 Author Interview with European donor agency official, Amman, April 2016. 23 Lamis Andoni and Jillian Schwedler (1996)  projects in urban and rural spaces, 26 the ostensible aim of which has been to get Jordanians out of a supposedly outdated orientation toward public sector employment, enhance 'human capital development,' and create better 'fits' with the labor market. While such initiatives have transformed conceptions and spaces of the political, 27 their successes-as defined by their own goals-have been questionable. Public sector employment remains crucial and desirable to many Jordanians, and microenterprise projects remain fraught with contradictions that are relatively apparent to their participants. 28 Most of the special economic zones not only have been relatively unsuccessful in attracting investment but also they have failed to create positive spillover effects on local economies and labor markets. This is not least due to an over-reliance on migrant labor. 29 While large-scale labor migration, both to and from Jordan, preceded the rise of these projects to neoliberalize Jordan's political economy, it has since become embedded in these attempted transformations. In the context of the oil boom in the 1970s, many Jordanian citizens sought opportunities in the oil-producing states of the Gulf. The remittances from this employment fueled a demand for services and investments in infrastructure, which in turn dramatically increased the number of migrant workers in Jordan, particularly from Egypt and South East Asia. 30 These trends, which continue to this day, reinforced the need for migrant labor to do many of the low-skilled jobs that the Jordanian labor market in fact offers-for example in sectors such as agriculture, construction, the service sector, and domestic labor. More recently, labor market reforms have focused on reducing the cost of labor and increasing labor market 'flexibility. ' 31 This has been achieved both through the use of migrant labor, and its spatial segregation into economic zones. The garment sector, one of the areas discussed in this article, provides a particularly striking example of this trend. The sector has a quota that allows up to 75 percent of its workforce to be non-Jordanian. This extensive use of migrant labor, and the specific labor regime to which workers are subjected, is a key factor in the success of the garment industry, which is one of Jordan's most significant export sectors. This labor regime, helped by various exceptions from the labor law, exploits the isolation of migrant workers from their surroundings, increases their availability to work longer and harder, and thereby enables factories to respond to the ever-increasing pressure on suppliers for on-time delivery and shorter production times in the global garment industry. With its reliance on the free availability of such dis-embedded workers, this regime has become the cornerstone of garment production in Jordan's Industrial Estates, formerly designated as QIZs. 32 Another way in which labor market 'flexibility' in Jordan has been produced and reinforced is informality and the non-enforcement of regulations. Although structural adjustment policies in Jordan did not lead to the same degree of informalization of labor as elsewhere in the Global South, 33 a 2013 United Nations Development Program (UNDP) study reported that informal employment accounts for 44 percent of all private sector employment in Jordan. 34 Its scope presumably further has increased since that time, in part due to Syrian (and other) refugees entering the labor market. Many migrant laborers working in the informal economy either do not have work permits or are working outside the sector for which they originally were granted a permit. 35 This form of (non-)regulation of migrant labor, which allows migrants into the country but only partly formalizes their status, constitutes a vital part of neoliberal forms of governance, in that it allows for further deregulation of labor markets, and creates downward pressure on wages. While not necessarily an intentional, top-down strategy controlled by political decision-makers, it becomes effective though a myriad of individual actors pursuing strategies (on a local level) that are politically tolerated or facilitated. 36 In Jordan as elsewhere, illegalization and the non-enforcement of labor market regulations have benefitted employers, who profit from a labor force that is made vulnerable to exploitation and that tends not to claim its labor rights. 37 As a result of these broader developments, the Jordanian labor market is characterized by a high degree of segmentation between the public and private sectors, formal and informal economies, genders, Jordanian nationals and migrant workers, 38 and also between different nationalities of migrant labor. Different nationalities are (formally or informally) recruited in accordance with particular skills ascribed to them, which supposedly qualify them for work in specific sectors. 32 Factories in these estates, which used to be designated, and are still often referred to as, Qualifying Industrial Zones (QIZs), produce exclusively for the US market under the Jordan-US Free Trade Agreement (FTA), and pay only a 5 percent income tax and no export taxes, while having to ensure that 35 percent of the value of exported goods is composed of local content. The origins of the scheme were tied to the Jordanian-Israeli peace agreement in 1994, and the QIZ regulations used to require a minimum of 8.5 percent of components to come from Israel. These regulations were Yet these neoliberalizations should not be conceived of as proceeding in smooth and homogenous ways. As a new generation of literature has demonstrated, such forms of governance are uneven and variegated. 39 In Aihwa Ong's terms, neoliberal reforms bring market-driven calculations 'in the management of populations and the administration of special spaces,' 40 but 'exceptions to neoliberalism are also invoked, in political decision, to exclude populations and places from neoliberal calculations and choices.' 41 Such exceptions can be understood as a way to appease important constituencies or, more abstractly, as articulations with alternative, residual or currently marginalized projects, which inscribe themselves into dominant political economic strategies. 42 One way in which such exceptions are evident in the Jordanian labor market is through the existence of 'closed sectors.' There are at least 17 job types or sectors, including engineering, teaching, medicine and many service sector jobs, that are formally closed to non-Jordanians, even though many non-Jordanians work in these sectors informally, and there is even some unclarity about exactly which sectors are 'closed.' 43 This exception to the neoliberal logics that drive the use of migrant labor in Jordan protects the interests of a domestic constituency, whose protestations about reduced public sector employment opportunities and high unemployment dovetail with a narrative that there are 'too many' labor migrants in the country, and that these migrants are competing with Jordanians for jobs. 44 The current attempt to integrate refugees into the labor market should be read as another rendering of these attempted transformations. Jordan is once again a laboratory for sweeping changes, but this time Syrian refugees have been singled out as the population that can reinvigorate Jordan's economy. Rather than being objects of humanitarian care and maintenance, Syrian refugees are framed as unused human capital, which can be made productive as a new category of migrant workers and help Jordan to overcome the long-standing deficiencies of its production regime. At the same time, attempts to preserve the aforementioned exceptions to neoliberalism remain apparent, particularly in the GoJ's staunch refusal to contemplate opening up the closed sectors-even though such a move could substantially increase the numbers of work permits issued to Syrians. 45 Given the Compact's articulation with these previous strategies of transformation, it is unsurprising that, in its implementation, it has encountered many of the same stubborn and persistent challenges facing previous attempts to restructure Jordan's political economy and labor market. The Compact particularly has run up against the problems inherent to attempts to create an export-driven economy through zonal development, the consequences of a nationally-segmented labor market and the regimes of control it relies on, and the country's informalized economy. This becomes apparent when analyzing the key strategies pursued to formally integrate Syrian refugees into the labor market.

Changing the Rules of Origin, or Special Economic Zones revisited
At the center of the policy instruments sketched out in the Jordan Compact was a renewed focus on Jordan's Special Economic Zones (SEZs) 46 as drivers of increased manufacturing production, combined with the renegotiation of EU-Jordan trade relations. Much like the Compact overall, this component signals the successful integration of the agendas of a wide range of actors and institutions. Yet, it has not been able to produce the promised rewards, or to overcome the underlying problems that belie the zonal development model. The expected reinvigoration of manufacturing through the Compact was based centrally on the EU's commitment to changing the rules of origin (RoO) that govern the Jordan-EU Association Agreement. Much like elsewhere in the Global South, these RoO largely have prevented significant (Jordanian) exports to the EU by setting a high threshold for the percentage of any exported good that needs to originate in the country in order to qualify for tariff benefits, and by defining 'origin' in a narrow sense. 47 Given the low percentage of local raw material input in most goods exported from Jordan, these rules have acted more like a trade barrier than a means to facilitate Jordanian exports to the EU. 48 Renegotiating them was thus high on the list of Jordanian priorities for the London donors conference. The subsequent agreement, concluded in July 2016, eased these regulations for a number of designated SEZs, and Jordanian officials celebrated it as a major breakthrough. They declared that it not only would make manufacturing more profitable, facilitating investment and expanded production levels, but also create tens, if not hundreds, of thousands of jobs for Syrians and Jordanians. 49 The agreement stipulates that for a period of 10 years, effective immediately, Jordan and the EU temporarily relax the applicable RoO for individual factories located in 18 industrial and development zones across the country, and that employ no less than 15 percent of Syrian labor. From the beginning of the third year, factories will be required to employ a minimum of 25 percent Syrian refugees to qualify for the new rules. These relaxations adopt a wider definition of the types of processing required for non-origin products to obtain originating status, thereby making it easier to export products to the EU for fewer or no duties. 50 The 46 These zones come under a variety of designations, with differing legal regulations and regulating institutions. The most important ones are classified as Development Zones / Areas, Industrial Cities or Estates, Free Zones, and the Aqaba Special Economic Zone (ASEZ) is classified as its own type. Given that the broader literature generally refers to such zones as SEZs, we use it as an umbrella term that encompasses them all. See further Hanieh,Lineages,  scheme is supported by a substantial concessionary World Bank loan and a 5-year tailored World Bank project, which particularly focuses on attracting investment, changing labor regulations, and improving Jordan's investment climate. 51 This striking deal marks a convergence of different logics and agendas and has created new alliances between hitherto unconnected actors. It uses Jordan as a laboratory and a showcase for a new global blueprint for economic development in refugee hosting states, thereby attempting to marshal the unused 'human capital' of refugees. Driven by donor pressure to bring down aid levels (as well as the desire to find solutions for refugees in their countries of first destination), humanitarian and development actors, for some time, have been thinking about ways to accomplish a shift in this direction. 52 Recently, this process has been connected with the idea of employing refugees in spatially contained zones. In an influential proposal, Alexander Betts and Paul Collier, two Oxford University professors who undertook a study tour of Jordan in conjunction with UNHCR, advocate such a fresh approach to refugee 'integration' in countries of first reception. The idea is to reconcile refugees' needs and wishes for self-sustainability with the security concerns of host states and their aversion to long-term integration, by fostering a refugee-driven economy that is removed from the broader labor market, thereby taking pressure off it. Rather than refugees being an economic burden, this zonal economy-a second-tier and temporary form of socioeconomic integration-would boost the overall local economy, particularly in export-oriented manufacturing. At the same time, formal employment possibilities in neighboring states would provide the incentive for Syrians to stay where they are, rather than attempt the journey to Europe. 53 Mediated through the global networks of expertise in which the promoters of this new scheme are embedded, 54 the model already has gone global. It has been taken up as a blueprint by the UK government, which recently announced that the scheme, originally developed for Jordan, also will be applied to Ethiopia, where the creation of industrial parks is supposed to create jobs for Eritrean refugees, and in effect stop them from coming to Europe. 55 The idea has engaged the imagination of a variety of players who currently are seeking to redefine responses to refugee crises, 56  Yet these supposedly global schemes always need to mobilize adherents in specific settings in order to become meaningful and effective there. 57 In Jordan, the WANA Institute, a socioeconomic think-tank, played a decisive role in fleshing out the idea. 58 Sponsored by King Abdullah II's uncle, Prince Hassan bin Talal, and headed by Dr Erica Harper, the wife of the (then) UNHCR country representative for Jordan, the institute's participation secured crucial royal buy-in, and created a close link with UNHCR, as well as the Oxford professors. 59 The scheme also was shaped by the agenda of the proponents of zonal development, for whom it has provided an occasion to re-create the raison d'etre of economic zones. 60 The reinvigoration of zonal development strategies is taking place despite their questionable record worldwide, 61 and despite a dispute about their overall benefits to Jordan's economy and labor market. 62 While the Industrial Estates/QIZs and ASEZA have managed to attract substantial investments, this is less the case for the Development Zones, some of which only exist on paper, whereas others have a mixed record at best. 63 Yet, with such an array of prominent agencies and individuals as driving forces, ostensibly united around a common set of aims, the idea gained considerable traction. At the same time, the heterogeneity of their agendas, as well as the detachment of many stakeholders from, and limited understanding of, the actual functioning of the manufacturing industry, Jordan's trade regime and/or the SEZs in Jordan, also has caused considerable friction.
In the lead-up to the EU deal, some conflicts of interest became visible, for example in discussions about whether the new regulations should apply by factory, by zone, or across sectors and spaces. The GoJ and Jordanian business associations pushed for a deal that was as open as possible, creating access to the EU market in return for a certain percentage of Syrians being employed in a sector, zone, or in the country as a whole. 64 Yet, in order to ensure adherence to the regulations could be monitored effectively, the EU negotiators insisted on making the new RoO applicable only by individual factory and only within SEZs, 65 thereby reducing the number of firms that could readily take advantage of the rule changes. This also de facto means that improved accessibility to EU markets will be limited to companies that actually succeed in employing the required percentage of Syrian refugees. Jordanian negotiators succeeded, however, in adding a substantial number of zones to the five government-run development zones for which the scheme originally was foreseen in 57 Cf. Anna L. Tsing (2005)  the Jordan Compact. 66 Factories and sectors located in Jordan's Industrial Estates, such as the garment sector that exports successfully to the US market, now-at least in theory-also can gain access to the European market. Many firms in Jordan, however, have long been unable to export to the EU because of the relatively stringent standards required of all goods entering the European single market, which this deal does nothing to tackle. 67 The deal that was negotiated in effect has meant that only a very limited pool of companies potentially could benefit from it.
Perspectives not only have differed on the opportunities offered by the specific regulations, but also regarding the very nature of the anticipated investments. Rather than discussing the economic utility and feasibility of the scheme, public discussions often have framed future potential investments as a European duty, or, alternatively, as a gift that Jordan deserves. To give just one example, an op-ed in the daily al-Ghad, echoing governmental statements, points out that given European governments' fear of (more) Syrians taking refuge in their countries, it is their duty to create investments in Jordan, which can provide hundreds of thousands of jobs for Syrians and Jordanians. 68 In such statements, investments appear not so much as driven by economic opportunity, but rather as a form of aid, which is assumed to be coming, if only you ask for it strongly enough. While this latter dynamic is a very well-established one in Jordan, particularly in times of crisis, 69 actually attracting investments may not be so straightforward.
For those who are supposed actually to establish businesses in the designated areas, profit motives and calculations remain a central consideration. The possibilities of preferential access to EU markets, the inclusion of refugees in corporate social responsibility (CSR) strategies, 70 as well as the waiving of fees for work permits plausibly could interest potential investors. Yet, factors like production and shipping costs, the quality of infrastructure and the transparency of regulations play a decisive role in these decisions as well, and Jordan fares relatively poorly on most of these fronts. Production costs remain comparatively high due to high costs of industrial inputs, especially fuel and electricity. Coupled with comparably low productivity levels among the local workforce, the costs hamper competitiveness with products from India, China or Turkey. 71 In spite of a new investment law that is supposed to facilitate investment, Jordan's rank in investment-related indices has deteriorated in recent 66 The deal includes the originally foreseen governmental Special Economic Zones as well as Industrial Estates (formerly QIZs), and a number of 'Industrial Areas' that do not have a designated status, but rather constitute sites or conglomerates of factories. The GoJ argued that these latter areas would be suitable for inclusion due to being characterized by low income levels, high unemployment, and a large number of Syrian residents. See Tariq  years. 72 While Jordan's political stability may have been a factor in establishing factories in the QIZs in the 1990s and early 2000s, 73 the Arab uprisings, the Syria conflict and the rise of Islamic State, as well as the closure of borders and trade routes since 2011, have changed prevailing perceptions of Jordan and the region more broadly, and may well deter many potential investors.
Given these obstacles, uptake has been slow. By May 2017, fewer than 10 companies had registered to export under the new RoO regulation, and only one actually had done so, 74 a turnout that has resulted in a blame game among donors, the government and the private sector, 75 but also in renewed efforts to find or mobilize potential candidates. 76 Yet it is uncertain from which countries and sectors new investments conceivably could come in the future. While Syrian investors are among those expected to take advantage of the new regulations, 77 many already have found other bases in the last five years and probably will be reluctant to relocate again. 78 Large Western companies, despite being involved in initial discussions surrounding the Compact, have not made the hoped-for investments, and Jordanian companies themselves are unlikely to fill in the gap. While they have suffered greatly from conflict and border closures in the region and are therefore looking for new markets, they often lack connections with potential buyers in the EU, and the knowledge about EU markets and product standards required for certification. 79 Accordingly, there is currently hardly any local industry that actually would be ready or able to export to the EU market.
These unresolved difficulties are compounded by the fact that companies in the SEZs find it very difficult to recruit a sufficient number of (qualified) Syrians to fill the required quota, 80 for reasons that are discussed below. Similar difficulties have plagued the attempts to use 72 al-Khatib et al.,  A factory manager in one of the Private Industrial Parks highlighted the tax-free regulation for these zones, as well the fact that Jordan was more stable than Bangladesh-where much of the global garment industry is located-as points of attraction. the scheme to employ Jordanians. This is an important aim for the GoJ, which needs to (be seen to) promote the interests of Jordanian workers, as governmental and media discourses systematically have portrayed Syrians as 'taking their jobs' and creating unmanageable burdens for Jordanian society. 81 To this constituency, the policy change has been sold not just as a way of taking pressure off the broader Jordanian labor market by 'zoning' Syrian labor, but to revamp the economy and solve Jordanians' economic woes. Jordan's politicians, international organizations and donors alike have envisaged that the scheme could boost the manufacturing sector and thereby contribute to reducing unemployment, particularly among the youth. 82 Initially, a 50/50 quota for the employment of Jordanians and Syrians in the SEZs was discussed in order to lend credence to these claims. 83 Yet, it is questionable whether the rate of Jordanians working in manufacturing 84 easily could be increased. The largely blue-collar jobs that the manufacturing sector can offer do not meet the expectations of many increasingly well-educated Jordanians, who usually aspire to white-collar jobs. 85 In light of this, within a few months the proposed quota had been dropped, with donors now discussing, in general terms, the benefits of investment and the resulting multiplier effect for the Jordanian economy. 86 From the GoJ's perspective, increased Jordanian employment (along with donor funding) is one of the key goals of this scheme, but it remains unclear how this is to be achieved. In sum, the scheme has managed to mobilize the imaginaries, interests and hopes of a number of crucial actors and institutions, making it one of the most discussed current policy strategies regarding refugees in the entire region, and other parts of the world. Yet, it has not been able to overcome the contradictions of the zonal development model, nor the long-standing dynamics that have impeded the development of a manufacturing industry.

Garment Factories as Laboratories for Syrian Employment
Proponents of the Jordan Compact are detached not only from the situation on the ground, including the realities of the Jordanian trade regime but they also are detached from the role of Syrian refugees as a (potential) labor force and the question of how their lives in Jordan fit with the production regime in the manufacturing sector. Attempts to recruit them to work in the sector, and replace other migrant workers with them, have proven very difficult. This became particularly apparent in a high-profile pilot scheme in the garment sector. Announced in April 2016, shortly following the Jordan Compact, the scheme envisaged that 2,000 Syrian refugees quickly would be recruited to work in the garment sector factories within 81 84 Noting significant difficulties in obtaining data, a recent report estimates that the percentage of Jordanians working in the development zones listed in the EU-Jordan agreement is around 50 percent, whereas it is only 24 percent in the Industrial Estates, which focus on textiles production. Company owners in both types of zones reportedly have trouble recruiting more Jordanian workers due to the bad image of working conditions in the manufacturing industry. See Toppen et al.,Work Opportunities,41. 85 See Assaad,'Structure and Evolution',Toppen et al.,Work Opportunities,41. 86 Author Interview with employees of MoL Syrian Refugee Affairs Department, Amman, April 2017; see also World Bank, 'Jordan'. Jordan's Industrial Estates. 87 With the garment sector being one of Jordan's most successful export-oriented sectors to date, 88 it attracted attention from a range of actors seeking to think creatively about solutions to the labor market integration of Syrians. The pilot, which brought together JGATE (The Jordan Garments, Accessories and Textiles Exports Association), UNHCR, the World Bank, the Jordan Investment Commission (JIC), the Jordan Industrial Estates Company (JIEC), the Ministry of Labor (MoL) and others, was hailed as a promising partnership with the private sector, one which could lead the way in making Syrians more economically self-sufficient and simultaneously benefit the Jordanian economy. However, so far it has failed to accomplish these goals.
To implement the scheme, UNHCR and Better Work Jordan, an organization formed by a partnership between the International Labor Organization (ILO) and the International Finance Corporation (IFC), organized job fairs in community centers close to the industrial zones in question, in an attempt to inform Syrians about the new opportunities in the garment sector, and to sign up those who were interested. These job fairs are widely perceived to have been unsuccessful-a fact that initially greatly surprised many of the architects of the scheme. 89 Syrians were reluctant to sign up to work in the factories, which offer very poor levels of pay and working conditions, and are often located far from where they are living. While this work still at least can be financially interesting for the female South Asian workers who make up the bulk of the workforce, and who usually come to Jordan without families, live on site and do not have many other expenses, the situation is different for Syrians living in Jordan with their families. They have many in-country expenses, and rent alone easily can exceed this meagre salary, which is much less than in sectors such as construction. 90 Despite assurances from UNHCR that receiving a work permit would not affect their status, Syrian refugees expressed concerns that taking formal jobs may do exactly that, or reduce their opportunities to access financial assistance or resettlement 91 -very valid concerns, given that donors are seeking to reduce funding levels as well as migratory movements by finding ways to make Syrians more 'self-sufficient' in the medium term. Syrian women were the (implicit) main target population of this scheme. If the target of 200,000 Syrians working is to be reached, it would require a large-scale recruitment drive among Syrian women. However, numerous sources reported that they were concerned about the lengthy journeys required, the lack of childcare facilities in the factories, and the prospect of mingling with unknown men in a large working environment, and therefore they did not pursue these work opportunities. 92 An equally, if not more, significant barrier to the successful implementation of this scheme has been the reluctance of factories operating in the zones to employ Syrians. At the heart of their reluctance, and of the wider dispute among different stakeholders, is the categorization of Syrians in the workforce. Currently, garment factories are allowed to employ migrant workers up to 75 percent of their workforce, while the remaining 25 percent should be Jordanians. This high proportion of migrant labor is of crucial importance to the functioning of the sector. Throughout the negotiation process, JGATE therefore has been demanding that the employment of Syrians should be outside the current migrant labor quota, and should be seen as a 'humanitarian measure.' 93 In other words, employers wanted to be able to replace their Jordanian labor with Syrians. Many factories already perceive themselves to be overly burdened by their obligations to employ a minimum of 25 per cent Jordanian citizens. Jordanians work shorter days than migrant workers (typically 8 hours rather than 10 or 11 per day). They are seen as a drain on the factories' productivity because they are only willing to do the easiest jobs and are more expensive to employ. 94 They therefore endanger the dormitory-based labor control regime described above, which has been key to the success of the garment sector.
In the imaginative framework used by many employers, Syrians, like Jordanians, do not fit into this regime. They are perceived as less productive, skilled and trained than the (potential) pool of South Asian labor normally recruited to work in the sector. 95 As the representative of JGATE explained, those Syrians who attended the job fairs 'did not meet the factories' expectations.' 96 The demand to replace South Asian migrant workers with Syrian refugees thus appeared impossible. As JGATE stressed in a communiqué, 'migrant employees are the CORE of this sector due to their higher efficiency and productivity, skill levels, willingness to work overtime, lower turnover rate. The migrant employees … cannot be REPLACED by Jordanians nor Syrians.' 97 These disputes betray, as discussed above, that the Jordanian labor market is highly segmented by nationality, both between and within sectors. The substitution of Syrians for South Asian migrant workers would require a concurrent belief among employers that the two nationalities are straightforwardly substitutable, a belief that is clearly not currently prevalent in the garment sector. Whether appropriate or not, these entrenched beliefs make it more difficult to change the regulatory framework in which they exist and operate. As authors like Bauder 98 show, they not only legitimate, but also shape and entrench the segmentation of labor markets, creating a boundary between relatively well-protected non-migrant workers and different groups of much more precarious and exploitable migrant workers, who are assigned a place in the labor market through ascribed corporeal, cultural or skills-related differences. 99 Although historically and contextually contingent, these attributions tend to solidify and thereby complicate attempts to enact changes quickly. In the garment sector, these obstacles of perception combine with practical and political considerations to shape the labor force composition. In response to the GoJ's requirement that 25 percent of factory employees be Jordanian, garment factories have opened 16 'satellite units' across the country, mostly in poor, rural areas, and with a workforce that is almost exclusively Jordanian. While these satellite units often do not even break even, and are subsidized by the GoJ, they have made it easier for Jordanians, especially women, to work closer to home, in a more familiar environment. Thereby, they have effectively increased Jordanian employment in the sector. 100 In principle, the same scheme also could be used to facilitate Syrian employment, and suggestions to this effect have been made, yet there have been no concrete moves to implement them. 101 Employing Syrians in satellite factories is a politically tricky suggestion for the GoJ. Despite creative attempts to reinterpret the quota, for example by declaring Syrians to be an 'additional' component of the labor force, rather than simply 'migrant workers,' government representatives stress that Syrian employment must not eat into the quota for Jordanians. 102 Jordanian unemployment is persistently and stubbornly high, and has risen since the start of the Syria crisis, from 11.4 percent in early 2012 to almost 16 percent by late 2016. 103 Not only that, but the government assured its citizenry that it was going to the London conference to put Jordan first, and that providing jobs for Syrians would not be at the expense of jobs for Jordanians. 104 In contrast to the priorities of the GoJ, employers have very little, if indeed any, desire to integrate Syrians into their workforce-as Dina Khayyat, then Chairperson of JGATE, stated clearly: 'We are happy with what we have.' 105 According to an official in the Ministry of Planning's Project Management Unit (PMU), these sentiments extend well beyond the garment sector, with many employers very reluctant to swap their current migrant workers for Syrians. 106 In the garment sector, this has meant that by the end of 2016, only 30 Syrians had been employed through the scheme. 107 Yet at least on paper, the employers are still willing to continue the scheme. 108 As Ms Khayyat explained, it is important for them to maintain their 'social solidarity with Jordan.' 109 This solidarity is arguably not just an altruistic sentiment: not only has the garment sector received positive publicity because of its commitment to employ Syrians, 110 but recent developments have opened up a large range of potentially useful and profitable connections with donors, European governments and markets, and UNHCR, which possesses a database of information about, and can provide access to, a highly exploitable labor force (in spite of the fact that their skill level is currently judged as low). Yet so far, two contradictory visions of who should and could conceivably be substituted by Syrians-South Asian labor migrants or Jordanian workers, combined with the (perceived) absence of decent working conditions in the sector, have prevented the scheme from being anything more than a blueprint.

Restructuring the (Informal) Labor Market
Given the difficulties of employing Syrians in a zonal development model, the broader labor market outside of the SEZs represents perhaps the only opportunity to get even close to 200,000 work permits. The two main ways through which this could be achieved are the formalization of Syrian workers currently working informally and the substitution of other migrant workers in favor of Syrians. A variety of initiatives pursuing one or both of these two strategies have been established. Yet as explored above, Jordan's informalized labor market relies, to a significant degree, on illegal migrant labor and the non-enforcement or selective enforcement of labor regulations. We argue that the attempts to increase Syrians' formal labor market participation do not sufficiently take these dynamics into account. Formalization or substitution strategies run into vested interests and political sensitivities that contribute to maintaining this informal system that relies on both Syrian and Egyptian migrant labor. While they may increase the security of those Syrians who get permits through these schemes, they may well leave large parts of both the Syrian and Egyptian populations in Jordan as precarious and vulnerable as they have been before.
The informalization and illegalization of the labor force, as well as the non-enforcement of labor regulations, are crucial when seeking to understand the status of Syrians in Jordan and current efforts to change it. Although Syrians' labor market participation was largely illegalized until early 2016, many have worked informally since coming to Jordan. According to the ILO and Fafo, Syrians living outside of refugee camps in Jordan primarily work in construction, wholesale, retail and repair trades, agriculture and manufacturing. These four areas together were believed to employ around 80 percent of Syrians working in Jordan prior to the Jordan Compact, 111 and are also areas in which, for all nationalities, there are high degrees of informality. 112 With the Jordan Compact in place, many actors involved in its implementation hoped that these informal jobs now simply could be converted into formal ones. Besides filling a large part of the stated work permit target, formalization also was expected to bring improved social protection, wages and working conditions for Syrians, and to increase taxation and 110 Such positive publicity is important, given the bad reputation the sector earned in Jordan, particularly UNDP et al.,Informal Sector, other forms of revenue for the GoJ. 113 Actors like UNHCR initially appeared keen to demonstrate how the formalization of labor could happen 'organically.' 114 In practice, however, many vested interests have worked the other way. While every fee waiver announcement, or potential end of the grace period (before its renewal was announced), led to a spike in permits issued, overall numbers are still considerably smaller than hoped for, and the number of new issues has slowed down significantly since fall 2016, making an 'organic' evolution unlikely. 115 It is difficult to see, for example, what incentives there are for employers to seek work permits for their employees on a large enough scale to come close to the ambitious targets that have been set. As is the case within the development zones, employers lack a clear financial incentive to employ Syrians formally. Even where employers take advantage of the temporary absence of work permit fees, workforce formalization still typically will increase their costs, because in many sectors they now will be formally liable to pay social security contributions for their registered employees. 116 Furthermore, while work permits were issued to Syrians without fees for most of 2016 and for 2017, they are renewable annually, so future renewals may well incur fees.
Given the number of Jordanians, Egyptians, Syrians and other nationalities who work informally in the Jordanian economy, 117 it is implausible to envisage that companies' workforces will in the future consist of Syrians with work permits alongside informally employed Jordanians, and Egyptians without permits or without the correct permits for their sector. Informal labor is too pervasive a feature of Jordan's economy to change by decree. Many of the small businesses Syrians work in do not even have a valid license. As a report by the Jordanian NGO ARDD points out: 'To achieve [a regularization of Syrians' status] it will be necessary for entities to become legal, which may present as an insurmountable challenge for small businesses. Unofficial entities may not see it within their interest to regularize Syrian labor.' 118 It is thus likely that most of the work permits given in retail, wholesale and trade to date-the sector that makes up most small businesses in Jordan-come from entities that have a formalized status already, and that the further issuing of work permits will be limited by the informality of both businesses themselves and their non-Syrian workforce.
It is also unlikely that threats of punishment will eliminate Syrian informal labor. Fines are the main potential penalty for employers who hire staff without work permits, if they are caught by the Ministry of Labor's inspectors. Yet it is not clear that these fines are an adequate and effective incentive not to employ Syrians without permits. Throughout Jordan stories abound of employers who, with the right connections in the Ministry, or in exchange for the right price, can find out when labor inspectors are coming with enough notice to send their Syrian employees away beforehand. 119 Therefore this symbolic threat does not actually incentivize formalization on a large scale. What it does though is make Syrians feel more insecure, thereby potentially increasing their exploitability. As discussed earlier, the selective enforcement of labor regulations is a central contributing factor to an informalized workforce. Events like symbolic round-ups of migrant workers simultaneously make an illegalized foreign labor force and its 'normal' economic integration in the country visible and render that population more precarious. By emphasizing the boundary between the national and the non-national, they remind migrant workers of their precarious status in a country. 120 Some of these structural difficulties gradually have become clear to those invested in the scheme. Accordingly, many efforts have been made to redesign the employment conditions in specific sectors such that they could facilitate a formalization process, with mixed results. Agriculture has witnessed the biggest strides to date. One key difficulty in this sector is that the work undertaken by Syrian labor is typically casual, varying from day to day, and season to season, to meet the needs of employers. This disincentives individual employers from applying for a work permit, which under Jordanian law is predicated on a year-long job contract. This was tackled by a June 2016 regulation that allowed Syrians to apply for work permits through agricultural cooperatives, enabling them to work for more than one employer. Largely through this mechanism, agriculture has been the sector that has distributed the largest number of work permits, totaling over 16,000 by May 2017. 121 This has been further helped by the fact that the agricultural sector is exempt from the requirement for social security registration. Therefore, it has come to be seen as a comparably easy path to a work permit, and many Syrians with agricultural permits actually work in other sectors, without encountering many problems during inspections. 122 It remains to be seen, however, whether this system will continue to work if/when work permit fees are re-introduced.
Attempts to replicate this model in the construction sector, however, have proved challenging, because of workers' desires not to be bound to one employer, and because of the absence of an organization to play a role parallel to that of the agricultural cooperatives. While many Syrians therefore continued to work in the sector informally, some reportedly resorted to brokers acting as stand-in employers, a tactic common among Egyptian migrant workers in Jordan. Such brokers often demand fees for their services, ranging from 100 to 400 Jordanian dinars (about $140 to $560), which prevent Syrians without means from accessing this informal system. 123 In the spring of 2017, the MoL signed a Memorandum of Understanding (MoU) with the Construction Contractors Association that enabled the association to issue work permits and for Syrians with such permits to move between employers. The MoU also waived the quotas for hiring Jordanian labor in the construction sector. Although there was an initial spike in work permits in construction after the signing of the MoU, it is too early to tell the full extent to which these arrangements will increase the number of work permits for Syrians in the sector. 124 The manufacturing sector also has witnessed a significant degree of formalization, with just under 10,000 work permits issued by May 2017. 125 This is arguably because, unlike agriculture or construction, manufacturing jobs are more likely to resemble permanent, fulltime jobs, even when being done informally. 126 The work permits granted to Syrians in this sector often formalized these pre-existing working arrangements. 127 However, while roughly 4,500 Syrians were granted work permits in manufacturing from March-July 2016, 128 it took almost a year for this number to be replicated. 129 This suggests that the manufacturing route for increasing work permit numbers may be approaching a limit, and it indicates a broader trend across the sectors: the issuing of new permits is stagnating, and there is a perception among many consulted stakeholders that a ceiling is about to be reached. 130 The relatively low-hanging fruit have been picked, but deeper constraints relating to widespread informality and the lack of enforcement of labor law remain relatively untouched.
While some of the movement in these sectors is due to these measures, the decisive reason why many Syrians have received work permits is they paid their employers, or brokers, to apply for one on their behalf, or they covered associated fees, as well as social security contributions, themselves. 131 In order to make Jordan's policies work in their favor, Syrians are thus shouldering the financial burdens of the new approach. This represents an important continuity with previous policies that left Syrians effectively responsible for paying whatever is needed in order to resolve legal status issues and get access to services in Jordan. 132 While those with better resources have been able to buy themselves some temporary security, many others have been left behind. That some formalization has occurred does nothing to lessen the conditions of vulnerability that have been created for this latter group.
The same can be argued for attempts to substitute another informalized migrant labor force-Egyptians-with Syrians. These efforts may well make Egyptians more vulnerable to exploitation and criminalization. The GoJ estimates that an estimated 520,000 Egyptians live in Jordan, and approximately 320,000 of them do not currently hold work permits. Many who hold work permits now work in a different sector than the one for which the permit originally was issued. 133 No new permits currently are being issued for Egyptians; while existing permits are eligible for renewal, this comes with a new flat cost of 500 Jordanian dinars, which represents a significant fee increase for sectors such as agriculture. There are also reports of an increasing number of deportations of Egyptians found working without a permit, a trend that has created shortages that are felt acutely in sectors like agriculture or construction. 134 This indicates a degree of rebalancing, a continuation of previous policies of selective formalization and informalization of Egyptian labor, toward a (temporarily) more restrictive approach, which may well push many Egyptians working in Jordan into an informalized situation.
Yet these new measures do not signify a wholesale policy change. Jordan's policies toward Egyptians working in the Kingdom have political, diplomatic, and economic repercussions for both states. Egyptian workers in Jordan represent an important source of remittances for the Egyptian economy, and therefore have been a source of political leverage for Jordan. 135 Jordan, however, heavily depends on the availability of cheap gas from Egypt, as Jordan's electricity bill makes up a substantial amount of its growing debt. Any moves to decrease substantially the number of Egyptians (formally) working in Jordan in favor of Syrians could damage its multifaceted bilateral relations with Egypt. 136 The reduction of the size of the Egyptian workforce, ostensibly, has been a goal of the GoJ for many years, but little has been done practically to achieve this aim. 137 Past attempts to deport Egyptians were stopped because they caused diplomatic upheavals, 138 and in 2016/17 countervailing measures, such as a grace period for Egyptians to formalize their status, have co-existed with, and to some extent balanced, policies that implied an attempt to reduce the size of the (formal) Egyptian labor force. 139 Therefore the GoJ's continuation of the selective formalization and informalization of Egyptian labor enables it to balance competing interests, while also allowing it domestically to maintain a public commitment to creating jobs for Jordanians through the substitution of migrant workers.
Besides diplomatic issues, substitution attempts are also inhibited by the prevailing dynamics and imaginaries of the informal labor market. As previously discussed in the case of the garment industry, there are widespread beliefs about the suitability of different nationalities for different labor market activities. While, in abstract economic labor market supply terms, Egyptian construction workers could be substituted with Syrian construction workers, it is 'known' in Jordan that while Egyptians can be relied upon to do long shifts of hard manual labor, Syrians are more suited to decoration and interior design. 140 While some substitution has occurred, Syrians' work in the sector-often mediated by Egyptian brokers and foremen-is typically limited to subordinate positions, which may well keep these ideas alive. 141 Given these political and practical intricacies it is very unlikely that the status of either Syrians or Egyptians will be redefined unambiguously or that formalization will happen across the board. Rather, we are witnessing a partial formalization of Syrian labor and a concomitant partial informalization of Egyptian labor. However, once the current donor focus on formalizing Syrian labor lapses, both Syrians and Egyptians may well remain semi-formalized, along with many other nationalities of migrants in the country. In this situation, employers are likely to continue to be able to find Egyptians as well as Syrians to employ with or without work permits. Although the possibility of formalization may benefit some Syrians, this situation may well leave other Syrians, as well as Egyptians, just as precarious and exploitable as before, in some cases even more so.
Despite the limitations of the formalization efforts, those Syrians who potentially may be eligible to receive work permits are very aware, and very supportive, of the changes. 142 Syrians who continue to work informally typically do so under the threat and fear of forced encampment or refoulement if they are caught working without a permit, especially if they are caught more than once. 143 Syrians are therefore eager to obtain work permits, but their ability to do so is limited by the sector they (wish to) work in, the widespread degree of informality in the Jordanian labor market, as well as the regulatory and sector-specific constraints detailed in this article. While helping Jordan reach its short-term work permit goals, current efforts to reform employment conditions do very little, if anything, to tackle the underlying structuring dynamics of the labor market. As a senior official working on the refugee response in Jordan self-critically noted: I think we are not really discussing the structural deficiencies endemic to the Jordanian economy, we are just looking at working inadequately within it. The GoJ and everyone is kind of coming to terms with just how informal the economy is .... [s]o you just keep putting these band-aids on a cut that is super big. But I guess it's good for making sure that Syrians don't get [sent to the camps]. 144

Conclusion
This article has analyzed the initiatives and measures taken since early 2016 that seek to facilitate Syrian refugees' access to the formal labor market and traced the twists and turns of these supposedly sweeping policy shifts to date. We have argued that the plan formally to 'make refugees work' successfully has mobilized a large number of important stakeholders in such a way as to turn it into a policy success, regardless of its actual effects on those supposed to be helped by it. The underlying claim that two birds-humanitarian obligations toward Syrian refugees and Jordan's economic development needs-can be killed with one stone has brought together humanitarian and economic actors, both national and international, governmental and non-governmental. This alliance, and the creation of a globalized 142 Author Interviews with 15 Syrian refugees, Irbid and Karak Governorates, April-June 2016; ILO, Work Permits, p. 33. 143 While it is impossible to obtain exact numbers for how many Syrians have been deported or forcibly encamped, reports and interviews with Syrian refugee families across Jordan confirm that many Syrians personally know people who have been refouled to Syria or forcibly encamped, and many more are aware of others who have faced this fate, demonstrating the dangers of working without a permit. Author Interviews with Syrian refugees, see n. 142; AP (2017) Aid Groups: Jordan Deports Thousands of Syrian Refugees. May 12. Available at: https://www.voanews.com/, accessed July 16, 2017. 144 Author Interview with senior official working on the refugee response, Amman, May 2017. blueprint, we have argued, was made possible by these actors' relative detachment from the realities of, respectively, the Jordanian labor market and the lives and perspectives of Syrian refugees. Looking at the three most prominent strategies attempted to make the Jordan Compact happen-the expansion, substitution and formalization of employment-has revealed various frictions that point to conflicting projects of government and development among the actors and organizations involved. These strategies stand in tension with a political economy and labor market that have evolved in the country over decades, and that have complicated the scheme's realization.
Firstly, the plan to employ refugees in special economic zones and to boost Jordan's economy by attracting investments through trade concessions has proven hard to roll out when confronted by these complexities. Many of the grand visions for increasing foreign investments stand in tension with the rationalities of potential (even ethical) investors, the unwillingness/inability of existing employers to find the requisite numbers of Syrians to employ, and the many remaining impediments for successfully exporting to the EU. The vision of refugee employment in SEZs is also at odds with a perspective that emphasizes not just work for Syrian refugees, but decent work. In those zones that actually do make profits, success often has been bought at the expense of harsh working conditions and low wages, which have deterred Syrian refugees from taking up employment there.
Secondly, the scheme's unfolding has revealed competing visions of Syrians' place in, and the effects of, Jordan's entrenched and highly segmented labor market. We have shown throughout this article that Syrians are located in a gray zone between local and migrant workers in a labor market that draws sharp divisions and distinctions between those two groups, as well as between different nationalities of migrant workers, genders, and along other axes. Attempts to replace migrant workers with Syrians have run up against these pre-existing labor regimes, and Syrians have become the object of unresolved debates about where the boundary between different parts of the labor force lies. Another effect of this segmentation is that, more often than not, the twin goals of bringing Syrians into formal jobs and creating employment for Jordanians (and, beyond that, fostering Jordan's economic and infrastructural development) are actually conflictual rather than complementary. While Syrian formal labor plausibly can benefit Jordan's economy in a number of ways, it has become relatively clear that Jordanian (un)employment is not one of them. Whereas some Jordanians benefit from employing migrant workers (and/or Syrians), this tends not to lead to improvements in Jordanian job prospects. The underlying mismatch between the jobs the Jordanian economy can offer, and the skills and expectations of Jordanians, cannot be solved through the proposed measures.
Thirdly, the donor-driven push to formalize (Syrian) refugee labor stands in tension with the long-established dynamics of an informalized labor market, and with the interests of the actors for whom this informalized or semi-formalized system works. Various efforts to facilitate this process have revealed that, while some jobs can be formalized through changed incentives and regulations, others defy such attempts, demonstrating how entrenched informal arrangements are in sectors like construction or retail, wholesale and trade. The non-enforcement of labor legislation has constituted a compromise in the past, and likely will continue to do so, only in a rebalanced version that formalizes some Syrian employment while keeping other Syrians, as well as many Egyptians and also Jordanians, in precarious informalized positions.
Frictions between the different projects of government and development do not just occur between 'humanitarian' and 'economic' actors. These groups, and their projects, are not so monolithic, and as our analysis shows their goals can and do overlap. This is not least due to shared or similar organizational logics. The number of work permits issued to Syrians quickly became, in part due to strong pressure from some donor agencies, a core indicator of the scheme's success, and has been adopted as such by the GoJ and other stakeholders. Yet the joint attempts to increase the number of permits as quickly as possible have favored quick fixes at the expense of more lasting solutions. This effectively has left behind many Syrians who do not fulfill the criteria, and also has left other refugees on the sidelines.
Our analysis of the Jordanian context has important implications for the future of the global governmental project to 'make refugees work.' While the labor market in Ethiopia (or to wherever the scheme next is exported) may or may not be similar to Jordan, it will be more complex than any blueprint could account for, and will contain its own stratifications, hierarchies, and exclusions. Due to their supposed universalism, such 'one-size-fits-all' approaches, which have gained popularity in humanitarian circles in recent years, often overlook or exclude those potential 'beneficiaries' who do not easily fit into predetermined frameworks. They also ignore the specific histories and socioeconomic contexts of the places they seek to reshape and 'improve,' yet these factors shape what is possible, and they cannot be wished, decreed or imagined away.
Within Jordan, these factors have played a large role in ensuring that the scheme has not gone according to plan. This arguably should be expected, as it is the norm rather than the exception that supposedly universal ideas take on a reality of their own in a specific context. The progress that has been made largely has been on the sidelines of the high-profile initiatives that were announced with much fanfare. The greatest numbers of work permits have been issued through processes of formalization, rather than in the factories of Jordan's Special Economic Zones. It remains uncertain how close to 200,000 work permits the Compact will reach through these projects, and even if the target were reached, whether these jobs would be decent and sustainable. What is certain is that Syrian refugees are very weakly integrated into the networks of humanitarian actors, donors, and governmental agencies that represent the trustees of this 'development opportunity.' As the objects of development, they do not have, and are very unlikely to develop, a strong voice in the worlds of those who negotiate their socioeconomic fate. In the face of attempts to 'make refugees work,' Syrians have been left with two choices: either to refuse to entertain the paltry options offered to them, or to try to turn whatever possibilities that emerge from the scheme to their advantage.