Asylum in Austere Times: Instability, Privatization and Experimentation within the UK Asylum Dispersal System

In 2010, the UK government passed contracts for the provision of dispersal accommodation and reception services for asylum seekers to three private providers. This article explores the causes and consequences of this process, arguing that dispersal has been reshaped through a confluence of ‘austerity urbanism’ and privatization. The article draws on fieldwork in four cities (Birmingham, Cardiff, Glasgow and Sunderland), including interviews with local authority representa-tives, politicians, and asylum and refugee support services. The article highlights the production of instability within asylum dispersal as an effect of austerity and privatization. As a result, we witness a narrative of political neglect, shrinking accountability and the slow recession of support services and expertise. Whilst instability has often been a common facet of asylum policy in the UK, austerity and privatization have meant that a limited concern with the social needs of asylum seekers has been replaced with an increasingly revanchist agenda.


Introduction
In January 2016, the issue of asylum seeker accommodation made national headlines in the UK, following revelations that houses used to accommodate asylum seekers in the city of Middlesbrough were identifiable by their red front doors. The doors had been painted by an accommodation contractor, Jomast, to help identify their properties for the purposes of monitoring, maintenance and inspection. However, such identification also served to expose the inhabitants of these properties to racism, harassment and vandalism. The red doors marked the residences of asylum seekers relocated to Middlesbrough as part of the United Kingdom's dispersal scheme. The outcome of this brief media attention was a Home Office inquiry into the accommodation involved and the immediate repainting of the doors. In this article, I want to explore some of the processes and policies that underlie this incident. In particular, I consider how the reception, accommodation and support of asylum seekers in urban Britain have changed through privatization. Moving beyond Middlesbrough, the article explores how the system of dispersal and accommodation for asylum seekers in the United Kingdom has been reshaped in response to demands for austerity that are both nationally articulated and locally enacted. Through considering the effects of privatization, austerity and policy instability on the ground, I argue that we see a model of accommodation that has failed to prioritize the needs of its clients and that has often ignored the lessons of the past. As I argue, the exclusionary marking of red front doors is emblematic of a system built upon a drive for fiscal savings at the expense of human dignity.
In considering the effects of privatization, this article makes two key contributions. First, it brings discussions of asylum support and refugee resettlement into conversation with an emergent body of work that examines the politics of austerity. I argue that neoliberal frames of fiscal constraint, productivity and competition increasingly underpin dispersal as both a social policy of accommodating and regulating asylum seekers (Darling 2011(Darling , 2016a, and an economic site of profit-making and efficiency savings (Darling 2016b). Second, it seeks to redress the focus on the nation state that has often dominated refugee studies. This dominant focus has meant that analysing the influence of domestic welfare regimes, fiscal policy and local political negotiations has been less widespread than a concern with international questions of refugee resettlement and protection. In doing so, the article argues for an engagement with the workings of dispersal policy as it is practised in relations between urban authorities, the Home Office, thirdsector organizations and asylum seekers themselves. Considering these tensions and negotiations highlights how government policy is interpreted, resisted and remade in practice.
In making these arguments, the article proceeds in three sections. In the first of these, I outline current work on dispersal within refugee studies and geography, and consider recent debates around austerity urbanism as a means of framing the contemporary dispersal system in the United Kingdom. The next section draws on empirical material from a study of four United Kingdom cities to consider how the effects of privatizing accommodation are shaped by austerity, and how these processes produce an increasingly unstable dispersal system. The third section provides a longer historical view on dispersal and its local politics, to suggest that aspects of instability and fragmentation have remained essential components of this policy and that a turn to 'austerity' and its socially revanchist agenda is not new for asylum seekers in Britain. I begin, though, by outlining recent discussions of dispersal and austerity.

Jonathan Darling
Downloaded from https://academic.oup.com/jrs/article-abstract/29/4/483/2741223 by guest on 28 July 2018 engage with asylum seekers and as an experience shaped by urban contexts of demography, social care and political support (Hynes 2009;Netto 2011b;Darling 2016a). Within these discussions, Netto's work in Glasgow finds that [p]roblems such as racial harassment and abuse were more extreme in areas with little previous history of accommodating people from other ethnic groups, indicating the need to incorporate a temporal dimension in considering conducive locations for refugee settlement (Netto 2011a: 126).
Findings from initial analyses of the United Kingdom's dispersal programme support this assertion, with Zetter et al.'s (2003) work for the Home Office concluding that shortcomings in information sharing were undermining the effectiveness of dispersal, producing an uneven geography of support across the country. Phillips (2006: 544) highlights the fact that, in the initial stages of dispersal, local authorities were 'key players' due to their 'partnership role in the regional consortia' that provided accommodation. Yet the reliance on local authorities to offer accommodation resulted in dispersal being focused in areas of low-cost and hard-to-let social housing. Often, as Phillips (2006) notes, this meant asylum dispersals to areas of existing social deprivation, economic stagnation and high unemployment. Importantly, the housing market-led geography of dispersal has been argued to both isolate and marginalize asylum seekers through their association with areas of territorial stigmatization (Phillimore and Goodson 2006;Spicer 2008), and has been cited as a source of tension in local contexts often unprepared for the arrival of asylum seekers (Finney and Robinson 2008).
Discussions of dispersal within the United Kingdom have thus examined the liminality experienced by asylum seekers through the imposition of mobility and fixity that this policy implies (Hynes 2009;Squire 2009;Darling 2011), and have also considered the local politics of dispersal as a policy that places asylum seekers in areas of existing social deprivation and often poorquality housing stock (Phillips 2006). Today, both of these effects are being reshaped in a landscape of fiscal austerity and widespread reductions in support for asylum services and third-sector organizations, and it is to this changing landscape that I turn in this article, to consider how dispersal changes as it comes to be practised in an increasingly austere context.
A focus on austerity has been growing across the social sciences, with the term referring to government policies that seek to reduce budget deficits and spending cuts by reducing or freezing labour costs, tax increases, privatization, [and] a reconfiguring of public services and the welfare state (Donald et al. 2014: 5).
Austerity policies are not new as means of governing; rather, they have a history of being evoked as a way to frame moments of 'crises' and to implement seemingly necessary 'emergency' responses (Blyth 2013;Donald et al. 2014;Stanley 2014). In this sense, austerity might be seen to operate as a 486 Jonathan Darling specifically neoliberal governmentality-part of a wider set of narratives, impulses and policies that contextually and contingently reproduce the primacy of economic modes of calculation, prioritization and market ideology as guiding social and political life (Peck and Tickell 2002;Brown 2006;Dean 2009). Crucially, though, discussions of the recent 'age of austerity' (Stanley 2014) have focused on the ways in which austerity is produced at the local level (Lowndes and Pratchett 2012), and through cities in particular (Donald et al. 2014). A turn to examining 'austerity urbanism' (Peck 2012) has thus drawn attention to how models of fiscal discipline, constraint and retraction have been transferred to municipal authorities, alongside expectations to 'do less with less'. In this context, Tonkiss (2013: 322) argues that an urbanization 'of (and by) enclosure is a keynote of austerity politics', enacted through the privatization of 'many spaces and services that used to be publicly owned and more open and/or affordable' (Mayer 2013: 9). The urban effects of austerity are thus argued to be in the closure or privatization of public space, social services and support structures.
Importantly, as Beatty and Fothergill (2014) highlight, in the United Kingdom, this process of 'enclosure' and welfare retrenchment is geographically uneven, with many of the country's poorest areas hit the hardest through an austerity agenda that 'rolls back' support for the unemployed and the economically marginalized (see Hamnett 2014). Whilst austerity policies draw on a concern with managing 'crisis', they have also been framed as a means of reworking the state so as to enhance its economic potential, to encourage efficiency and to reproach those seen as economically unproductive. Austerity and its effects have thus been argued to produce 'rivalries rather than building solidarities amongst those who ''have little''' (Hoggett et al. 2013: 567), as socially marginalized groups are placed in competition for scarce resources. In the context of asylum dispersals, the disproportionate impacts of austerity on areas of existing deprivation, combined with a discursive focus on efficiency and competition in the allocation of what resources remain, means that dispersed asylum seekers are now often located in areas at the centre of the 'age of austerity' and its most divisive impacts. It is, therefore, vital to consider not just how asylum dispersal is practised in varied local contexts, but also how such policies have been reshaped by their proximity to a politics of austerity and divisive competition.
In making this argument, the article draws on fieldwork from a larger project examining the relationship between asylum seekers, dispersal and urban governance in the United Kingdom. Focused on four cities (Birmingham, Cardiff, Glasgow and Sunderland), this project involved 106 interviews with stakeholders in the asylum accommodation and support sector, including local authority representatives, councillors, service providers, support organizations, third-sector groups and asylum seekers. This article examines interviews with those within the 'asylum sector' of each city, namely those involved in either local authority, third-sector or voluntary service provision, in order to trace their experiences of asylum policy and how this has Asylum in Austere Times 487 shifted throughout the period of dispersal (2000-present). In what follows, I focus on how the privatization of asylum accommodation in response to austerity has led to an increasingly unstable dispersal system, in which support organizations and local authorities are readily side-lined and their expertise lost. Yet, I also contextualize this shift through considering the longer history of dispersal, and through narratives of austerity, loss and political indifference that pre-date privatization. In doing so, I make the case that austerity, and its urban manifestations, matters not just because it shapes national policies and priorities in terms of support for asylum seekers. Rather, it matters because austerity moulds the local context into which asylum seekers are dispersed. The intersections between austerity and dispersal are thus multiple and multi-scalar and, in the remainder of this article, I shall begin the work of articulating a few of these connections.

COMPASS and Local Neoliberalisms
In examining these questions, this article explores the effect of transferring contracts for the provision of asylum seeker accommodation from a mixture of consortia of local authorities, social housing associations and private providers to just three private contractors. The multinational security services company G4S, the international services company Serco and the accommodation partnership Clearel. These contracts, collectively known as COMPASS (Commercial and Operating Managers Procuring Asylum Support) marked a significant shift in the landscape of asylum support in the United Kingdom.
Whilst the COMPASS contracts were signed in March 2012, the dispersal system itself has been in force since 2000. It was initially proposed in the 1998 Fairer, Faster, Firmer, White Paper in the following terms: The administration of a new support scheme for asylum seekers, entirely separate from social security benefits, will require new national machinery to plan and co-ordinate provision, obtaining information from around the country and purchasing places either directly or by contracting with local agencies. Asylum seekers would be expected to take what was available, and would not be able to pick and choose where they were accommodated, but where possible placements would take account of the value of linking to existing communities and the support of voluntary and community groups. . ..This nationwide approach will help to relieve the burden on provision in London, where the majority of asylum seekers are currently concentrated (Home Office 1998: 8.22). Imagined as a means to 'relieve the burden on provision in London', dispersal created a new geography of asylum across the United Kingdom. As asylum seekers were relocated to social housing in towns and cities across the midlands and north of England, Scotland and Wales, the notion of 'spreading the burden' of provision became commonplace rhetoric for describing the accommodation process (Darling 2016b). Throughout the 2000s, the dispersal policy faced a series of challenges, from concerns over 'community cohesion' and resources, to questions of integration and the need for longer-term planning for the establishment of 'new communities', once individuals had received refugee status (Robinson 2003;Zetter et al. 2003;Griffiths et al. 2005). As Andersson (2003) and Robinson (2003) both note, dispersal within the United Kingdom was heavily criticized from three perspectives: first, for the imposed mobility and regulatory oversight it entailed for asylum seekers, producing an exclusionary system of control (Squire 2009;Webber 2012); second, for its entanglement with a newly restrictive regime of welfare entitlements and conditionality separated from the mainstream benefits system (Sales 2002;Gill 2015); and finally for the delays in sourcing adequate accommodate, the lack of communication with local authorities, and the lack of support structures in place, all of which reflected the absence of a coherent approach to, and effective management of, this new process (Andersson 2003). As Webber (2012) notes, it was this confluence of factors that led dispersal to the United Kingdom to be more heavily criticized than in other European countries.
Yet, the changes brought in through COMPASS were not framed as responses to these challenges. Rather, they were positioned in the context of fiscal austerity and a desire to cut the budget of the then United Kingdom Border Agency (UKBA). In their investigation into the procurement of the COMPASS contracts, the House of Commons Public Accounts Committee found that the cost of providing accommodation in 2011-12 was £150 million. In March 2012 the Department decided to introduce a new delivery model involving fewer and bigger housing providers than under previous contracts . . . . The Department, through the introduction of these new contractual arrangements, aims to save around £140 million over seven years (House of Commons Public Accounts Committee 2014: 4).
Similarly, in describing the rationale behind COMPASS, a UKBA spokesperson stated that [c]ontracts for asylum services have been awarded to providers that demonstrated that they could meet our high standards of support and ensure the welfare of individuals. As well as providing a quality service, these contracts will help to deliver estimated savings of £150m over seven years (UKBA, cited in Hegarty 2012).
Whilst a range of government departments were under pressure to reduce their budgets, asylum accommodation was targeted within UKBA as a relatively easy area for austerity measures. Asylum seekers are unable to vote, unable to work, are often demonized in the popular press and are constructed as figures whose entitlement to any form of support is readily questioned (Sales 2002). As such, they represent a group unlikely to garner a high degree of public sympathy or political capital when faced with the impact of austerity, not least at a time of wider public-sector cuts.

Asylum in Austere Times 489
The driving force behind COMPASS was therefore a desire to contribute to the government programme of deficit reduction and austerity. The importance of this connection to austerity is two-fold. First, austerity has been argued to have played a central role in 'amplifying' and 'extending' the reach of neoliberal governmentality into new domains of public life (Mayer 2013). As Newman (2014: 3292) states, within the orthodoxy of austerity, 'we must all change the ways we live and work according to an economic calculus of debt reduction and financial stringency', for fear of 'the market as an invisible but potent and demanding force' (ibid.). In this context, the financial crisis of 2008 and its subsequent fiscal aftermath presented 'an opportunity to implement more severe cuts than ever seemed justifiable before, and to push through more privatization' (Mayer 2013: 10), whilst maintaining, and even enhancing, the importance of an economic rationale at the heart of various domains of public policy.
Second, the impact of austerity is localized to municipalities, such that understandings and experiences of austerity are refracted through the local state and its services (Blanco et al. 2014). Thus, whilst Brenner and Theodore (2002: 349) argue that cities are 'strategically crucial geographical arenas' for the articulation of neoliberal policies, Peck (2012) suggests the emergence of 'austerity urbanism' has led to 'the cumulative incapacitation of the state', as local authorities must 'do less with less' (Peck 2012: 630, 647). In contemporary asylum accommodation, this incapacitation is vitally important, as it presents a geographically uneven range of responses to the privatization of housing. For Peck, the public-sector cuts most commonly associated with austerity serves as a prelude to political instability and institutional degradation, to crisis management, to backfilling efforts on the part of non-profit or business interests, and in some cases to de facto abandonment (Peck 2012: 629).
These effects might all be considered as impacts of a drive to reduce publicsector support and provision through local authorities. In the context of COMPASS, it is worth considering how austerity has shaped accommodation and policy on the ground. To do so, I want to consider how COMPASS might be seen to have produced moments of instability within dispersal.

Austerity and Instability
In considering the effects of COMPASS, I want to focus on Peck's (2012: 629) claim that austerity provides 'a prelude to political instability'. Instability was a common theme expressed by those working within local authorities, the third-sector and those supporting asylum seekers throughout this research. For example, Laura, who worked for an asylum support organization across the north of England, argued that in terms of COMPASS you definitely lost council officers who were trained up in asylum issues. It's undeniable really. . .. I think generally across the council services there is a lack of knowledge about asylum but that's probably because they don't actually have any mandate to work with asylum seekers anymore (Laura, interview, 2014).
Similarly, Farah, a representative from a third-sector organization in Birmingham, noted that when the local authorities were in contract, they obviously had a vested interest in it and the money from the contract would then go back into support services in that area, and what's happened now is they're not in contract, they have had to cut money to those voluntary sector agencies and those support services, demand is going up and up and up, you've got a private organisation that doesn't appear to always be listening, and that money isn't being reinvested into the community. So from the providers' point of view I do see where they're coming from, they're holding a contract directly with the Home Office and they don't think they need to do other things. . .But it is very different, and public sector thinking and private organisation thinking is very different (Farah, interview, 2014).
In both of these accounts, Laura and Farah highlight the distinctions of investment, knowledge and capacity that marked the transition from a mixed provision of accommodation between local authorities and some private housing associations, to the privatization of accommodation under COMPASS. The shift associated with COMPASS not only gave contracts to private contractors; it necessarily also meant the end of public provision of accommodation and support services through local authorities. As both Laura and Farah argue, this has reshaped the relationship local authorities have with asylum seekers as a constituent group. With no obligations to provide accommodation, and no financial support to maintain asylum expertise within the local authority, those who formerly worked in the asylum sector for local authorities were either transferred to new roles or sought employment elsewhere.
The effects of this transition were discussed by David, a case worker for a national refugee organization: Some of those staff members may have transferred over to the private sector so actually the local authority has lost its memory because the memory is sat with individuals. . . local authorities are not great in terms of institutional memory as are the Home Office because the staff turnaround is so frequent. . . I've been to Home Office meetings where they've presented something and we've said this is very similar to what you presented four years ago and they're surprised because they think that it's a new idea. . . that happens quite a lot and I think you see the same with local authorities, they start out knowing so much and then a lot of the asylum departments or the specialist teams that were working with the asylum seekers while they were being housed, those individuals have moved on. So, it does feel like a bit of a retrospective step (David, interview, 2013).

Asylum in Austere Times 491
As David recounts, the transition of responsibility for asylum accommodation that came through COMPASS led to a loss of knowledge and expertise within the local authority. This 'institutional memory' was not only of the asylum system and its varied legal complexities and distinctions (Zetter 2007), but also of the local context of dispersal, community relations, and those practices and experiments that had made dispersal function more or less effectively in the past. With such memory lost, the ability to reflect on past mistakes, contextualize new policy approaches and crucially historicize what may, or may not, work within differing local contexts of support, community relations and demographics was dramatically reduced. Memory loss forms part of the wider 'institutional degradation' that Peck (2012) suggests marks urban austerity. As fiscal support for local authorities is removed, so the capacity to retain staff, manage responsibilities and meet statutory obligations is diminished, leading to local authorities that hold knowledge and expertise over an ever-decreasing range of public service requirements (Blanco et al. 2014). Whilst some staff and expertise have been transferred from local authorities and support organizations to private providers, as Farah points out, 'public sector thinking and private organisation thinking is very different' (Farah, interview, 2014), leading to frustrations that the approach to asylum had changed from being service-driven to profit-driven.
The implications of this transition effect both private providers and local authorities. With the movement of former local authority staff to private providers, we see a process of knowledge 'downloading' in action. In each of the four cities studied, respondents cited examples of former colleagues who had transitioned to work for providers such as G4S and Serco, but who had left those organizations within 12 months of that transition. Varying reasons were attributed to this turnover of experienced staff, but the distinction of ethos that Farah notes above was highlighted as significant. Individuals felt that they could 'make a difference' to the delivery of these newly privatized services, and often grew frustrated when it became clear that the public service ethos they had worked to previously could not be so easily translated. The inability to translate this ethos was put down to the fundamental importance of providing only those services accounted for in the COMPASS contracts (see Darling (2016b) for a fuller discussion of the role of contractualism), as a means to ensure the profitability of such contracts. This meant that any additional forms of discretionary support or attempts to go beyond strictly defined parameters of service delivery were to be avoided, serving, as Gill (2015) argues, to reinforce the moral distance between bureaucratic functionaries and asylum seekers. This is not to suggest that previous forms of local authority service delivery were orientated around forms of care and compassion, but rather that the discretionary space in which moments of care could occur was more readily available to service providers less tied to the delivery of a profitable contract.
In this context of 'downloading', private providers were able to utilize the knowledge and expertise of specialists for a brief period of time so as to help establish themselves in the field of asylum accommodation. Whilst it might be argued that this process does ensure some level of continuity and means that expertise is not entirely lost, it has two key consequences. First, it means that knowledge is now mobilized for the purposes of a profit-driven adherence to contract delivery, rather than to the public service ethos associated with local authorities. And, second, as David points out, it means a crucial memory loss from local authorities. In a context of tensions between local authorities, the Home Office and private accommodation providers over asylum housing and support, this loss of expertise and institutional memory to another key actor in the dispersal system places local authorities further on the back foot in negotiating such tensions (Darling 2016b). This form of 'downloading' knowledge is not unique to the asylum-dispersal field under COMPASS. Rather, as varying public services have been privatized or co-opted by private interests, the accumulation of institutional knowledge, networks and understanding has been a key resource for private providers in social care, housing and probation (Clarke 2004;Williams et al. 2012;Fitzgibbon and Lea 2014). The types of knowledge and expertise gained here, even for a short period of time, are critical to the evolving neoliberalisation of dispersal accommodation (see Darling 2016b). That is to say, the privatization of accommodation that emerges with austerity is not a fixed expression of a neoliberal orthodoxy; rather, it illustrates how neoliberal approaches to service delivery are adaptive and incorporate new forms of knowledge and expertise so as to enhance their effectiveness (Peck and Tickell 2002). The 'downloading' of local authority expertise thus offers an opportunity to more effectively govern and manage the asylum-dispersal system in the interests of profit.
The second implication of this loss of institutional knowledge and memory is for local authorities and their capacity to manage the outcomes of asylum dispersal. For, whilst expertise is lost, local authorities still retain responsibilities for community cohesion and the social integration of refugees. In this context, the loss of institutional knowledge and memory of asylum that COMPASS affects is significant. As a range of analyses of the early period of dispersal in the United Kingdom highlight, dispersal is successful most readily where existing local communities are assured that their interests are being protected, and that information is sensitively and clearly distributed in advance of dispersal (Robinson 2003;Zetter et al. 2003;Netto 2011b). Community preparation work and information sharing of this kind is not an easy or quick task; it demands understanding and communication of the asylum process, and a contextual grasp of local politics, communities and sensitivities. Where such work has been most successful, in Glasgow for example, changes in attitudes towards asylum seekers and refugees have been effected by committed and long-standing work from integration networks, local authority teams and third-sector groups working together (Wren 2007;Bowes et al. 2009). In Glasgow, such work was required after an initial Asylum in Austere Times 493 period of poor communication, racial harassment and violence towards asylum seekers (see Hickman et al. 2008), and it is precisely this form of incrementally achieved knowledge and expertise that is lost through the privatization of asylum accommodation.
Furthermore, the impetus to undertake preparation work is undermined through reduced local authority budgets and a political climate where publically supporting asylum seekers is potentially divisive. This confluence of factors is cited by Laura as a reason why community preparation work for asylum dispersals across the north of England is highly geographically uneven: Council budgets are being slammed down to zero, it's hard to persuade councils to take this one as a big issue when there's lots of other things that they feel they should put their money in. Sometimes though. . . they kind of do good things without telling anyone about it because they know it's not a particularly strong local vote winner. . .. It's very mixed. . . if you have the right people working in some councils you're likely to get more awareness. . .. The worst scenario is when you've got no awareness, no council mandate to work with asylum seekers, no asylum teams, and in a place with high dispersal (Laura, interview, 2014).
With COMPASS, the 'council mandate' that Laura references here has largely been removed and, with a loss of institutional memory and knowledge, the awareness and support teams also mentioned here have dwindled in many dispersal areas. Indeed, of the four cities examined in this research, only Glasgow had managed to retain within its local authority any formal support structures for working with asylum seekers. In studying policy discourses, Barbeho¨n and Mu¨nch (2016: 49) find that Glasgow has promoted a narrative of 'fairness' towards asylum seekers that contrasts the image of a 'caring city' with that of a 'cold' and distant 'UK bureaucracy', and retaining some level of support within the local authority might be seen as one means to maintain this image (see also Darling 2013).
The shifting fiscal context of COMPASS had further impacts within dispersal areas, most notably through knock-on effects for third-sector organizations. As Sarah, a policy coordinator from Glasgow, highlighted in relation to COMPASS: one of the big side effects of that has been that it was more. . . there was more money coming into Glasgow per person than there is now and Glasgow City Council provided wraparound services that were factored into that in their calculations. So there was excess money that allowed them to put more resources into English for additional language provision. . .it's to do with the type of provision and the amount of money that's available because what a private company sector would see as profit local authorities were able to reinvest into those additional services and they did do that and there was significant added value to what they provided (Sarah, interview, 2013). In the Glasgow context, when the local authority held contracts to provide accommodation, funding was able, at times, to be used to support wider services and integration work, as also noted by Farah previously. In this way, services such as language training, drop-in support and integration networks were maintained, often through multiple small sources of income. Yet, as Sarah notes, with the privatization of this contract, any money over and above the running costs of accommodation is cycled back into the shareholders of a private contractor, rather than into third-sector organizations or asylum support programmes. COMPASS meant not just a loss of control and capacity for local authorities, but also a concurrent removal of support for third-sector organizations and groups. At the same time, the wider politics of austerity has produced an increasingly challenging funding environment for third-sector organizations, with two effects. First, as Williams et al. note: public austerity is resulting in funding cuts to the voluntary sector-a disinvestment that is undermining the capacity of third-sector organisations to sustain their presence in a landscape of escalating need (Williams et al. 2014(Williams et al. : 2803. Thus, within dispersal cities such as Sunderland that have historically had a small but committed refugee sector of support groups and organizations, funding cuts have led to their closure and a dramatically reduced capacity to provide advice and support to asylum seekers. Second, an austerity-driven focus on funding competition has placed third-sector organizations in '''defensive'' mode, developing strategic and operational solutions for their own survival' (Lowndes and Pratchett 2012: 29-30) and encouraging an acquiescence to neoliberal norms of market efficiency and economic calculation. Where organizations are able to survive funding cuts, a reliance on competitive and insecure funding streams through increasingly rare government grants presents a barrier both to developing longer-term strategies for support and to performing advocacy work that may be critical of government policy (Gill et al. 2014).
The tensions of change noted above are further exacerbated when we consider the model of outsourcing and subcontracting that has emerged through COMPASS. As Warren, an asylum advocate from Birmingham, argues, the centralized contracts provided by the Home Office mask a far more fractured reality: G4S will have the contract and then they'll subcontract to someone, who then may subcontract to someone else. So for the service users and for the agencies trying to help and resolve things, it's impossible to ever find out who's going to take responsibility for something. So the service user will say, I don't know, I haven't got any hot water, and they'll tell that to the guy that comes to see them once a week. He'll say, that's nothing to do with me, you need to speak to G4S. They'll phone the G4S line. They'll say, no, you need to speak to the guy that comes. And it's this endless deferral of responsibility and simple, simple things. . . can go on for weeks and weeks and weeks (Warren, interview, 2014).

Asylum in Austere Times 495
The 'deferral of responsibility' that Warren recounts here was a common concern among asylum advocates and support groups. The disengagement of local authorities who no longer had contractual ties to asylum accommodation or financial means to support asylum expertise meant that the space for dealing with accommodation complaints or issues was evacuated and left solely to private providers and the Home Office. As reports from both the Scottish Refugee Council (2014) and the National Audit Office (2014) have illustrated, the standard of accommodation in some dispersal regions has been poor and complaints procedures are lengthy, complex and difficult to navigate for those without high proficiency in the English language. Whilst expertise and support have been lost at a local authority level through the loss of funding, the subcontracting undertaken by private providers replaces past provision with a complex system of multiple landlords, authorities and intermediaries.
Reflecting upon the changes of COMPASS, we might argue that this transition has led to increasing instability and uncertainty within the asylum accommodation system. The shift in provision has meant not only the loss of some expertise, but also increased distancing between providers and asylum seekers in their ability to complain and to receive adequate and timely responses.
The instability that arises through COMPASS has also been extended through those public service cuts and fiscal constraints that Peck (2012) suggests are shaping austerity urbanism. Whilst COMPASS itself is partly a response to austerity, the lack of subsequent support for asylum from local authorities once accommodation has been removed from their remit reflects the realities of local authority cuts and the need to prioritize resources ever more carefully. The context of austerity in which the COMPASS contracts were designed has meant that these contracts are modelled on the 'bottom line' of provision and profit. The confluence we see here between a context of austerity as a driving logic for the privatization of accommodation costs and a political demand to regulate and control those seen as 'unwanted' within the state reflects what Tonkiss (2013: 315) has termed the 'dual logic of cutback and crackdown' at the heart of austerity. Similarly to Smith 's (1996: 10) description of revanchism as the 'ugly cultural politics of neoliberal globalization', in asylum dispersal, we see the fusion of an exclusionary politics of citizenship that demands 'crackdowns' on those 'unworthy' of acceptance and who are feared to be abusing what support is provided (Sales 2002;Bloch and Schuster 2005) and a neoliberal governmentality of the market that demands austerity for housing the unwanted.

Cutbacks and Crackdowns
Whilst it is tempting to view the changes of COMPASS as ushering in a period of significant instability, it is important to contextualize this account 496 Jonathan Darling by considering the wider political context of dispersal as a policy in place for 16 years. Doing so highlights that, whilst the austerity politics of COMPASS has certainly extended and enhanced the neoliberal framing of asylum accommodation as an issue of cost-efficiencies and the 'rolling back' of state support, aspects of this process pre-existed the imposition of COMPASS. Taking this longer-term view illustrates that, whilst dispersal may be subject to a 'dual logic of cutback and crackdown' (Tonkiss 2013: 315), this 'dual logic' has a long history when discussing dispersal. Indeed, the turn to an 'age of austerity' that has been marked since the election of the Coalition government in 2010 masks a much longer history of austere politics and cumulative cutbacks in asylum policy. In the following section, I want to sketch some of the contours of this politics to illustrate the continuities that mark asylum accommodation as an arena of inherent instability. To do so, I focus on three aspects of dispersal.
First, since its inception, instability can be seen as a constituent part of the dispersal process due to the changing nature of asylum policy itself. As Ruth, the policy coordinator for a national refugee organization, highlights: the asylum process is complex, and shifts really quite often, even the language in it shifts, you know. UKBA is no more, it's now the Home Office, but nobody quite knows what model they are working to now (Ruth, interview, 2014).
The shifting nature of asylum policy and its priorities is well documented, with the New Labour government in particular producing new legislation on asylum and immigration at a rapid pace throughout the 2000s (Squire 2009). The effect of this turnover of policies is two-fold. First, as Ruth alludes to, it becomes hard for those working in the asylum sector to keep up with the demands of new policy priorities, restrictions and languages. When matters as seemingly minor as the number of bags allowed to be taken into dispersal accommodation, and as major as the institutional structure and oversight of the Home Office, are both altered on a regular basis, this produces a context in which continuity of service provision is incredibly difficult to achieve. Second, the changing policies and priorities of the asylum system help to generate a considerable turnover of staff within the Home Office and within private accommodation providers. This is both because staff tire of keeping track of the latest priority, and because those who are seen to be adaptable and effective are often moved on to other areas of social policy seen to be of greater importance. Whilst the transition to COMPASS led to a loss of expertise and staff within local authorities, concerns over a loss of knowledge, expertise and 'institutional memory' pre-date this shift in policy and have been argued to be endemic within the Home Office (Webber 2012;Gill 2015). The result of this turnover for those involved in asylum accommodation is a frustration at a lack of continuity, as Farah illustrates in relation to her role as part of an asylum networking group: Asylum in Austere Times 497 when you have a person from the Home Office at meetings. . . and they don't have that remit for housing or the background in housing then that's been quite difficult getting things escalated or things resolved. . .. [In the past] we have had somebody from the Home Office, but we've had a number of changes in people. And then it's telling that whole conversation again to a new person, and so that's stopped conversations or stopped things happening (Farah, interview, 2014).
The turnover of staff involved here is significant, as it adds frustration and delay to discussions between actors within the 'asylum sector' as key issues are repeated and newcomers to discussions often lack the contextual knowledge required to effect change. A loss of staff expertise and knowledge also produces a loss of personal connections and relationships developed over considerable periods of time. In working with the Home Office and others, relationships of trust are essential for both asylum support organizations, local authorities and third-sector groups, yet these are undermined if there is never an opportunity to develop connections with specific individuals responsible for policy or decision-making. Furthermore, the turnover of staff creates a context in which addressing concerns is a drawn-out and hardfought process, often being passed from one intermediary to the next as individuals shift positions and roles. Again, this creates a highly unstable environment for the effective implementation of asylum policy and for those seeking to address the failures of asylum accommodation in terms of quality standards and the rights of asylum seekers.
Second, whilst the provision of accommodation seen under COMPASS has focused on sourcing housing from the lowest-cost markets in the private rental sector, previous models still largely relied upon an economic rationale of low-cost provision above other concerns. As Paul, a policy coordinator for a third-sector organization in the midlands, outlines: looking at a map of dispersal areas in the UK it's quite clear that the Home Office have made that decision or processed that decision based on housing but there's so many other factors that need to be considered they're always kind of a bit of an after-thought (Paul, interview, 2013).
Similarly, Gareth argues that, in the initial phases of dispersal, the focus was placed upon procuring accommodation in areas of low-cost housing, often relying upon local authorities with empty social housing stock: talking about when dispersal first happened and we started to develop orientation projects what we found was that generally the local authorities weren't that well consulted on dispersal. . . some local authorities blatantly didn't want it to happen and dispersal was really geared towards, well, the main driving factor was availability of low cost housing (Gareth, interview, 2013).
Both Paul and Gareth highlight two significant aspects of dispersal: first, the varying levels of political support and engagement that local authorities have 498 Jonathan Darling had with the process; and, second, the dominance of an economic model focused on utilizing available low-cost housing. For Paul and Gareth, as for a number of academic explorations of dispersal (Robinson 2003;Phillips 2006), it was the housing market and its areas of low growth that dictated the map of asylum dispersal in Britain. This focus on the housing market has continued through COMPASS but with an important distinction: now property is sought in the private rental sector, rather than through social housing. Whether in social housing or the private rental sector, there is a continuity in how dispersal has focused on those areas of British cities most stigmatized as 'failing' and 'unproductive' (Phillimore and Goodson 2006;Spicer 2008). In this context, resentment towards asylum seekers has been a consistent feature of dispersal (Bowes et al. 2009), and one made all the more acute as austerity policies bite at the local level. As Lowndes and Pratchett (2012: 24) argue, if it is 'the poorest groups that suffer most through the Government's austerity measures then, by extension, it is the most deprived localities that will be most challenged', not least because the effects of austerity are 'politically, socially, institutionally and fiscally cumulative' (Peck 2012: 630). The cumulative effects of austerity combine to enhance resentments felt around asylum accommodation in the most deprived areas of the United Kingdom, and indicate a dispersal system that has always placed economic calculations of cost over and above the needs of both asylum seekers and the communities to which they are dispersed.
Finally, as alluded to by both Paul and Gareth, the history of dispersal highlights a dwindling of support for dispersal from many local authorities. Warren summarizes this history in the context of the West Midlands: When the whole dispersal programme came into effect. . . I guess there was quite a lot of idealism about what that was going to look like. So the talk had been from government about, well, we're actually going to be, effectively, creating these new communities in different regions of the country. . .. People are going to build, kind of, self-sustaining communities that were going to be appropriately resourced and supported. And then they never were. And there was also. . . quite a lot of buy-in and support from local authorities. . . as time went on, that [local authority support] just fell away until you got to the point where. . .the administration in Birmingham were just saying actually, we're going to wash our hands of this entirely. . . and also, we're going to make quite punitive noises about it, as though the city's actually been losing out from its relationship with the dispersal process, which actually wasn't ever the case (Warren, interview, 2014).
Initially, as Warren notes, there was some enthusiasm around dispersal from local authorities, as they viewed it as a means to fill otherwise hard-to-let social housing (Cohen 2002). Yet, over time, as concerns around community cohesion grew and funding levels to support dispersal fell, enthusiasm diminished. In the case of Birmingham, these concerns were interwoven with a context of political debate that Barbeho¨n and Mu¨nch (2015) argue views Asylum in Austere Times 499 immigration as a point of potential urban 'crisis', thus shaping the context of local political discussions of diversity, migration and asylum (Wilson 2015;Barbeho¨n and Mu¨nch 2016). As enthusiasm for dispersal fell away, Birmingham City Council withdrew from their contract with the Home Office to accommodate asylum seekers and argued that they needed to 'help the citizens of this city first and foremost' (Yeoman 2010). Framed in the context of a rise in homelessness in the city and in relation to urban austerity, the issue of prioritization was at the fore of this public rhetoric: 'With a long waiting list for homes, we really need all our properties for our people in these difficult economic times' (ibid.). The Birmingham case illustrates again that, for many local authorities, dispersal was viewed through a predominantly economic lens, such that it was supported when profitable as a means to gain rent from otherwise empty properties. As that income fell with cuts to the level of funding available from the Home Office, the gains of dispersal were calculated to be less than the political capital of opposing asylum support at a time of austerity. In this context, the question of prioritizing 'our people' becomes both an economic calculation of diminishing value and a political calculation of reinforcing distinctions between the 'needy' citizen and the 'abusive' non-citizen (Darling 2016b).
The asylum-dispersal system has therefore always represented an unstable assemblage of interests, authorities and priorities. It has shifted as local authority support has waxed and waned and as new policies and approaches are experimented with. However, the shift undertaken through COMPASS highlights the role instability plays in the neoliberal governmentality of dispersal. The unstable nature of asylum policy and provision on the one hand demands a response that is flexible, adaptive and responsive to changing political and economic conditions. This is precisely the mode of policy and subjectivity that neoliberalism valorizes and promises-an ability to mutate, adapt and create in response to changing external conditions (Peck 2010;Haughton et al. 2013). At the same time, instability may offer moments of 'impasse' that 'serve as a potential moment of dislocation in which ''normal politics'' is no longer perceived to deliver satisfactory outcomes' (Blanco et al. 2014: 3137). Instability in this way allows for, and legitimates, experimentation. If 'normal politics' and policies are no longer satisfactory, then alternatives must be sought, not necessarily to reduce that instability, but rather to meet the demands of efficiency and effectiveness that have underwritten dispersal. In this context, we might question how readily the privatization of asylum accommodation may be used to experiment with the austerity-driven privatization of wider areas of social housing and social care (see Daly et al. 2005;Smyth 2013).

Conclusion
In this article, I have offered an initial examination of the United Kingdom asylum-dispersal system and its changing nature under privatization. Through exploring the effects of COMPASS contracts for local authorities, support organizations and third-sector groups at an urban level, I have argued that we see a system of asylum accommodation being reworked into an increasingly insecure and unstable model of provision. Yet, at the same time, these facets of instability and political insecurity have long marked dispersal as an approach to housing vulnerable individuals. The focus on low-cost housing and efficiency that runs throughout dispersal may have been extended through COMPASS, but this does not mean that previous models of accommodation were focused on the needs of asylum seekers or on social care over economic necessity. Rather, we should critically reflect on the confluence of an unstable system of accommodation, with a politics of austerity that seeks to 'cutback and crackdown' (Tonkiss 2013: 315) on those seen as problematic to the state. It is this confluence that has shaped asylum accommodation and support in the United Kingdom over the last 16 years, and that continues to produce a system in which asylum seekers risk social isolation, harassment and marginality.
Responding to such a system demands a renewed focus on the local contexts in which dispersal is lived, so as to both address the daily exclusions experienced by asylum seekers and to practise alternative accounts of how asylum seekers fit within communities across Britain. On the one hand, this means utilizing the possibilities of local media and information campaigns to 'reinvent the national through re-imagining local identity and re-negotiating local networks of power and information', as Finney and Robinson (2008: 409) suggest (see also Wilson 2015). And, on the other, this means employing austerity not as a legitimation for exclusionary practices and cuts in support, but to articulate the common challenges faced by those present within areas hit hardest by austerity, regardless of immigration status. As Netto (2011b: 300) argues, 'responsiveness to refugees needs to be undertaken within the context of anti-poverty initiatives that target the socially disadvantaged neighbourhoods that they may occupy' (see also Bowes et al. 2009). At a time of divisive austerity measures and pervasive neoliberal logics of competition and economic worth, this form of responsive identification across difference becomes even more important.
issue. The article has benefited from the guidance, advice and critical commentary of Kevin Ward and Helen Wilson-my thanks to you both. All errors remain my own.