All-foreign prisons in the United States, England and Wales, and Norway: Related logics and local expressions

Norway, England and Wales, and the USA are among a small number of affluent Western countries to establish ‘all-foreign’ prisons in response to public concerns about the growing threat of foreign-national prisoners. Drawing on collaborative analysis of empirical data collected at all-foreign prisons in these three countries, this article traces the conditions in which all-foreign prisons emerged, the position and function of all-foreign prisons in specific national systems of criminal justice and immigration control, and the operation of all-foreign prisons within each context. The article points to a shared logic, while drawing attention to the local expressions of bordered penality.


Introduction
At the turn of the 21st century, Norway, England and Wales, and the USA (hereafter United States) each established all-foreign prisons in response to public concerns about the growing threat of foreign-national prisoners. In the United States, the Federal Bureau of Prisons began procurement for the first two privately contracted, allforeign 'criminal alien requirement' (CAR) prisons in 1999 (Department of Justice Office of the Inspector General [DOJ], 2016). At its peak in 2016, the Bureau of Prisons operated 15 of these facilities, housing roughly 22,600 sentenced prisoners (DOJ, 2016). A similar system was established in England and Wales in 2006, as part of the 'hubs and spokes' policy (Kaufman, 2013), which deliberately segregated foreignnational citizenss into specific prisons with a designated immigration presence. In 2012, the Norwegian Department of Justice's Correctional Service announced that the Kongsvinger prison facility would become the country's first dedicated all-foreign facility (Kriminalomsorgsdirektoratet, Letter 12/7475, 2012).
This integration of punishment and migration control is a shared feature of three national penal systems otherwise thought to exemplify different ends of the spectrum of criminal justice. Whereas Norway is considered an exemplar of 'Scandinavian exceptionalism', the United States and England and Wales are said to exemplify 'Anglophone penal excess' (Pratt, 2008;Pratt and Eriksson, 2013). Yet, the growth of research on the entanglement of criminal law and immigration control ( García Hernández, 2013;Stumpf, 2006) has led scholars to question this characterization of relative penal excess and moderation (Aas, 2014;Barker, 2013Barker, , 2018Franko, 2020). The emergence of all-foreign prisons in the United States (Bureau of Prisons, 1999;Kaufman, 2019), Norway (Ugelvik and Damsa, 2018;Ugelvik, 2013Ugelvik, , 2017 and England and Wales (Kaufman, 2015;Singh Bhui, 2008;Warr, 2016) has given rise to a growing body of research that has charted how the goal of deportation has been integrated into the prison setting.
Following Katja Franko's (2020: 87) call to investigate the 'numerous national variations of bordered penality and crimmigration control', in this article, we draw on original empirical research to develop a comparative analysis of all-foreign prisons in the United States, England and Wales, and Norway. Our analysis focuses on three areas of comparison: the conditions in which all-foreign prisons emerged, their position and function of in specific national systems of criminal justice and immigration control, and their operation within each context. The intention of our analysis is not to chart policy transfer or to claim specific causal factors led to penal convergence or divergence (Jones and Newburn, 2008), but rather to develop a more nuanced understanding of the emergence of what is variously called 'crimmigration' (Stumpf, 2006), 'bordered penality' (Aas, 2014;Franko, 2020) or 'penal nationalism' (Barker, 2018;Barker and Smith, 2021). This analysis may then help to nuance claims about the global, unilinear trend towards the mixing and intensification of criminal justice and migration control.
Our findings highlight the disjuncture between bordered penality in rhetoric and practice. In response to anxieties about 'foreign criminals' (BBC News, 2006a;Solheim, 2014) or 'criminal aliens' (United States Senate, 1995), each country shifted the focus of penal intervention from moral censure towards determining 'who has the right to be here' (Aas, 2014: 539). Yet the material practice of segregated punishment varies substantially by jurisdiction. In the United States, the scale of incarceration and deportation is far greater, and foreign nationals are explicitly excluded from many key prison services. All-foreign prisons in the United States are the only privately operated prisons in the Federal system and are not subject to the same legal regulations as other facilities run by the Bureau of Prisons (DOJ, 2016). By contrast, the scale of segregated punishment in Norway and England and Wales is much smaller and is held to the same legal standard as other penal facilities. Building on Vanessa Barker's concept of 'penal nationalism' (2018: 27), we argue that variations in the material conditions of all-foreign prisons can be partially explained by how policymakers defined and located public welfare services: whereas officials in Norway and the England and Wales employed all-foreign prisons to exclude non-citizens from social entitlement programmes, US officials framed segregated punishment as a tool to exclude non-citizens from prison services and rehabilitation programmes.

A bordered penality
The entanglement of crime control and migration control has received significant criminological attention, described variously as 'crimmigration', a 'bordered penality' or 'penal nationalism' (Aas, 2013(Aas, , 2014Aas and Bosworth, 2013;Aliverti, 2013;Barker, 2018;Barker and Smith, 2021;Bosworth, 2019;di Molfetta and Brouwer, 2019;Fekete and Webber, 2010;Franko, 2020;Guia et al. 2013;Stumpf, 2006;Ugelvik, 2017;Zedner, 2016). 1 In each of these descriptions, the integration of migration control and crime control has been explained as a response to periods of crisis that threaten state authority. Where public anxieties about migration have produced the perception of national crisis, border criminologists have argued that the integration of punishment and deportation works to re-assert the primacy of state authority and the privileges of formal citizenship (Barker, 2017(Barker, , 2018Bosworth, 2008;Franko, 2020;Stumpf, 2006). In her influential description of 'crimmigration law', Juliet Stumpf drew on membership theory and a Durkheimian account (Garland, 2001: 134-135) of the expressive function of punishment to argue that the integration of criminal law and deportation worked to define and reinforce the boundaries of national membership (Stumpf, 2006(Stumpf, , 2013. Building on this work, Katja Franko has described the emergence of a new 'bordered penality', defined by 'a differentiated, two-tier approach to criminal justice and a more exclusionary penal culture directed at foreign nationals' (Aas, 2014: 521;Franko, 2020;Ugelvik, 2017). Although the terms bordered penality and crimmigration (Stumpf, 2006) are increasingly used interchangeably, we use the term bordered penality here in an effort to highlight the specific forms of differentiation produced by a prison system that segregates prisoners on the basis of citizenship (Aas, 2014: 525-526). Indeed, the defining feature of this exclusionary penality is the combination of the 'power to punish and the power to banish': non-citizens are no longer to be returned to society, but instead expelled from the territory of the state (Bosworth, 2011;Franko, 2020: 54).
Although the Scandinavian states have been seen by some scholars to favour a more 'humane' approach to punishment stemming from 'cultures of equality' (Pratt, 2008;Pratt and Eriksson, 2013), others have described in detailed fashion how the differential penal treatment of foreign nationals is, in fact, key to the maintenance of nationally egalitarian welfarist regimes Smith, 2021: 1540;Smith and Ugelvik, 2017). In her work on 'penal nationalism', Vanessa Barker (2017: 442-443;2018: 79-80) has argued that new formations of bordered penal power have emerged to reaffirm state authority and the value of national membership in Nordic societies where the coherence of the welfare state is perceived to be under threat from migrants and asylum seekers. Rather than a shift towards Anglo-American penal trends, Barker (2017: 27) and others have argued compellingly that the merger of border control and penal power reflected the welfarist desire to protect public resources from the 'crimmigrant other' (Aas, 2014;Barker, 2018;Franko, 2020;Ugelvik, 2013).
The concepts of crimmigration law, bordered penality and penal nationalism share key theoretical features. Although these authors disavow Durkheim's functionalism, they draw on his description of punishment's unique ability to communicate moral values and reinforce social boundaries (Durkheim, 1902(Durkheim, /1983 to describe how the application of penal power in the context of migration control communicates specific exclusionary social orders (Barker, 2017;Franko, 2020). These accounts also draw on critical criminology (Hall, 1978) to explain how crises of governance (Simon, 1998) constrain and mobilize specific political responses to migration and crime. Border criminologists have also placed the criminalization, detention/incarceration and deportation of non-citizens within the context of European colonialism and North American settler colonialism, drawing attention to racialized, classed and gendered inequalities (Aas and Bosworth, 2013;Bhui, 2016;Bosworth et al., 2018;De Giorgi, 2010;Franko, 2020;Kaufman, 2015;Melossi, 2003;Parmar, 2018Parmar, , 2020Yuval-Davis et al., 2017;Shull, 2022).
Bordered penality has been most tangibly expressed through a new kind of penal institution: the all-foreign prison. Scholars have highlighted how these facilities work to define and reinforce the boundaries of citizenship and social welfare (Kaufman, 2015;Barker, 2018;Barker and Smith, 2021;Franko, 2020). Within all-foreign prisons in each country, scholars have described how 'foreign national prisoners are increasingly placed in separate institutions, and afforded different procedural treatment and standard of rights than citizens' (Aas, 2014: 525-526; see also Kaufman, 2015Kaufman, , 2019Shammas, 2016;Ugelvik, 2017). In Scandinavia, the all-foreign prison has been understood as a key technique of penal nationalism, working to remove 'unwanted' populations, and thereby preserve egalitarianism for the nation Smith, 2021: 1212). In England and Wales and the United States, the differential treatment of non-citizen prisoners has been linked to ethnic and racial discrimination (Bhui, 2007(Bhui, , 2016Kaufman, 2019) and the legacies of the post-colonial period (Kaufman, 2015).
Strikingly, each country implemented systems of segregated, all-foreign imprisonment in response to crises of governance (Simon, 1998) triggered by periods of sustained public anxiety about 'foreign criminals.' Drawing on the intellectual tradition of critical criminology (Hall, 1978;Simon, 2007) and recent work on 'migration crisis' (Dines et al., 2018), we approach these crises less as temporal conditions triggered by specific causes, and more as narrative devices that define and describe social phenomena in ways that constrain and mobilize specific political outcomes. Rather than ask why specific events led to crisis, we compare how periods of crisis enabled certain changes to penal orders and prevented others.
Although existing research has pushed forward the concepts of 'bordered penality', empirical questions about the extent to which bordered penality in each of these national contexts may have shared or distinct features remain open. Indeed, there has not yet been a sustained comparison of how similar institutions in the United States, England and Wales, and Norway emerged (although see Pakes and Holt, 2017). We draw on the analysis of the relationship between crises of governance and the communicative function of bordered penality (Aas, 2014;Barker, 2017Barker, , 2018Franko, 2020) to help explain all-foreign prisons in the United States, England and Wales, and Norway at the turn of the 21st century. This framework also helps us explain how the logics of bordered penality embedded in all-foreign prisons in each country reflect significantly different material conditions in practice.

Methods
This article is the product of three years of discussion and collaborative scholarship driven by the authors' shared interest in the treatment of foreign-national prisoners and the all-foreign prison system. Each of the authors has focused their doctoral research on all-foreign prisons in one of either Norway, England and Wales, or the United States, providing a unique combination of jurisdictional expertise and diverse points of view. During the initial phase of this collaboration, the authors noted points of continuity within their research: the emergence of 'crises' in each country in relation to foreignnational prisoners, the importance of race and gender to discursive constructions of these offenders, and the materialization of the all-foreign prison as the solution to these periods of 'crisis'. Because these three states are generally understood to be fundamentally different in contemporary penal studies, these points of continuity were the starting point for this comparative analysis.
Our analysis builds on the theoretical framework described in the previous section to examine and compare the emergence, consolidation and operation of the all-foreign prison in these three countries. Notably, Pakes and Holt's (2017: 73) comparison of allforeign prisons in England and Wales and Norway found that 'similar pressures have produced similar solutions in two contrasting West European prison systems. ' Following Nelken (2009) , our analysis centres on the agency of local political structures, describing how specific practices and policies emerged. Our goal is to develop a comparative method that helps us to identify cross-national similarities in the rhetoric and logic of bordered penality, while nonetheless attending to the many differences in the way all-foreign prisons were put into practice.
To compare how all-foreign prisons embody the rhetoric and logic of bordered penality, we draw on close textual analysis of legal documents, parliamentary debates and media coverage from each country. Before beginning archival research, we collaboratively established criteria for the selection of media and policy documents, focusing on sources using the terms 'criminal alien' or 'foreign criminal' (translated as kriminelle utlendinger in Norwegian). To identify sources, we each used archive search engines (ProQuest, Retriever and Newspapers.com) and drew on existing research (inter alia De Noronha, 2015;IMDi, 2014IMDi, , 2017Todd-Kvam, 2019). Echoing Nelken's (2009) scepticism of the possibility for comprehensive, like-to-like comparison, our analysis instead focuses on identifying key themes in the discursive constructions of non-citizen offenders and the all-foreign prison in the United States, England and Wales, and Norway. Nonetheless, our comparison of the practical operation of all-foreign prisons is grounded in the collaborative analysis of similar types of sources produced by national institutions in each country: legal texts, prison policy and administrative records, and state-produced prison population statistics.
To address questions of how all-foreign prisons operate in practice, we examine the legal and administrative regulations governing the treatment of foreign-national prisoners in each country, as well as audits, inspections and other monitoring records produced by official sources (see, for example DOJ, 2016; Her Majesty's Inspectorate of Prisons [HMIP], 2020). We also draw on fieldnotes, interviews and ethnographic observation conducted at North Lake Correctional Institution (CI), CI McRae and CI D Ray James in the United States, Her Majesty's Prison (HMP) Huntercombe in England and Wales, and Kongsvinger and Oslo prisons in Norway. In the United States, interviews were conducted via phone or by mail correspondence with 32 men in CAR prison custody during the period January 2021 to October 2021. In England, ethnographic fieldwork was carried out at HMP Huntercombe from April 2019 to October 2019. In addition to periods of immersive observation, the author interviewed 38 incarcerated people. In Norway, fieldwork was carried out at Oslo prison from August to December 2016, and at Kongsvinger from August to December 2017, conducting 30 interviews in total with foreign-national prisoners. These fieldnotes were further supplemented by interviews with prison administrators, legal practitioners, and non-government organization staff and grassroots advocates to situate the operation of all-foreign prisons within national structures of criminal justice and migration control. Totalling 31 in all countries, these interviews focused on day-to-day prison administration, the process of immigration proceedings in the penal setting, and the resources available to foreign-national prisoners in each facility.

The evolution of the all-foreign prison
In 1994, Texas Republican Lamar Smith, chair of the House of Representatives Subcommittee on Immigration, addressed the first of a series of high-profile hearings on a subject believed to be of urgent importance: 'criminal aliens' (House of Representatives Subcommittee on Immigration, 1997). Smith summarized the problem succinctly: 'criminal aliens are a drain on the American taxpayer while they are in prison, and when they get out of prison, they pose a significant threat to the public safety' (House of Representatives Subcommittee on Immigration, 1997: 2). Smith's summation highlights the shared features of public anxieties in the United States, England and Wales, and Norway: as an offender, the 'foreign criminal' (House of Commons, 2006) was framed as a threat to public safety; as a prisoner, the foreign-national offender was framed as an unwanted burden whose presence constituted a waste of public funds on incarceration and a threat to the welfare state.
criminal' (BBC News, 2006a;in Todd-Kvam, 2019: 305). These terms were novel labels developed to invoke and reflect public concern in specific ways. In the UK, the term 'foreign criminal' came to widespread use in 2006 during the 'foreign prisoner crisis' (BBC News, 2006ain De Noronha, 2015. With no legal definition, the term 'foreign criminal' was used to encompass a wide range of distinct categories of migrants and offenders (Daily Mail, 2006;De Noronha, 2015: 10). The term 'criminal alien', which came to prominence in the United States in the period after the Mariel Boatlift in 1980 (Government Accounting Office [GAO], 1986;Loyd and Mountz, 2018) has no specific legal definition. Like 'foreign criminal' the term 'criminal alien' is a shifting signifier, heavily shaped by racial hierarchy, working to justify who may be arrested, detained and deported (Chazaro, 2016: 659). Although Norway might be expected to differ from these examples, Thomas Ugelvik (2012: 67-68) and John Todd-Kvam (2019: 304) have both observed how depictions of foreign-national prisoners invoked the term 'foreign criminal' (kriminelle utlendinger) to define a novel category of subjects within the criminal justice system. Whereas Norwegian citizens are termed 'offenders', foreign nationals are frequently the one group to which the label of 'criminal' is applied in public discourse (Todd-Kvam 2019: 304; Kvittingen, 2015;Solheim, 2014). The rhetorical construction of the 'foreign criminal' relied on representations of specific types of migrants committing specific kinds of crimes. De Noronha (2015: 14) argues that these representations mobilized race and gender in particular to glue together migrantness and criminality in ways that made foreign-national prisoners 'ideal villains' (Christie, 1986).
'Foreign criminals' were portrayed as violent racial or ethnic minority men who victimized vulnerable people. Representations of the 'foreign criminal' in the UK focused on violent crimes, often involving the victimization of women and children committed by Black immigrants (De Noronha, 2015: 16). At the height of public concern about the failure to deport foreign-national prisoners in England and Wales in 2006, a Conservative Party statement to parliament focused on Mustaf Jama, a Somali asylum seeker suspected of murdering female police officer Sharon Beshenivsky (House of Commons, 2006). Similarly, US portrayals of 'criminal aliens' focused on spectacular depictions of violent crimes committed by Latino immigrants Getter and Alvarez, 1993;Means, 1997). At a 1997 Congressional Hearing, Republican Representative Elton Gallegly told the story of Jose Zavalla, a Mexican man who had been convicted of rape three times and returned to the community in California each time, rather than being deported (House of Representatives Subcommittee on Immigration, 1997: 61). In both countries, these stereotypes belied the fact that foreign-national prisoners were disproportionately incarcerated for drug or immigration offences (Kaufman, 2015(Kaufman, , 2019.
In Norway, police officials sought to emphasize that open borders and the low punitiveness of the system encouraged 'foreigners' to come and 'commit crime' (Solheim, 2014). This 'problem' 'worsened', following the expansion of the Schengen Area and the European Union/European Economic Area (EU/EEA) to Central and Eastern Europe (Hammerstad, 2013ab, Solheim, 2014. Following the 2007 enlargement of the EU to include Romania and Bulgaria, Norwegian press described an 'Eastern European crime wave [that] crashes over Norway' (News in English, 2009). During this period, migrant men from Romania, Poland and Lithuania, became 'the usual suspects' (Franko, 2020). Police officials developed specific policing strategies to target Central and Eastern Europeans for 'mobile organized crime groups' and 'pickpocketing' (Franko, 2020). Roma Romanians in particular captured the attention of authorities and the public, with several attempts being made to criminalize some of the survival strategies employed by the Roma, such as begging and living without formal housing (Friberg, 2020). These targeted policing strategies and tougher sentences, according to police and justice officials, were needed to deter 'tough Eastern European crime gangs' (Gregersen, 2012, Hammerstad, 2013a. In each of these cases, foreign-national offenders were portrayed as burdens on the system, who sought to use up welfare assistance when free, and wasted taxpayer money crowding prison systems when incarcerated. In the UK, a report by the National Audit Office (NAO, 2005) that would help to precipitate the 'foreign prisoner crisis' (BBC News, 2006a) foregrounded the extravagant cost to British taxpayers of continuing to incarcerate and support non-citizen prisoners who were not deported. In the United States, efforts to target 'criminal aliens' in Federal prison custody after 1994 were linked rhetorically to efforts to exclude immigrants from access to public services. As Florida Representative Tom Lewis put it to Congress in 1994, incarceration costs the taxpayer between $800 million and $1.5 billion keeping these criminals aliens in our overcrowded prisons. We clothed them, housed them, and fed them, we put them through the drug treatment and job training programmes to make them better citizens, then we deported them. (Lewis, 1994: 142) This rhetoric was most visible in Norway, where politicians rationalized the need to target foreign offenders specifically as a measure required to protect the solvency of the welfare state (Barker, 2018;Franko, 2020). In 2014, Norwegian police produced a report detailing the costs of 24 foreign nationals to the criminal justice system (e.g., police work, court case and days in prison), ranging from over a 100,000 to 2.8 million crowns; a day in prison was billed at 1800 crowns (Solheim, 2014). Progress Party politician Ulf Leirstein complained in 2013 that 'prisons that were meant to be reserved for Norwegian inmatesinmates who should have been rehabilitatedhave been inundated with foreign criminals' (Leirstein in Todd-Kvam, 2019: 305).

Three foreign-national prisoner crises
In the United States concerns about the 'criminal alien problem' (House of Representatives Subcommittee on Immigration, 1994) came to a head during the period 1994 to 2000. During the early 1990s, the Federal government's failure to identify, apprehend and deport foreign nationals released from Federal prison was portrayed as a grave threat to the safety of the American public. In the run-up to the 1994 Congressional elections in the United States, a wave of media coverage (Sontag 1994a(Sontag , 1994bBerger, 1994;Abrams, 1994) raised concerns about the release of violent 'criminal aliens' onto US streets. In turn, Republican politicians used this news to attack the Clinton administration's record on drugs, crime and immigration. Throughout Congressional hearings on the 'criminal alien' problem from 1994 to 1999 (United States Senate, 1995; House of Representatives Subcommittee on Immigration, 1994, 1997, Rabkin, 1999 foreignnational prisoners were portrayed as dangerous recidivists, likely to 'go back on the street to be arrested at least one more time' (House of Representatives Subcommittee on Immigration, 1994: 112).
Non-citizen offenders were also portrayed as a key cause of prison overcrowding, and a threat to state capacity. As Kentucky Democrat Representative. Romano L. Mazzoli put it, because of 'the growing numbers of deportable alien criminals … there are simply more felons than there are beds' (House of Representatives Subcommittee on Immigration, 1994: 3). After 1994, newly elected officials in California, Arizona, Texas and Florida sued the Federal government to recoup costs associated with incarcerating non-citizens (Abrams, 1994: 379). These suits argued that the Federal government's failure to sufficiently enforce border controls was the root cause of the 'criminal alien problem' (United States Senate, 1995). Therefore, it was argued, the Federal government should re-imburse State governments for public services used by unauthorized migrants, including incarceration. Although these suits were largely dismissed in court, sympathetic officials in Congress responded by enhancing the Institutional Hearing Program, a system of prison-based deportation courts formalized in 1986, and creating a Federal prison-subsidy programme called the State Criminal Alien Assistance Program (House of Representatives Subcommittee on Immigration, 1997; Eagly and Shafer, 2020).
Concerns about foreign-national prisoners in England and Wales reached a fever pitch in 2006 during what came to be known as the 'foreign prisoner crisis' (Daily Mail, 2006;De Noronha, 2015: 1). In 2005, the NAO published a report which found that immigration authorities were unable to provide details on a number of foreign offenders who had been released from HMP custody, but were not deported (NAO, 2005: 7). After weeks of pressure from Conservative MP Richard Bacon (Kaufman, 2013: 167;BBC News, 2006a) the Home Office published statistics showing that 1023 foreign offenders who had been recommended for deportation had been released at the end of their sentences (Travis, 2006). The British media framed these revelations as a 'crisis' of public safety and state incompetence (Daily Mail, 2006;Travis, 2006;The Independent, 2006). An article in The Telegraph from 2006 told of how Rashid Musa was 'released to rape and kill', focusing on how Musa re-offended after his release despite the fact that his application for asylum had failed . In an interview with the BBC, Conservative MP for Reading Rob Wilson claimed that there was a 'whole raft of… illegal immigrants' who were about to be released from prison custody in his constituency (BBC News, 2006b).
Although the policy discourse surrounding the release of foreign-national prisoners was more explicitly tied to the failure of the asylum system in the UK (NAO, 2005: 23), these releases were nonetheless understood as a failure of the state's custodial capacity to adequately control non-citizen offenders. The NAO report found that the directorate's criminal casework team was underfunded and inefficient, and regularly failed to meet removal targets for individuals in HMP custody (NAO, 2005: 24). A similarly scathing inquiry conducted by the Select Committee on Home Affairs (SCHA) in July 2006 portrayed the Immigration and Nationality Directorate as an agency that had failed on all accounts: allowing offenders to walk free, wasting public money through inefficiency, and mismanaging the use of detention in the cases in which prisoners were transferred to immigration custody (SCHA, 2006: 520).
In Norway, bordered penal populism dominated public debate in the 2000s (Todd-Kvam, 2019), with politicians, police and prison officials emphasizing that foreign nationals 'filled up' overcrowded prisons, while Norwegian citizens waited in the 'imprisonment queue' (Rønneberg and Solvang, 2006). Tore Leirfall, the spokesperson of the Correctional Service's Professional Association, reframed the targeted deportation of foreign nationals as a solution to limited prison capacity. At Ullersmo prison alone, a facility with a total capacity of 238 cells, he argued, 36 foreign nationals could be 'sent out' (Rønneber and Solvang, 2006).
Moreover, deportation was believed to reduce the costs experienced by the Norwegian criminal justice system. Leaders like police chief Atle Roll-Matthiesen routinely emphasized how 'criminal foreigners inflict enormous costs on the police, the court and the Correctional Service' and the only solution was expulsion (Solheim, 2014). In 2010, a scandal erupted over the opening of the 'world's most humane prison' in Halden, Norway, when it became apparent that 70% of the offenders in the prison were to be foreign nationals (Dagbladet, 2011in Ugelvik, 2012. Under pressure from the farright Progress Party, the Correctional Service swiftly announced that the 248 beds in Halden prison would be reserved solely for Norwegian citizens (Dagbladet, 2011in Ugelvik, 2012. This reaction symbolized the framing of foreign nationals as costly, unwanted burdens on the Nordic welfare state. In the years following the Halden debate, detailed police reports on the average 'price tag' of a 'foreign criminal' regularly made the news, as evidence of the need to expel 'foreign criminals' quickly to reduce their 'enormous' cost (Solheim, 2014).

The all-foreign prison as the solution to the crisis
Across the United States, England and Wales, and Norway, public anxieties about foreign-national prisoners produced perceptions of a crisis of governance (Simon, 1998). By narrating the relationship between crime and migration in specific ways, these periods of crisis proved critical to the emergence of a new form of bordered penality: the all-foreign prison.
In a primary sense, the all-foreign prison promised to prevent the possibility that foreign-national offenders would be released by seamlessly suturing together punishment and deportation. From 1986 to 1994, Federal prison officials experimented with the use of four privately contracted facilities in Texas and Oklahoma as prisons for 'deportable criminal aliens' (House of Representatives Subcommittee on Immigration, 1994: 167-168). Although these facilities were not specifically designated as all-foreign prisons, this process of segregation was designed to improve the efficiency of the process of prison-based deportation proceedings (Eagly and Shafer, 2020). During a 1994 Congressional Hearing, Immigration and Naturalization Service Commissioner Chris Sale suggested that this outsourced, all-foreign model was best to ensure that 'sentenced aliens are funneled into a single prison intake center where [immigration officials] can work most efficiently' (Sale, 1994: 174). The 'hubs and spokes' (Kaufman, 2013) policy implemented within the prison system in England and Wales in 2006 followed a similar logic. A report produced by the Prison Service and the UK Border Agency in 2009 detailed how HMP Canterbury and HMP Bulwood Hall had been re-designated specifically for prisoners subject to deportation, in an effort to reduce the possibility of release, and to 'maximize the capacity of the detention estate by holding some ex-prisoners as immigration detainees' (Ministry of Justice andUKBA, 2009: 9 in Kaufman, 2013).
In 2011, in Norway, amid rising pressure from the far-right Progress Party, Norway's 'Red-Green' coalition government announced that the Department of Justice's Correctional Service would be concentrating foreign-national prisoners in a wing of Ullersmo to improve the efficiency of the process of post-sentence deportation (Justis og politidepartementet, Prop. 1 S, 2011-2012). A year later, the same officials designated the entire Kongsvinger prison as an all-foreign facility. Much like all-foreign prisons in the United States and England and Wales, the logic of bordered penality was clear: foreign nationals were a burden on the system, 'to be deported' and not 'returned to Norwegian society', so Kongsvinger was to function as a one stop deportation prison to ensure the 'fast' and 'effective' deportation of 'foreign criminals' (Justis og politidepartementet, Prop. -1 S 2011-2012).
All-foreign prisons also promised to reduce public spending by substantially cutting the cost of incarcerating foreign-national prisoners. In the United States, Federal officials used the fact that 'criminal aliens' were to be deported to rationalize excluding all foreignnational prisoners from access to critical prison services and regulations under BOP programme statements (Kaufman, 2019). After experimenting with private contracts for more than a decade, the first two formal all-foreign CAR prisons were procured through private contracts in 1999 (Bureau of Prisons, 1999). For the Clinton administration, prison privatization offered to reduce overcrowding in Bureau of Prisons-run facilities and cut the burdensome costs of incarceration (National Performance Review, 1994).
In England and Wales, where anxieties about 'foreign criminals' focused on the concern that this group of prisoners were overwhelming the state's carceral capacity by abusing systems of humanitarian protection (De Noronha, 2015), policymakers framed 'hubs and spokes' as a way to speed up the deportation process and reduce time spent in custody (HMIP, 2006;Bhui, 2007;Kaufman, 2012). In Norway, efforts to implement all-foreign prisons were framed explicitly as an effort to protect precious welfare resources from the burden of foreign criminals (Franko, 2020;Todd-Kvam, 2019). If the 'enormous cost' incurred on welfare services and prison budgets by these undeserving foreigners was widely publicized in the Norwegian press, the priority deportation schemes embedded in places like Kongsvinger aimed to protect the public by getting rid of the kriminelle utlendinger as rapidly as possible (Solheim, 2014).
In practice, there is scant evidence that all-foreign prisons in the United States, England and Wales, and Norway have lived up to their promise as the solution to the 'criminal alien problem' (DOJ, 2016). Emma Kaufman's analysis of all-foreign prisons in the United States and the UK suggests that these facilities have had a limited effect on deportation rates (Kaufman, 2015(Kaufman, , 2019. Rather than reducing the number of 'foreign criminals', lower thresholds for sentencing and deportation in Norway have increased the number of foreign nationals in prison (Franko, 2020). Audits conducted by the US Department of Justice (DOJ) found that all-foreign facilities experienced a significantly higher number of incidents related to safety, understaffing and medical care (DOJ, 2015(DOJ, , 2016. Advocates in the UK and the United States have rightly raised significant concerns about inequality and potential rights violations within these facilities (Webber 2009).
That the logic of penal nationalism was so central to the turn towards all-foreign punishment in each country further confounds expectations of the relationship between welfarist institutions, immigration, and punitiveness (Lacey, 2008). It is not surprising that officials in Norway, another 'coordinated market economy' (Lacey, 2008: 144) with strong welfarist institutions, relied on the logic of penal nationalism to rationalize the segregated incarceration of foreign-national offenders. It is more surprising, however, that concerns about the security, capacity and stability of public resourcesthe hallmark of penal nationalism (Barker, 2018: 43-47)were so central to the shift towards all-foreign punishment in the United States. Indeed, reports issued by the Norwegian police on the 'price tag' of a 'foreign criminal' (Solheim, 2014) are remarkably similar to estimates of the cost of incarcerating 'criminal aliens' included in the lawsuit filed by California Governor Pete Wilson in 1994 (Abrams, 1994). On the one hand, this suggests that even in post-Fordist contexts like the United States, rhetoric that scapegoated migrants as a threat to public services was remarkably politically powerful (Calavita, 1996). On the other hand, as we explore further in the following section, this suggests that political leaders in the United States, England and Wales, and Norway defined public services in radically different ways.

Operating the all-foreign prison
Although all-foreign prisons share the logics of bordered penality, they are embedded in specific national systems of criminal justice and migration control. In the United States in 2018, 66,429 people were transferred from prison custody to immigration detention under Immigration and Customs Enforcement's Criminal Alien Program (TRAC, 2018). By comparison, in England and Wales 4971 offenders were transferred from prison custody to immigration custody, and 80 were deported from Norway, of which roughly one-third were first transferred to a detention centre (Kriminalomsorgsdirektoratet, pers. comm., 2022). More broadly, the US prison population rate is nearly five times the rate in England and Wales, and roughly eleven times the rate in Norway (World Prison Brief, 2021).
Although the all-foreign prison operates, at least nominally, to facilitate deportation, the way migration control is practised within these prisons varies significantly in each country. In Norway, the division of labour between prison and deportation is relatively defined (Ugelvik and Damsa, 2018: 1034-1036. The integration of the deportation process at all-foreign facilities involves an established division of labour between the Correctional Service, the Police's Immigration Service and the Directorate of Immigration. In theory, all-foreign facilities in England and Wales should have a similar division of labour: HMP staff are primarily responsible for prison custody, and facilitate the work of a dedicated team of on-site immigration officials. In practice, however, the boundaries between prison work and immigration work are blurred: at HMP Huntercombe, prison staff regularly serve prisoners with immigration and deportation paperwork, and HMP employees often complain about increasingly 'bureaucratic immigration duties' (HMIP, 2021). In US all-foreign prisons, the process of migration control is remarkably hollow and immigration officials are almost completely physically absent. The majority of staff with whom prisoners interact are employed by contractors and have no role in the deportation process. Indeed, immigration officials rarely appear in these facilities. After Federal criminal courts register that a prisoner is a non-citizen, that information is shared with a central Immigration and Customs Enforcement (ICE) liaison, who determines whether foreign-national prisoners are eligible for a court hearing, or whether they are subject to extrajudicial 'speed deportation' (Wadhia, 2014). During the period from 2014 to 2018, only 34% of prisoners in CAR facilities were determined by ICE to be eligible for a court hearing, and only 7% of this group avoided deportation (TRAC, 2018).
Although each system is premised on the notion that all foreign nationals will be deported after they finish their criminal sentence, the actual likelihood of this outcome depends significantly on jurisdiction. In the United States, 96% of prisoners in all-foreign facilities during the period from 2014 to 2018 were ordered removed (TRAC, 2018). By contrast, deportation rates in both England and Wales and Norway are significantly lower (Ministry of Justice and UKBA, 2009: 9;Kaufman, 2015: 126). In the UK, 3837 offenders were transferred from prison custody to immigration custody in 2020, of whom 67% were eventually released from immigration custody on bail or granted immigration relief. In the calendar year January to December 2020, only 26% of foreign nationals in UK Immigration Removal Centres were removed, whereas the rest were released from detention (AVID, 2020). By comparison, in Norway, in total, in 2020, 17% of the 470 offenders were deported or transferred abroad (Kriminalomsorgsdirektoratet, pers. comm, 2022). Although Emma Kaufman has argued persuasively that all-foreign prisons in the United States have not meaningfully improved the efficiency of deportation (Kaufman, 2019(Kaufman, : 1385, the relative efficiency of the deportation of foreign-national offenders in the United States compared with England and Wales and Norway is nonetheless stark. This comparison highlights the need for a better understanding of the legal and administrative practices that have enabled mass deportation (Gibney, 2008).

Differential treatment in all-foreign prisons?
The material consequences of differential treatment vary significantly by country. Whereas conditions in all-foreign prisons in Norway are materially identical to those for citizens, all-foreign facilities in England and Wales often face unique constraints on funding and programming. By contrast, foreign-national prisoners in the United States are explicitly excluded from a wide range of prison services, and all-foreign facilities are held to a significantly lower regulatory standard than Bureau of Prisons-run facilities. Notably, the all-foreign prison system in the United States is entirely privately operated (DOJ, 2016), shaping practices and conditions of confinement in critical ways.
In the United States, foreign-national prisoners have been legally excluded from many of the prison-based services afforded to citizens. 2 These exclusions were compounded by the fact that all-foreign prisons are the only privately operated prison facilities in the Federal prison system. Analysis of contract documents from eight all-foreign facilities showed that none contained language holding contractors to the same rules applied to Bureau of Prisons-run prisons. 3 This lack of regulatory specificity has led to alarming levels of understaffing, overcrowding and a number of deaths in custody resulting from poor medical care (DOJ, 2016;Wessler, 2016). Although poor conditions and deaths in custody are endemic throughout the Federal prison system (Guernsey, 2021), for many men in CAR custody, these conditions exacerbated the impact of bordered penal exclusion in all-foreign prisons (No Detention Centers in Michigan, 2020).
Although foreign-national prisons in both England and Wales are held to the same legal standard as prisons for citizens, in practice foreign-national prisoners are often exposed to significant material differences. Foreign-national prisoners in England and Wales are not legally excluded from the services or programmes afforded citizens in other facilities, but in practice all-foreign prisons tend to face greater budgetary and resource constraints (Kaufman, 2013). Often prisoners' sentences required them to take part in rehabilitative programmes that are not available to them within an all-foreign facility, which in turn impacts their future immigration claims (HMIP, 2006(HMIP, , 2017(HMIP, : 14, 2021. Moreover, the resources available in all-foreign facilities were often of little use, exacerbating the impact of exclusion. An audit of HMP Huntercombe in 2017 found the provision of work and exercise activities to be either irrelevant or insufficient, and noted that only 35% of prisoners had received support for planning their life after release (HMIP, 2017: 14). A similar inspection conducted in 2021 found that there was no dedicated rehabilitation provision, and little specialist support relating to life after deportation (HMIP, 2021: 23-24).
By contrast, foreign nationals in Norway had access to all the welfare provisions afforded to any other Norwegian prisoners (fieldnotes). Like prisoners who were Norwegian citizens, foreign nationals also had a 'contact officer' to help them with planning for their release, despite the fact that, in some cases, they would be deported. Despite the rhetoric of burdensome foreign nationals, the Norwegian Corrections Service explicitly stated that all-foreign prisons would operate no differently to other prisons (Kriminalomsorgsdirektoratet Letter 12/7475, 2012). Although all-foreign facilities offered a more restrictive regime than other Norwegian prisons, with regards to leave or activities outside the prison, officers sought to counter the stricter regime by providing alternatives; and developed skills to address the specific needs of foreign-national prisoners (fieldnotes; see also Mulgrew, 2016 ). The prison followed what Ugelvik (2017: 185) calls 'a double vision', hosting 'a welfare-oriented system with inclusion as its fundamental logic' but 'accompanied by an alternative substitute system where exclusion is the desired end' with regard to foreign nationals.
Although these national prison systems share logics and symbolic meanings, our comparison highlights how the material terms of exclusion differ substantially in each national context. These differences in treatment can be partly explained by how national policymakers defined public services: in Norway and England and Wales framed incarceration as a tool for protecting the welfare state, rather than a public service in and of itself. In the United States, by contrast, officials like D'Amato, Smith and Gallegly clearly defined incarceration as a public service in and of itself. In both cases, penal nationalism worked to frame as a threat to state capacity. Yet, whereas penal nationalism in Norway worked to protect the capacity of social entitlements, in the United States it worked to protect the capacity of the carceral estate. In the United States, the formal legal exclusion and privatization of all-foreign facilities reflected a logic of penal nationalism internal to the Federal prison system. The outsourcing of all-foreign prisoners, and the formal legal exclusion of foreign nationals from key prison services in the United States constitutes a kind of differential treatment intended to protect the capacity of the 'public' prison system. By contrast, the absence of material differences between all-foreign prisons and the rest of the penal estate in Norway belies the broader symbolic function of penal nationalism. Post-colonial scholars have noted how intertwined colonial, nationalist, racialized and gendered ideologies are embedded within the Nordic welfare state and continue to be projected onto different groups in the region, especially non-citizens (Barker, 2018;Franko, 2020;Keskinen et al., 2009;Loftsdóttir and Jensen, 2012). In this context, the segregation of penal subjects according to citizenshipand in practice, race and ethnicity represents a culmination of this logic, ensuring that white Norwegians retain privileged access to welfare resources external to the prison walls. Although the 'cultural nonmemory' of colonialism facilitates the view of exceptional Nordic identities as anti-racist, gender-equal and classless, the contemporary Nordic welfare state is no exception (Keskinen et al., 2009;Loftsdóttir and Jensen, 2012).

Race, nationality and the evolving demographics of all-foreign prisons
All-foreign prisons, as Emma Kaufman has observed, alter the meanings of race, ethnicity and nationality in prison (Kaufman, 2015: 8). Our comparison highlights differences in the racial and national demographics of all-foreign prisons, and the process through which deportation and punishment are integrated in the prison setting. Although a multitude of nationalities are represented in all-foreign prisons in each country, US facilities are more homogenous in racial and national terms than all-foreign prisons in England and Wales and Norway. Nearly 90% of prisoners in US CAR facilities are Latino, mainly nationals of Mexico, Honduras, El Salvador and Guatemala (TRAC, 2018;Eagly and Shafer, 2020: 816). In both England and Wales and Norway, EU expansion since 2004 has changed the racial and national demographics of foreign-national prisons. In England and Wales, Polish prisoners and Albanian prisoners have accounted for the largest national group in recent years (House of Commons Library, 2021: 14). After the first EU enlargement in 2005, 11.5% of Norway's prisoners were foreign nationals, with Iraq, Poland, Lithuania and Somalia making up the four largest groups. The percentage of prisoners who are foreign nationals peaked at 34%, in 2017, with Poland, Lithuania, Romania and Sweden making up the four largest groups (Kriminalomsorgen, 2018).
These statistics highlight how Eastern European migrants have become an increasingly important racialized and national 'other' in both Norway and England and Wales. This suggests a shift to xeno-racism in addition to more traditional racism (Fekete 2001;Weber and Bowling 2008). Xeno-racism highlights and reasserts hierarchies of belonging in England and Wales and Norway, where national identity is prioritized and outsiders are increasingly excluded and demonized (Bhui, 2016;Franko, 2020;Ugelvik, 2013). In the UK, in the years coming up to and following Brexit, EU migrants became synonymous with criminality. One group of MPs highlighted how Polish prisoners had taken prime position in the 'league for foreign inmates' (Duell, 2014, ). This elision between race and national origin has also occurred in Norway, exemplified by the police projects targeting Eastern Europeans described in the previous section. Compared with nationals of other Nordic countries, nationals of countries like Romania and Lithuania can only attain 'precarious membership' (Franko, 2020), as their membership can be quickly rescinded when considered a criminal or security threat.

Conclusion
Seeking to answer Katja Franko's (2020) call to empirically explore the dynamics, processes and practices of a bordered penality within specific contexts, this paper has compared all-foreign prisons in Norway, the US, and England and Wales. In doing so, it has demonstrated the similarities in the logics of bordered penality across each jurisdiction. Whereas constructions of the 'criminal alien' or the 'foreign criminal' displayed the rhetoric of penal populism in of the three countries, local discourses of 'race', 'ethnicity' and gender played key roles in gluing together foreign-ness or migrant-ness and criminality in different ways. In both Norway and England and Wales, in all-foreign prisons, the intersection of racialization and ethnicization gave rise to distinct forms of xeno-nationalism (Fekete 2001;Weber and Bowling 2008), whereas in the United States it resulted in the racialization of non-citizenship (Kaufman, 2019).
In each country, 'foreign criminal' men were not only portrayed as a threat to public safety, seen as likely to reoffend, but also to the welfare state, as an unwanted burden who wasted public funds on overcrowded prisons. Despite notable differences in the scope and scale of national welfare institutions, our comparison found shared logics of penal nationalism (Barker, 2018) at work in each country. In Norway, England and Wales, and the United States, excluding foreign nationals from public services was a key rationale behind the segregation and targeted deportation of foreign prisoners, despite significant variations in the welfare system in each country (Lacey, 2008). This article has also highlighted the difference between symbolic values and practical arrangements: although all-foreign prisons consistently worked to reinforce the symbolic value of membership through penal exclusion, the material consequences of this exclusion varied substantially by jurisdiction. Unsurprisingly, the scale of all-foreign prisons and the broader prison-to-deportation pipeline in the United States dwarfs similar systems in England and Wales and Norway. Differences in the material conditions of segregated punishment in each country can be partially explained by different definitions of public services. Whereas officials in England and Wales and Norway framed the penal system as a tool to exclude foreign nationals from public services, officials in the United States framed prison as a critical public service in and of itself (Abrams, 1994;GAO, 1997GAO, , 1998. In Norway and England and Wales, bordered penality reflected a concern that the growing 'price tag' of 'foreign criminals' would threaten state spending on social services (Solheim, 2014). By contrast, in the United States the growing cost of punishing non-citizens was understood first and foremost as a threat to the apparatus of mass incarceration itself (House of Representatives Subcommittee on Immigration, 1997). In this sense the differential treatment of foreign-national prisoners and the privatization of all-foreign prisons worked to protect the capacity of the 'public' prison system. The logic of a bordered penality observed in Norway, the United States, and England and Wales reveals the repressive and exclusionary nature of citizenship within these societies (Bhambra, 2015). The incarceration, detention and deportation of those deemed not to belong in any of these 'nation'-states is not new (Franko, 2020;Kaufman, 2019;Weber and Bowling, 2008) and from this perspective it is perhaps unsurprising to find these three jurisdictions together in comparison. What is novel is the emergence of an 'abnormal' administration of justice (Aas, 2014;Fraser, 2008) that may be at odds not only with contemporary aspects of punishment in Norway, but also with the principles of democracy in all three countries (Bosworth, 2021).

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.