Unfree labour in immigration detention: exploitation and coercion of a captive immigrant workforce

Abstract This paper focuses on labour within immigration detention in the United Kingdom, offering an original national case study as well as a new conceptual framework for analysing such practices. It does so through an innovative engagement with recent literatures on forced labour, unfreedom and hyper-precarity, particularly amongst irregular migrants. We advance two key arguments in this paper. First, that the available data on labour within immigration detention indicate that detainees should legally be considered employees and granted access to labour protections, including the national minimum wage. Second, that work in immigration detention is an example of state-sanctioned exploitative, coercive and unfree labour amongst a hyper-precarious group of the population. This case has implications for other country contexts where immigration detention is used.


Introduction
The year 2016 marked the largest increase in global rates of immigration detention since 2010 (Global Detention Project, 2017). In the United Kingdom, despite a decrease in asylum applications since the mid-2000s (Home Office, 2016, the size of the detention estate has grown from a capacity of 250 detainees in 1993 to over 4,000 in 2017 (Shaw, 2016, p. 33). Redolent of the US 'prison industrial complex', this estate can be conceptualized as an 'immigration industrial complex', comprising a system of mutually beneficial relations linking corporations, government and media to the incarceration industry (Conlon & Hiemstra, 2014;Doty & Wheatley, 2013;Douglas & Saenz, 2013;Golash-Boza, 2009). But for immigration detainees such facilities are not only sites of incarceration, they are also places of work.
The United States first introduced 'voluntary work programmes' for immigration detainees in the 1950s in order to save on administrative expenses, a practice which continues today in the United States and has been adopted elsewhere. Detainees in the United States are paid $1 a day and are not considered to be employees as the relationship is 'rehabilitative' and ostensibly based on the upkeep of detention centres rather than profit (Sinha, 2015). Though it remains unclear exactly how many other countries have adopted detainee labour within immigration detention, the increased use of immigration detention on an international scale means that this practice is likely to be expanding. This paper focuses on labour within immigration detention in the United Kingdom, offering an original national case study as well as a new conceptual framework for analysing such practices. It does so through an innovative engagement with recent literatures on forced labour, unfreedom and hyper-precarity, particularly amongst irregular migrants (Cruz, 2018;Lewis et al., 2014Lewis et al., , 2015. Though the media have sporadically documented the existence of immigration detainee labour in the United Kingdom (Khan, 2014;Layton, 2009;McVeigh, 2011;Nye, 2014;Rawlinson, 2014;Standoff Films, 2015;Wheelan, 2014), there is very little data on its impact or any thorough academic analyses of the system in practice (notwithstanding a short commentary in Race & Class by Burnett and Chebe in 2010). This paper, in part, addresses this neglect.
We advance two main arguments. First, that the available data on paid activities in immigration detention indicate that detainees working within immigration removal centres in the United Kingdom should legally be considered employees and therefore granted access to labour protections and the national minimum wage. Second, that work in immigration detention centres within and beyond the United Kingdom may not constitute forced labour, particularly when viewed through the current liberal International Labour Organization (ILO) guidelines, but that it is an example of state-sanctioned exploitative, coercive and unfree labour. While we suggest here that detainee work does not appear to constitute forced labour under the ILO definition, we acknowledge that this is based on limited evidence and that numerous additional circumstances, such as threatening or violent behaviour on the part of guards, could render this type of labour 'forced' according to this definition.
In exploring unfree labour practices within immigration detention, we use the political economy framework adopted by LeBaron (2015), which focuses on three key domains of unfree labour: (i) compulsion and coercion of entry and exit, (ii) the wider social relations surrounding labour experiences and (iii) the conditions involved in, and experiences of, the labour process. We layer on to this framework the wider context of hyper-precarity (Lewis et al., 2015) which is experienced by many immigration detainees prior to, during and post detention, arguing that this, alongside a number of structural unfreedoms, renders detainees more willing to accept low-paid menial jobs, despite their own recognition that such practices are exploitative.
Engaging with the concept of 'hyper-precarity' and unfree labour distinguishes this paper from the US literature on labour within immigration detention, which situates detainee work within the historical context of slavery and its abolition (Sinha, 2015). This literature is nevertheless informative in drawing attention to the historically embedded, racialized nature of prison as punishment, as well as the privatization of incarceration. Here, we situate 'race' within a broader sphere of factors that further compound labour unfreedoms. In addition, we analyse the utility of the ILO Forced Labour Convention 1930 with regard to immigration detainees and discuss the complexities arising from their situation, particularly in terms of the liberal binary between 'voluntary' and 'involuntary' labour which fails to account for economic pressures.
Whilst this paper has a conceptual focus, it also draws upon a body of empirical material. This includes reference to international instruments; domestic statutes and case law; parliamentary debates on the introduction of rules excluding immigration detainees from national minimum wage regulations (accessed via Hansard); a parliamentary inquiry on immigration detention led by the All Party Parliamentary Group (APPG) Refugees and APPG Migration (2015); and the 2016 Shaw Review into the Welfare in Detention of Vulnerable Persons which was commissioned by the Home Office. In addition, Home Office data and guidance are drawn upon, as well as documents detailing the nature and extent of paid activities in detention obtained via freedom of information (FOI) requests submitted to the Home Office by the authors. We also utilize (with due caution) evidence produced by campaign groups including a 2014 report produced by Corporate Watch which contains detainee testimony, and a 2015 documentary film, Working Illegally, produced by Standoff Films which further documents the findings of the Corporate Watch report. While these are not academic research sources, there is very little empirical data on the impact of the paid activities regime. Accordingly, this evidence and the accounts contained in the two parliamentary inquiries constitute the only testimony from detainees on the paid activities system currently available.
The scarcity of empirical data is perhaps due to the lack of transparency surrounding immigration detention. This arises because of outsourcing arrangements with private corporations, including G4S, Mitie, GEO and Serco. Indeed, after submitting a FOI request to the Home Office, we were informed that part of our request could not be completed as 'service providers for our privately operated centres are not obliged (or required) to release information pertaining to internal guidance or information that they have prepared and hold in relation to paid work' . Such a position reveals the disjuncture between private institutions that deliver a significant proportion of detention services (and that are not bound by the terms of the Freedom of Information Act 2000) and public institutions that are obliged to release information upon request, unless an exemption applies (s1, Freedom of Information Act 2000). Despite these difficulties, we obtained quantitative data across all of the detention centres engaging in 'paid activities' and qualitative internal guidance on paid activities for two detention centres -The Verne and Morton Hallboth of which are run by Her Majesty's Prison Service (HMPS).
In the next section we map the legislative basis for immigration detention in the United Kingdom and the outsourcing arrangements of the detention estate. The second section demonstrates that this form of work has all of the legal characteristics of employment under UK law and should thereby be treated as such. The third section then contends that this form of labour, though regarded by policy-makers as 'consensual', should not be considered free or non-exploitative, while the fourth section explores how such practices align with some key indicators of unfreedom.

Immigration detention in the United Kingdom: a captive immigrant workforce
The United Kingdom has one of the largest immigration detention estates in Europe (APPG Refugees & APPG Migration, 2015;Shaw, 2016); 29,762 people were detained in the year ending September 2016 (Home Office, 2016). The power to detain derives from the Immigration Act 1971, the Nationality, Immigration and Asylum Act 2002 and the UK Borders Act 2007 which justify detention in a number of circumstances: to establish a person's identity claim; to effect their removal; where there is reason to believe that a person will fail to comply with the conditions attached to their grant of temporary admission or release; and where Home Office officials are deciding whether to grant temporary admission (see Immigration Act 1971, Schedule 2 para 16(2) and Schedule 3, para 2; Nationality, Immigration and Asylum Act 2002, s62;UK Borders Act 2007, s36). Whilst the majority of those detained are asylum seekers and refused asylum seekers (APPG Refugees & APPG Migration, 2015), a number of other status groups are subject to immigration detention, including: newly arrived migrants who have been refused permission to enter the United Kingdom; visa overstayers; those in breach of visa conditions; those likely to fail to comply with conditions attached to their admission or release; and those who have committed a crime and have been issued with a deportation order. As asylum seekers and refused asylum seekers are the most common category of detainee held within immigration detention, the circumstances of asylum applicants and those who have had their asylum applications refused are important in drawing out the exploitative nature of the 'paid activities' regime. Throughout the paper, this group will be referred to as 'the asylum-seeking community'.
There are now 10 immigration removal centres in the United Kingdom, the majority of which are run by private corporations, making immigration detention a highly profitable sector of the incarceration industry as contractors are provided with a fee per inmate, per day. In 2013/2014 the average daily cost of detaining one person was £98.70 which is the equivalent of £36,026.00 per yearthis far outweighs the costs of reporting or using monitoring systems such as radio frequency bracelets to reduce the risk of absconding (Gower, 2015, p. 6).
In spite of demands from NGOs and widespread criticism of the practice, there remains no time limit on the length of detention for immigration purposes within the United Kingdom. As a result of this open-ended approach, numerous people have been held within immigration detention for significant periods, as evidenced by the Gov.uk statistics for September 2015-2016 which reveal that of the 29,762 people detained that year, 6 per cent had been detained for over one or two years, 12 per cent had been detained for between two and four months, 19 per cent had been detained for between 29 days and two months and 63 per cent had been detained for less than 29 days. In 2015 the (then) Immigration Minister, James Brokenshire, submitted written evidence to the All Party Parliamentary Inquiry on the use of detention in the United Kingdom stating that the power to detain, though not subject to statutory limitation, is subject to the 'Hardial Singh' principles which alongside judicial oversight protects against the arbitrary use of detention (see R v Governor of Durham Ex parte Hardial Singh [1984] 1 WLR 704). There are, however, a number of detention powers to which the Hardial Singh principles are irrelevant, including the now inactive detained fast-track procedure, and where persons are detained at ports of entry in order to establish permission to enter.
Concerns have also been raised regarding the inconsistent application of the Hardial Singh principles by case owners (HM Inspectorate of Prisons and the Independent Chief Inspector of Borders and Immigration, 2012) and regarding the varied interpretation of what constitutes a 'reasonable' period of detention (APPG Refugees & APPG Migration, 2015). For example, in the case of R (on the application of Giwa) v Secretary of State for the Home department (2013) EWHC 3189, a detention period of 53 months pending deportation was considered reasonable. During this time frame the applicant aged from 18 to 24 years of age. As will be addressed below, we argue that this uncertainty regarding an individual's length of detention contributes to the positioning of immigration detainees as hyper-precarious and more willing to accept highly exploitative working conditions as they 'have the simultaneous concern both that there will be sudden change and never-ending stasis' (Griffiths, 2014(Griffiths, , p. 2004. Immigration detention centres have been subject to consistent criticism since their proliferation in the early 2000s, most notably from the detainees themselves. Throughout the 2000s detainee riots took place in response to conditions of detention which led to the near-destruction of both Yarl's Wood in 2002and Harmondsworth in 2004and 2006. There have also been repeated allegations of sexual abuse, inhumane and degrading treatment, racism and malpractice on the part of guards, which was most recently exposed in a number of undercover documentaries shot by television broadcaster Channel 4 during which Serco guards were caught describing detainees as 'animals', 'beasties' and 'bitches' (Sorrell, 2015); see also HM Chief Inspector of Prisons, 2013a, 2013b). The 2015 Parliamentary Inquiry into Immigration Detention found that many detainees have mental health conditions, and that the experience of detention exacerbates post-traumatic stress disorder and depression, both of which are common amongst the detainee population.
Conditions of detention within immigration removal centres are governed by the Detention Centre Rules 2001 which are explicitly modelled on the Prison Rules 1964, and instructions and procedures for staff are contained in Detention Service Orders and Chapter 55 of the Enforcement Instructions and Guidance. Though some may consider the application of these rules inappropriate to those seeking asylum, the justification, as reported to the HM Inspector of Prisons in 1999, is that 'detainees are in an equivalent position to prisoners remanded before trial, in that they are prima facie offenders against the immigration rules' (Sawyer & Turpin, 2005, p. 708). It is perhaps then unsurprising that a number of removal centres, such as Brook House, Colnbrook and Harmondsworth, were constructed to category B prison standards rendering them 'somewhat claustrophobic [… with] the "feel" and look of contemporary gaols' (Shaw, 2016, p. 32). This context is, we argue, important in assessing the voluntariness or otherwise of paid activities in detention centres.
With regard to the outsourcing of the detention estate, the corporations currently running the removal centres are G4S, Mitie, GEO and Serco, while a small number of centres remain under the management of HMPS including sites at Morton Hall and The Verne (Shaw, 2016). As well as managing the immigration detention estate, all four of the private corporations listed above have commercial government contracts to provide prison services and police custody. In this context, Bacon (2005, p. 4) links the privatization of immigration removal centres to a growth in the use of detention, a willingness amongst immigration officers to detain, the growing secrecy and lack of accountability present in the detention estate and the move towards increasingly harsh detention practices. Though such claims continue to be denied by the Government, the campaign group Corporate Watch suggests that companies such as GEO could be saving £1.5 million annually by paying detainees £1 per hour to undertake jobs that would otherwise be performed by workers entitled to the national minimum wage (Miller, 2014). As well as constituting the raw materials processed by for-profit private corporations involved in the immigration industrial complex, inmates are also then directly exploited as a source of cheap labour power (O'Connell Davidson, 2015, p. 98).

Paid activities or underpaid labour?
In 2006 the Blair government introduced 'paid activities' into immigration removal centres in the United Kingdom, ostensibly to prevent instances of boredom and frustration amongst detainees (Tony McNulty, Hansard, 16 November 2005, column 1016. Concerned that detainees would be regarded as 'workers' and thereby entitled to a range of employment protections, the Government excluded detained persons from protection under the National Minimum Wage Act 1998 on grounds that granting detainees the national minimum wage 'would not be viable financially, nor reflect the true economic value of the work likely to be carried out' (Tony McNulty, Hansard, 16 November 2005, column 1016; see also s153A Immigration and Asylum Act 1999 as inserted by s59 Immigration, Asylum and Nationality Act 2006). As a result, immigration detainees engage in paid work activities such as cleaning, cooking, hairdressing and decorating for £1 or £1.25 an hour in immigration removal centres throughout the United Kingdom (s4-6 Detention Services Order 01/2013). This is in contrast to the national living/minimum wage paid to British citizens, which is currently £7.83 per hour for over-25s, £7.38 for 20-24-year-olds, and £5.90 for 18-20-year-olds (The National Minimum Wage Amendment Regulations 2018 SI 2018/000).
Research conducted by a number of NGOs and journalists suggests that detainees' work is integral to the running of the detention centres, reducing the need for paid staff who would otherwise fill these positions (Miller, 2014;Standoff Films, 2015;Wheelan, 2014). A FOI request submitted by the authors (FOI-41219) reveals that detainees undertake a myriad of jobs whilst in detention, which resulted in 495,270 hours-worth of work in 2014; 923,154 hours-worth of work in 2015; and 537,160 hours-worth of work between January and July 2016 (when the request was submitted). The data reveal that detainees are providing significantly more in working time than previous studies, such as Miller (2014), have suggested. The rules maintain that these activities are designed to tackle boredom; however, it has been noted that, in contrast to prisons, immigration removal centres are not obliged to provide educational or therapeutic boredom-alleviating activities such as arts and crafts, English language classes or training in IT support (Bosworth, 2012).
As part of our FOI request we were also provided with the internal documentation and guidance relating to 'paid work' at The Verne immigration removal centre and Morton Hall, both run by HMPS. Job roles available to detainees at The Verne consist of: cleaners, laundry workers, gardeners, litter pickers, kitchen assistants, waste disposal workers, barbers, librarians, decorators and gym orderlies. Referring to prison labour in the United States, Zatz (2008, p. 870) labels this form of labour 'prison housework' in which 'a prison manages production and also consumes its output, as inmates contribute directly to prison operations by cooking meals, doing laundry, or cleaning the facilities'. Evidence suggests that detainees are able to put themselves forward for whichever role they desire; however, similarly to the prison industrial complex, eligibility to work is linked to compliance and cooperation, underpinned by control. Accordingly, those wanting to engage in paid work must comply with the UK Border Agency in helping to resolve their case. Non-compliant or violent behaviour (such as failing to attend an interview or causing disruption) will automatically exclude individuals from paid work opportunities (Detention Services Order 01/2013 s10). This was criticized in the Shaw Review (2016) as 'redolent of the prison system' and an unnecessary limitation on the eligibility to work.
Once detainees are deemed eligible and agree to undertake a job they are provided with a timetable which specifies their place of work for both the morning and the afternoon of each day. For example, one of the forms specifies that, on Monday mornings, the cleaning orderly for the activities centre must clean the art room and, in the afternoon, the activity offices, alongside daily duties such as: the sweeping and mopping of corridors and the foyer; the emptying of bins in the offices; the cleaning of the kitchen area; and the checking of toilets. Detainees are also provided with an 'Employment Job Description' which confirms their title, responsibilities, pay, hours of work and the area in which they will be working. In the documentation provided for Morton Hall, the resident pay policy sets out that detainees should not work more than 30 hours per week and specifies a pay scheme for sickness and absence which consists of £1 per day for 'acceptable' absences. The pay policy also contains information on the consequences of 'sub-standard performance', such as poor timekeeping and absence, which can result in the detainees' removal from paid activity. Training (such as manual lifting) is also provided depending on the job role.
In light of this data we suggest that detainees working within immigration removal centres should be considered employees as defined under section 230 (1)  The irreducible minimum required for a contract of employment to exist is: personal service, meaning individuals cannot use unapproved substitutes to carry out their duties; a sufficient level of control exercised by the employer; and mutuality of obligation between the parties, which relates to an obligation upon the employer to provide work, and a duty upon the employee to accept and perform that work in return for a wage. The courts will also consider whether there are any other factors indicative of a contract of employment, such as the existence of sick pay or holiday pay, and whether an individual is integrated into the business, which includes consideration of uniform policies and the provision of equipment.
Applying the common law tests of personal service, control and mutuality of obligation to the 'paid activities' arrangement, detainees should be considered employees. Of particular relevance is the fact that detainees are given a written job description which requires them to provide their own services with no further details on the issue of substitution (Express and Echo Publications v Tanton [1999] ICR 693). Whilst at work detainees remain under the control of the centre which is demonstrated by the specific duties given to them in their job description and the application of disciplinary codes for inadequate performance; this means that the centres exercise a significant degree of control over what the detainees do and how they do it, reflective of a traditional master and servant relationship (Ready Mixed Concrete (South East) Ltd v Minister of Pensions and National Insurance [1968] 2 QB 497). The timetables provided at The Verne also suggest that mutuality of obligation exists as detainees are provided with 30 hours of work each week and must notify the centre of any 'acceptable' absences for which sickness pay will be granted (O'Kelly v Trusthouse Forte Plc [1984] QB 90). We also note that even if detainees are not considered employees, they would most certainly be workers via the status test set out in s230(3) of the Employment Rights Act 1996 which requires a contract by which an individual undertakes to perform work personally and where that individual is not in business of their own account (i.e. selfemployed). This is an easier test to meet than employee status (which grants more rights) and provides the right to the national minimum wage (see Bates van Winkelhof v Clyde & Co LLP [2014] UKSC 32).
The suggestion that detainees are employees, or at the very least workers, is supported by comments from the (then) Home Office minister Tony McNulty when introducing the paid activities regime to Parliament, during which he claimed that 'detainees may be regarded as "workers" for purposes of the National Minimum Wage Act 1998 if they perform paid activity of any sort, and would therefore be entitled to receive the national minimum wage' (Tony McNulty, Hansard, 16 November 2005, column 1016). Rather than recognizing immigration detainees as 'workers', however, the Government consider the paid activities regime to be analogous to the paid work arrangements in UK prisons (Prison Rules 1999SI 1999 on grounds that custodial work is in the detainees' interests (Tony McNulty, Hansard, 16 November 2005, column 1016 and 1017). Immigration detainees and prisoners can be distinguished, however, as those within immigration detention are not held on account of criminal activity. To be a criminal is to be recognized and subjected to the law and its punishments. To be an immigration detainee is to be subject to an administrative apparatus and thus 'figured outside of the purview of the law altogether' (De Genova, 2016, p. 6).
Unlike prison work, detainee work is also not designed to rehabilitate as deportation is the central premise behind immigration detention. In spite of this, a significant proportion of detainees are released, and reports suggest that those who are detained for longer periods are less likely to be deported (Bail for Immigration Detainees, 2015). Accordingly, some parallels can be drawn between immigration detainees and prisoners concerning the effects of working within the detention regime. LeBaron (2012), writing in the US context, argues that in the face of resistance from a significant proportion of the labouring population to precarious forms of working, prison labour has been used as part of a broader strategy to instil the market and social discipline needed to sustain the neoliberal order (see also De Giorgi, 2010;Schumpeter, 1942;Wacquant, 1999). From this perspective, prison work is a means of indoctrinating the incarcerated class into precarious forms of labour that serve the capitalist agenda both inside and outside of prison. Though immigration detainees are, in both the United Kingdom and United States, legally restricted from working once released (unless given leave to remain), the impact of their labour within detention may have the paradoxical effect of pushing them towards more exploitative forms of 'illegal' working when released.
Categorizing those engaging in 'paid activities' as employees is significant in this regard because employee status is attached to a number of employment rights which were designed to counteract the power imbalance between employers and employees which often leads to exploitation. These rights include but are not limited to: unfair dismissal; entitlement to notice; a duty of care; redundancy pay; the national minimum wage; controls over deductions from wages; rest breaks and protection from making lawful disclosures. At present, it is clear that a number of these rights are not enjoyed by detainees, most notably with regard to the national minimum wage. Yet as the UK Parliament has specifically legislated to remove this right from the group, there is little room for the courts to manoeuvre with regard to enforcement. Similarly, s45 of the National Minimum Wage Act inhibits prisoners from claiming the national minimum wage when engaging in prison labour. Yet, it is the immigration status of detainees which determines the value of work, as despite the fact that they perform the same jobs as their citizen counterparts, they are paid at a rate of 13 per cent of the national minimum wage. This position conflicts with a number of international instruments that afford the right to just and favourable remuneration, including Article 23(3) of the Universal Declaration of Human Rights, Article 4 of the European Social Charter, Article 7 of the International Covenant on Economic, Social and Cultural Rights and Articles 2 and 3 of ILO Convention No. 131 (1970); yet these instruments have little impact within domestic law in the absence of national legislation which incorporates the rights. Claims for enforcement of domestic labour standardssuch as notice paymay, however, have greater success as the UK Government has not expressly legislated to remove these rights from detainees.
Ironically, detained migrants are prohibited from engaging in employment outside of detention via the Immigration Act 1971 as they are persons liable to detention, and as a result of changes to policies on asylum seekers' rights to work in 2002 (see Mayblin, 2016Mayblin, , 2017b. As well as being justified as a form of deterrence (deterring irregular migration including disingenuous asylum applicants), restrictions on working are presented as a means of protecting British workers and business from illegitimate competition (Anderson, 2010;Bales, 2013;Mayblin, 2016Mayblin, , 2017b. Detention centre practices in which detainees work for low pay, potentially displacing British workers, thus conflict with the Home Office's own position that immigrants who 'end up in shadowy jobs in the grey economy undermine the terms and working conditions of British workers' (John Reid, Home Office, 2007, p. 2).
The proposition that detainees are replacing paid workers in detention centres has been refuted by the Home Office which maintains that paid work practices are 'not intended to substitute the work of trained staff' (Miller, 2014). This is also supported by Shaw who claimed that nothing he saw during his review of immigration detention supported the view that paid work is used by contractors to increase their profits (Shaw, 2016, p. 136). Yet if labour in detention is intended as primarily a means of alleviating boredom, why focus on activities which render detainees complicit in the maintenance of their own incarceration (as cleaners, cooks and handymen) at the expense of activities such as sports or language lessons? Perhaps because by offering jobs in detention, those running the centres receive cheap labour, while expanding the provision of leisure activities would cost money. Burnett and Chebe (2010, p. 98) note that in this way 'a core rationale of detainee labour is to reduce expenditure on essential aspects of the day-to-day running of the detention estate'. This also has the potential to degrade the wages and working conditions of paid staff who cannot parallel the subservience of the detained population (LeBaron, 2012).
Though detainee accounts of the paid work regime denote feelings of exploitation (Miller, 2014;Standoff Films, 2015;Wheelan, 2014), the work undertaken by immigration detainees may not fall under the strict liberal definition of forced labour adopted by the ILO (discussed below). Even if such practices are not constitutive of forced labour under the ILO definition, we argue that detainees' labour should not be considered 'free' as the ILO's emphasis on consent fails to capture the ways in which immigration status intersects with social, economic and legal relations to produce hyper-precarity amongst the detainee population (see Anderson, 2010;Lewis et al., 2014Lewis et al., , 2015. The next section explores how the concept of 'unfreedom' can aid our understanding of this issue.

Hyper-precarity and the continuum of 'unfreedom'
The ILO's approach to forced labour is widely recognized as dictating the parameters of institutional engagement with the concept at an international, European and national level (Lerche, 2011;Lewis et al., 2014). Accordingly, the ILO's construction of forced labour and the forces sustaining its existence are indicative of policy responses more generally. Article 2(1) of the ILO Forced Labour Convention 1930 identifies forced labour as exacted under the menace of penalty and for which persons have not offered themselves voluntarily. Article 2(2) provides for certain exceptions to the rule, such as the work of convicted prisoners or military service; however, none of the Article 2(2) exceptions apply to immigration detainees who are not detained on account of criminal conviction. Consent will also be rendered irrelevant where there is evidence of trafficking such as deception, fraud or the retention of identity documents (ILO, 2012); yet none of these factors apply to the situation of immigration detainees, who appear to be presented with an accurate account of the work before entering the agreement. The forced labour definition is further substantiated by the 11 ILO forced labour indicators which are designed to help 'front line' actors, such as criminal law enforcement officials, identify individuals that are potentially trapped in situations of forced labour. When analysing whether an individual is subject to forced labour, the assessor should have regard to the recruitment of the individual, their life at work and their ability to leave employment (ILO, 2011).
The indicators themselves are further qualified into strong, medium and weak categories. If two or more indicators are present, and at least one of these indicators is categorized as 'strong', the practice under scrutiny is likely to be deemed forced labour (Skrivankova, 2010; Allen v Hounga [2014] 1 WLR 2889) though at least one indicator should relate to involuntariness and the other to penalty (ILO, 2011, p. 27). The 11 ILO indicators are: abuse of vulnerability; deception; restriction of movement; isolation; physical and sexual violence; intimidation and threats; retention of identity documents; withholding of wages; debt bondage; abusive working and living conditions; excessive overtime (ILO, 2012). Many of these indicators resonate with the position in which immigration detainees are placed, particularly with regard to the abuse of vulnerability and the restriction of movement and isolation; however, there remains an implicit involuntariness within a number of the indicators. Restriction of movement, for example, is explained with reference to being 'locked up' and/or guarded to prevent escape from labouring, and the indicator relating to physical or sexual violence refers to the use of violence in forcing workers to undertake tasks (ILO, 2012).
The liberal concept of 'voluntariness' thus remains central in drawing the line between free and unfree labour practices. This liberal 'consent' based approach, however, fractures the social realities of many workers' experiences, which are uncomfortably forced into the rigid binary of 'free' or 'unfree' labour practices. Within this paradigm, unfree labour is regarded as an exception to the normal functioning of the market, promoting an image of the 'ideal victim' who is identifiable by subjection to force, control and/or deception (Cruz, 2018;Hoyle et al., 2011;Munro, 2005). Conversely, 'free wage labour is viewed as the consensual exchange of non-exploited labour power between equals in response to market forces' (Cruz, 2018, p. 68). Lerche (2007) contends that the ILO's strategy in 'cocooning' forced labour issues from other labour practices depoliticizes forced labour and isolates the issue as an 'unnatural' element of capitalism. In doing so, the ILO is able to maintain 'safe' relationships with governments and run anti-forced labour programmes. However, the deeper issues of rehabilitating forced labourers and preventing their return to forced labour relations cannot be tackled in isolation from its broader social and economic context.
In contrast to the liberal approach, we argue that wage labour can only ever be 'formally' free in so far as individuals have the capacity to sell their labour as a commodity (Fudge, 2014, p. 38;Holmstrom, 1977, p. 357). Within this paradigm the worker is 'compelled to sell himself of his own free will' in order to survive (Marx, 1976(Marx, [1867, p. 178; see also Hale, 1923, p. 473), thus rendering the concept of 'free' labour incoherent as it camouflages the fact that all wage labour is subject to some level of compulsion within a capitalist system (Banaji, 2003, p. 87). From the liberal viewpoint, labour that is compelled by economic reasons but subject to abysmal pay or conditions is thus non-exploitative and non-coerced as the price of labour and conditions of work are decided by the market and hence represent a 'just' price … It is only the deviation from 'free' labour markets that should be avoided, even if the outcomes from 'free' labour and from labour markets where non-economic force prevails are more or less the same. (Lerche, 2011, p. 16) Banaji therefore argues that all wage labour is 'more-or-less coerced'/'more-orless free ' (2003, p. 80). Such a perspective accords with the work of Lerche (2007), LeBaron (2015 and Phillips (2013) who also reject the free/ unfree binary, instead construing 'free' and unfree labour as part of a continuum of labour relations based on degree rather than binary opposition. This continuum is not pre-set but rather evolves with the processes which influence it, including for example changes to legal, gender and labour relations. The likelihood that individuals or groups will be subject to the more exploitative end of the continuum is related to a broad set of intersecting factors, of which immigration status is one.
In reference to this literature, Lewis et al. (2015, p. 582) argue that 'many exploited migrants' lives are best characterized by hyper-precarity that emerges from the ongoing interplay of neoliberal labour markets and highly restrictive immigration regimes'. The lens of 'hyper-precarity' recognizes that the intersection of a number of structural processes, such as the context of an individual's migratory movements, gender and socio-legal status 'create multi-dimensional insecurities that contribute to the necessity to engage in, and close down exit from, severely exploitative or forced labour' (Lewis et al., 2014, p. 171). State stratification of socio-legal statuses are here 'central to the production of (hyper)-precarious migrant workers' (Lewis et al., 2014, p. 582) which distinguish migrants from the wider precariat and necessitate risk taking and engagement with exploitative forms of work for lack of a viable alternative. This situation of hyper-precarity amongst irregular migrants, created in no small part by government policy, is not just significant in understanding the conditions outside of detention, but is also highly relevant in analysing the coercive nature of work within detention.
Though it is recognized that in a capitalist society 'freedom' can never be truly achieved as individuals cannot secure a livelihood without accessing a wage, labour might be considered more-free where it complies with international and national labour regulations such as the ILO's decent work indicators which include: adequate earnings, equal opportunities and a safe working environment, as well as national minimum wage rates (Skrivankova, 2010). Degrees of coercion, however, are more difficult to define, notwithstanding the 11 forced labour indicators identified by the ILO, which exist at an extreme end of the spectrum. In analysing unfree labour practices that fall outside of the strict 'involuntary' definition adopted by the ILO, it is necessary to look at the range of factors influencing an individual's employment opportunities, including the wider legal, social and economic conditions governing their circumstances.
Unfreedom is not therefore an intrinsic attribute of migrant workers or their employment but rather 'describes relationships that are actively produced and institutionalised by employers and the state' (Strauss & McGrath, 2016, p. 5). This accords with the political economy perspective which promotes a broad definition of unfreedom taking into account 'economic compulsion, exploitation and mechanisms of control and abuse (from the most mundane and routine), as well as the legal, gender and racial relations that facilitate capitalist accumulation' (Cruz, 2018, p. 68). Accordingly, we here move beyond a liberal analysis of the interpersonal factors that result in detainees' agreement to engage in work. The fact that immigration detainees are not physically threatened into agreeing to work, or held in debt bondage, is not therefore indicative of 'free' labour practices. We explore these issues in more detail in the next section.

Indicators of unfreedom
In order to encapsulate a broad definition of (un)free labour, LeBaron (2015, p. 8), drawing on Marx, suggests three areas of consideration: (i) the systemic compulsions and forms of coercion that underpin people's entrance into labour and their (in)ability to leave the labour contract; (ii) the social relations surrounding experiences in the sphere of circulation; and (iii) the conditions involved in, and their experience of, the labour process itself. Following this framework, we are able to analyse labour within immigration detention in dialogue with Lewis et al.'s (2014Lewis et al.'s ( , 2015 work on hyper-precarity. As the asylum-seeking community makes up a significant proportion of the detainee population, our analysis will primarily focus upon the unfreedoms of this group.

Systemic compulsion and coercion underpinning entrance into and exit from labour
As addressed by LeBaron (2015), in order to assess the forms of coercion that underpin detainees' entrance into the labour market of the detention centre, we must acknowledge that global processes shape unfree labour relations at different scales and levels of governance. Detainees are usually poor. They have been prohibited from working in the formal economy prior to detention and are likely to have been destitute and/or living in poverty (see Allsopp et al., 2014). Working in detention entails entering the internal labour 'market' of the centre in which pay rates are fixed low, but such jobs represent the only opportunity for earning money and thus providing for needs. Hale argues that it is the law of property (the privatization of basic amenities such as land and water) which coerces individuals into wage-work under the penalty of starvation (Hale, 1923, p. 473). Whilst basic food and water are not a concern for immigration detainees, Standoff Film makers revealed that multiple detainees reported lack of food provisions during Ramadan, meaning many Muslim detainees had to buy snacks from the over-priced shop in order to eat at all. A number of detainees featured in the Standoff Films documentary also note their compulsion to engage in paid work in order to obtain other goods and services such as phone credit for the purpose of familial contact or contacting their solicitor. As one interviewee explained: An Immigration Officer will approach you and say: 'There are jobs to do here, do you need to work?' It's where you see that all your humanity has been stripped off and, you know, it's like, you need to get out of there. You don't have a calling card to call your solicitor. You don't have … a way out. You just bury your pride, somewhere, somehow and go back to the officer and say, 'you know what, I think I'll do it'. (Anonymous interviewee, immigration detainee, Standoff Films, 2015) Similarly to US detention centres, UK providers thus appear to have created a cyclical process whereby detainees are driven to purchase goods and services at inflated prices as a result of their detention, which then necessitates working for minimal pay. The bulk of their earnings are then returned to the detention facility through the purchasing of goods and services (Conlon & Hiemstra, 2014, p. 341).
The need to work to provide remittances for family members both inside and outside of the United Kingdom is a further incentive, as one detainee stated in the Standoff Films documentary: 'I'm working here not for me. I'm not working for myself, I'm working for my daughter' (Anonymous interviewee, immigration detainee, Standoff Films, 2015). Here then, we see a form of social reproduction, that being 'the labour of individuals to sustain, care for, and attend the survival of, well-being, and reproduction of themselves and each other' (Hardy, 2016, p. 880), which is also one of the most common pressures leading to exploitative work amongst asylum seekers and refused asylum seekers outside of detention (Waite, 2017). The compulsion to enter into these types of jobs as the single means of earning money within detention is not therefore a voluntary choice made within a suite of options; there is no other option.
Whilst entrance to or exit from labour does not appear prohibited under threat of violence or other means (a common indicator of forced labour), such labour is not demonstrably more-or-less-free in that the circumstances of entry into underpaid labour (which would in any other circumstance breach UK employment law) are more-or-less coercive. Indeed, detainees are not physically free; they are incarcerated, setting the decision to undertake paid work in detention within the immediate context of carceral unfreedom. The conditions of detention therefore produce some degree of coercion, even where detainees are keen to work (Conlon & Hiemstra, 2017, argue as much in the US context). However, the compulsion and coercion of entry and exit can only be fully accounted for within the context of wider social relations and the hyper-precarity which this gives rise to, a topic which is discussed in the next section.

Social relations surrounding experiences in the sphere of circulation
The conditions of unfreedom within immigration detention are embedded within wider (unequal) social and economic relations. This type of unfree labour is therefore (re)produced by wider structural hierarchies including those of 'race', gender and citizenship (see LeBaron, 2015). It is salient to note here then that asylum seekers are generally people of colour, coming from formerly colonized countries. During the period of July to September 2017, statistics show, for example, that the top asylum sending countries to the United Kingdom were Iran, Pakistan, Iraq, Bangladesh and Sudan, four of which were directly colonized by Britain (Home Office, 2017). Yet, even where asylum seekers from particular countries are not connected to Britain's colonial histories, as acknowledged by Gilroy (2004), they may nevertheless be caught up in a pattern of hostility as a consequence of its aftermath. Though it is beyond the scope of this paper to delve into the implications of these histories more fully, we have elsewhere argued that these historical-contextual foundations are crucial in understanding how differential worth is assigned to various human bodies and why particular 'dehumanizing' policies, such as immigration detention and 'paid activities', continue to be implemented despite their widespread critique (Mayblin, 2017a).
Gender is also an important component in understanding the unfreedoms of immigration detainees and the coercive nature of the 'paid activities' regime. In the United Kingdom the vast majority of asylum seekers are young and male (Home Office, 2017). Traditionally, ideas of masculinity are linked to the 'breadwinning' male as an active, labouring body or to positions of authority, expertise and knowledge (Conaghan, forthcoming). 'Prison housework' (Zatz, 2008), however, forms socially reproductive work which is often devalued due to its association with 'women's work' which is regularly low paid or not remunerated (Conaghan, forthcoming;Cruz, 2018). The performance of these gendered forms of work may be experienced by male detainees as 'emasculating' forms of work further augmenting conditions of coercion, which is linked to unfreedom.
In addition to the structural hierarchies of gender and race, immigration controls work with and against migratory processes to produce workers with particular types of relations to employers and labour markets (Anderson, 2010). In this context, the imposition of work restrictions and criminal sanctions for illegal working renders detainees much more willing to engage in work whilst in detention as it is the only form of legal work which they can perform without the overarching fear of prosecution (Immigration Act 1971, s24B, as inserted by Immigration Act 2016, s34; see also Bales, 2017). Equally, the uncertainty of immigration status renders many individuals precarious and susceptible to high levels of exploitation (Anderson, 2010;De Genova, 2002). For many detainees, work undertaken outside of detention is likely to have been in the informal economy where very poor working conditions and high levels of exploitation are known to occur, particularly under the threat of immigration-related sanctions (Doyle, 2009;Lewis et al., 2014;Williams & Kaye, 2010). In addition to this, the policy context for irregular migrants situates them as socially, economically, politically and geographically marginalized, and their disempowerment through measures of containment and control sees them enter detention from a situation of vulnerability, even hyper-precarity.
The condition of 'deportability', as Nicholas De Genova (2002) has pointed out, permeates every area of the lives of those who are liable to deportation, including migrant labour spaces (Lewis et al., 2015). De Genova (2002, p. 438) contends that 'it is deportability, and not deportation per se, that has historically rendered undocumented migrant labour a distinctly disposable commodity'. One witness in the 2015 parliamentary inquiry into immigration detention explained the disempowerment inherent in this relation as follows: The uncertainty is hard to bear. Your life is in limbo. No one tells you anything about how long you will stay or if you are going to get deported. (Anonymous, written evidence, APPG Refugees & APPG Migration, 2015, p. 20) For those who have fled contexts of persecution, 'there is a qualitatively different risk', Lewis et al. (2015, p. 593) suggest, as they fear 'not only the loss of face and changes in family relationships confronted by many migrants returning without the status or income expected from migration' but also 'the risks of persecution, torture, and other threats to themselves and their families in states known for human rights abuses and conflict'. If the threat of deportation produces subaltern migrant subjectivities, its constant immanence in immigration detention introduces a permanent threat, an extra level of punishment, that detainees must live under. They are produced as imminently deportable, and this condition, we argue, frames their engagement in underpaid captive labour.
The wider social relations of gender, racialization, marginalization and hyperprecarity within a stratified regime of rights are highly pertinent in understanding both the condition of deportability and the coercive nature of labour within immigration detention. These connections between the inside and outside of immigration detention, and their relation to the conditions involved in and experience of the labour process, are discussed further below.
The conditions involved in and experience of the labour process Despite having many of the characteristics of decent work, work in detention is very low paid, is entered into under conditions of compulsion and coercion and resides within highly unequal social relations. Detainee workers are physically immobilized, alienated and more-or-less coerced. The low levels of pay are one element of the poor conditions, but a compounding factor is that in some detention centres detainees are paid in the form of electronic credits that can only be spent at a shop inside the detention centre (Miller, 2014). That shop is of course run by the same company that detains them, the implications of which were drawn out earlier.
As well as being poorly paid and working in jobs which contribute to the everyday running of the facility which imprisons them, detainees by their very circumstance do not have 'control over their own bodies, movement, privacy and social life'a measure of unfreedom noted by LeBaron (2015, p. 9). They work alongside non-incarcerated workers who may be members of a trade union, but they cannot collectivize as workers themselves. As noted earlier, reports suggest that the oppressive nature of the physical environment is coupled with racism, verbal abuse, sexual abuse and bullying, which exacerbate the conditions of those with mental health problems. The 2015 detention inquiry reported that a number of detainees who gave evidence to the inquiry were victims of trafficking or torture. One detainee, who received £17 per week for serving breakfast, lunch and dinner over four days at the canteen in Harmondsworth detention centre, asked Corporate Watch: 'do they think we just got off a slave ship?' These conditionsof the work, of the working environment and the coercive nature of entryas well as the possibility that private immigration detention centres are making profit from this practice, lead us to the conclusion that work in detention centres is coercive, unfree and exploitative.

Conclusion
In many ways labour in immigration detention holds several key features of regular employment. It is legal, workers have contracts, use time sheets and are not (as workers) thought to be subject to arbitrary violence, the threat of violence or to be working in dangerous conditions absent of health and safety considerations. And yet the work carried out in detention is not free; it is exploitative, coercive and undertaken within the broader constraints of immigration control and the threat of deportation. This paper has first argued that detainees engaging in 'paid activities' in UK detention centres are working under contracts of employment and should thereby enjoy the range of employment rights set out in law. Though such improvements would be welcomed, the granting of employment protections would not render the 'paid activities' regime non-exploitative or non-coercive. This is because of a number of structural factors which cross borders. The socio-economic position of detainees, their immigration status and their deportability compel detainees to sell their labour power and place them under the dominion of the detention centre.
At a minimum we would suggest that the labour exploitation of a highly vulnerable group of individuals does not constitute an enriching activity. On the contrary, it is a form of exploitation which inevitably emerges from the hyperprecarity produced by exclusionary immigration regimes and international political-economic power relations. Here we have presented a detailed analysis from a UK perspective, but analysis of this case has broader implications.
The past decade has seen the proliferation of immigration detention globally, including its increased privatization. 'Deportability' is an experience shared by migrants across many country contexts, as are mental health problems due to the experience of violence, trauma, torture and persecution at the hands of state authorities. As set out in this paper, a number of factors must be taken into account in determining whether the labour of immigration detainees would constitute 'forced' labour under the ILO definition. Regardless of this qualification, however, we argue that the structures, social relations and position of immigration detainees as largely 'hyper-precarious' render this type of labour 'unfree'.