The ‘rent a cell principle’ is a common way of dealing with prison overcrowding in the USA, where agreements between states facilitate the transportation of prisoners from one state to another (Christie, 2000). In Europe, these collaborations between nation states were unheard of until 2010, when Belgium and the Netherlands started the so-called “Nova Belgica” co-operation. The agreement permitted the Netherlands to relocate prison capacity (staff included) to Belgian prisons in return for a yearly fee. The partnership was a solution to two problems. First, Belgium has struggled with prison overcrowding for decades (Beyens et al., 1993). The agreement with the Netherlands offered an additional prison capacity of 500 (and later 681) places in Tilburg prison from 2010–2016 (Beyens & Boone, 2013). Secondly, the prison population in the Netherlands was decreasing, so the deal filled the Dutch prison, providing work for their prison officers (Liebling et al., 2021). Norway followed Belgium’s example in renting prison places from the Netherlands, but for reasons other than overcrowding. In Norway, the backlog of sentences had grown to over one thousand in 2013 and 2014, leaving many convicted prisoners ostensibly free in “prison queues”. To reduce the prison queue, Norway made an agreement with the Netherlands to rent vacant prison cells (Todd-Kvam, 2019). Between 2015 and 2018 a total of 650 prisoners convicted in Norway were sent to the Dutch prison of Norgerhaven to serve their sentences (Liebling et al., 2021:43).

Like Belgium, Denmark faces problems with both overcrowding and staff shortages (Danish Prison and Probation Service, 2021). In an internal negotiation note about the economy of the Danish prison system, it was estimated that Denmark would have a deficit of over 1000 prison cells in 2025 (Jensen & Jørgensen, 2021). By 2020, 368 prisoners due to be deported when their sentences expired occupied places in the Danish prison system (Danish Government et al., 2021, p. 6). This group of foreign national prisoners without residence permits and deportable under migration law have for years acted as a red rag for Danish politicians who favour a restrictive immigration policy. Different strategies have been implemented to force so-called “criminal foreigners” to leave Denmark as soon as possible with measures such as early release and repatriation after serving half of their sentence (Law no. 628, 2013) and limited access to education, employment, and services (Law no. 429, 2017). However, very few deportable foreign national prisoners have accepted deportation either during or after their sentences. To overcome prison overcrowding, and at the same time remove imprisoned non-Danish nationals for whom deportation was a condition of their sentence, in December 2021 the Danish government agreed to rent 300 prison cells from Kosovo. The Social Democratic Minister of Justice claimed the agreement would ease prison overcrowding and, in addition, send a clear signal that deportable prisoners should leave Denmark (Danish Government et al., 2021, p. 2). The law was implemented on 21 June 2022 (Law no. 893, 2022), and the deportation of Danish prisoners to Kosovo is planned to begin in 2023 (Danish Government et al., 2021, p. 7). However, on 19 September 2022, the Danish Trade Union for Prison Officers announced that the agreement was dormant due to a lack of finances (Danish Trade Union of Prison Officers, 2022). On the same day, the Danish Minister of Justice claimed that the rebuilding of the Gjilan prison in Kosovo was experiencing delays, and that consequently the agreement with Kosovo had not been legally ratified and would be slightly delayed. Nevertheless, intervention from the Danish Minister of Justice eventually made sure that the Danish prison in Kosovo would become a reality, albeit later than expected (Hagemann-Nielsen, 2022).

This article contributes to the field of comparative penology in Europe (e.g. Sparks, 2001; Cavadino & Dignan, 2006; Brodeur, 2007; Laursen et al., 2020; Ievins & Mjåland, 2021; Crewe, 2021) by drawing on experiences from Belgium and Norway, as well as Danish political debate and legislation. We explore the problems that arise when nation states rent prison cells abroad for deportable foreign national prisoners.

The policy of renting prison cells abroad can be criticised in different ways. First, when national governments engage in rental of prison capacity in foreign countries it raises concerns about the “commodification of prisoners”, where prisoners are both incapacitated in prison and mobile as a commodity between national states where they are warehoused as cheaply as possible (Pakes & Holt, 2017, p. 90). Second, the target group to be transferred are treated unequally if they experience a different prison regime (Todd-Kvam, 2019, p. 296). Thirdly, legislation and jurisdiction for both the sending and receiving states are complex matters, which heighten concerns about the violation of prisoners’ human rights (Sivilombudsmannen, 2016). In addition, not all prisoners volunteer to be transferred to another country to serve their sentence (Liebling et al., 2021; Sivilombudsmannen, 2016). This raises concerns about the selection criteria for being transferred and how transfers are carried out.

The methodological approach consists of document analysis, which is particularly applicable to qualitative case studies producing rich descriptions of a single phenomenon (Yin, 2014). The documents include background papers, law text, political program proposals, organisational reports, survey data reports, scientific articles, reports from control and monitoring bodies, newspapers, and press-releases. All documents have been reviewed in order to elicit meaning and gain understanding of national reasoning and principles of renting cells abroad (Bowen, 2009). The languages spoken by the authors allowed them to analyse these documents in the original languages (i.e., Danish, Norwegian, Dutch, French or English).

The article is structured in the following way. We initially explore the extant research about overcrowding problems in (Danish) prisons considering the various perspectives that emerge in the literature. Following this, we outline and place in comparative perspective the three agreements we are attentive to: “Nova Belgica” (Belgium/The Netherlands), “Norgerhaven” (Norway/The Netherlands), and “Gjilan” (Denmark/Kosovo). Thereafter, we explore the experiences of renting prison cells relative to legislative expectations. Finally, the implications of easing overcrowding problems in prisons by renting prison cells are critically explored within the crimmigration debate.

Like many other European countries (Penal Reform International 2021, p. 9) the prison population in Denmark has increased during recent years, bringing heightened occupancy rates and problems of overcrowding (Danish Committee on Legal Affairs, 2021). In an internal negotiation note about the economy of the Danish prison system, Demark is estimated to have 5,250 prisoners in 2025, which will leave the system short of more than 1000 prison cells (Jensen & Jørgensen, 2021). Figure 1 illustrates the development of the occupancy rate in the period between 2010 and 2021, contrasted with estimated figures (in black) between 2022 and 2025.

The data show that between 2010 and 2014 the occupancy rate in the Danish prison system was relatively stable, with the drop in 2015 explained by a reduction of unsuspended sentences and fewer remand prisoners (Danish Prison and Probation Service, 2016, p. 5). Since 2016 the occupancy rate has increased. Notwithstanding an overall reduction in recorded crime during this period, Danish prisons have witnessed a widespread and growing problem of overcrowding. Some of the overcrowding can be explained by the combined effects of longer sentences, a reduction in the number of prisoners being paroled, and fewer alternative sanctions such as community service (Minke, 2021). It is also worth mentioning that Denmark was among a small group of European countries (including Andorra, Denmark, Greece and Sweden) that experienced a four percent growth in the prison population during the Covid-19 pandemic in 2020 (Aebi & Tiago, 2020, p. 2). This occurred despite a short period of ceasing to summon prisoners to avoid Covid-19 from spreading in Danish prisons (Storgaard & Minke, 2022).

Prison overcrowding constitutes a serious problem. Research has revealed that it undermines the well-being of inmates and is a strong predictor of heightened suicide risk (Huey & McNulty, 2005, p. 508). Overcrowded cells are found to positively correlate with increased instances of rioting and unlawful protest (Useem & Reisig, 1999) and crowding in general is a significant predictor of misconduct among prisoners (Wooldredge et al., 2001) – especially among young inmates (Franklin et al., 2006). Meško et al. (2011) found that prison overcrowding in Slovenian prisons led to a lack of privacy, fewer hygiene standards, lower employment, and fewer free-time activities. In addition, it let to a higher number of conflicts between prisoners and attacks on prison staff (p. 407). In 2013, the United Nations published a handbook on strategies to reduce overcrowding in prisons, highlighting several problems:

In 2016, the European Committee on Crime Problems published a white paper on prison overcrowding stating that prison overcrowding could lead to “violating Article 3 of the ECHR [European Convention on Human Rights] because of overcrowded and insanitary conditions which facilitate or lead to inhuman or degrading treatment” (European Committee on Crime Problems, 2016, p. 26). As Hans Toch puts it, “the most serious consequence of crowding is warehousing, which creates a prison climate that prevents inmates from serving time in customary ways” (Toch, 1985, p. 58). In sum, a significant body of research stresses that overcrowding constitutes major problems for both prisoners and prison staff – explaining why nation states use different strategies to ease it.

The variety of political responses to prison overcrowding in Europe usually ranges from reductionist to expansionist government approaches towards imprisonment. Reductionist approaches exclude the construction of additional capacity, and instead seek to reduce overcrowding by using front-door and back-door strategies such as reducing criminal penalties, using community sanctions (instead of imprisonment), increasing suspended sentences, cutting the length of sentences, and increasing early release (Beyens & Maes, 2020). A reductionistic approach is also used to place sentenced people in a prison queue until there is vacant capacity in the prison system for them to serve their sentence. This, however, is arguably problematic. Some suggest that the gap reduces the preventive effects of punishment, notably the idea that offences should be met with a quick response. In addition, if a punishment is not executed immediately, the public might lose faith in the criminal justice system whilst convicted persons can feel as if their life is “on hold” (Laursen et al., 2020).

In contrast, the expansionist approach seeks to open additional prison capacity. Renting prison cells abroad fits within an expansionist view but is a very particular form of expansionism that entails challenges to prisoners’ legal certainty. How agreements on renting prison cells abroad are framed, and which encounters they face, is the focus of the following section.

The “Nova Belgica” agreement between Belgium and the Netherlands, in force between 2010 and 2016, was the first of its kind in Europe, while the “Norgerhaven” agreement between Norway and the Netherlands existed from 2015 until 2018. The “Gjilan” agreement between Denmark and Kosovo was signed in December 2021 and was slated to begin in 2023.

In the Nova Belgica agreement, the selection criteria for being transferred were restricted: male inmates who would serve a minimum prison sentence of one year. However, in reality, the majority of prisoners served a sentence of five years or more. In October 2011 this was 60 percent of the population (CPT/Inf (2012)19). Moreover, prisoners could not possess Dutch nationality, nor have an existing judgment of expulsion against them from the Netherlands. Prisoners could not be the subject of an arrest warrant, be under suspicion of criminal activity in the Netherlands, or pose a flight or security risk that could not be routinely met by Tilburg prison.

The Norgerhaven agreement stated that only male adult prisoners serving a longer sentence could be transferred. All prisoners subject to transfer should have started serving their sentence in a Norwegian prison and had to be transferred back two months before they were due to be released (Sivilombudsmannen, 2016, p. 43). Dutch nationals or prisoners expelled from the Dutch territory could not be transferred. Moreover, no transferred prisoners could be the subject of criminal investigation in the Netherlands, nor could they be sought for prosecution or the execution of a sentence. The prisoner should not pose a risk of escape, or a risk to society that exceeds the security level of the prison as determined by the Netherlands. In addition, the prisoner should not be in need of medical or other care which could not be provided in prison (Stortinget, Justis- og beredskapsdepartementet 2015, article 8). Finally, the Dutch government did not wish Norgerhaven to be overrepresented by foreign nationals and requested a representation of Norwegian prisoners (Liebling et al., 2021).

In contrast with the Nova Belgica and Norgerhaven agreements, the Danish Government and the contracting parties agreed that the target group to be transferred to Gjilan (Kosovo) would be deportable foreign national prisoners who will be deported from Denmark when their sentence is spent. As underlined in the multi-year agreement for the Danish Prison system, the purpose of a deportation sentence is that the convicted person must leave Denmark. Consequently, resources such as those aimed at reintegrating prisoners into Danish society should not apply to prisoners who are going to be deported (Danish Government et al., 2021, p. 22). This approach echoes a new rule in the Danish Prison Act, which states that “an inmate who is expelled from Denmark by judgement, cannot be employed in education, teaching or program activities unless certain circumstances speak for it” (Danish Prison Act no. 1333, 2019, § 38, 2.). In this case the Danish Prison Act explicitly advocates for, and requests, inferior treatment for prisoners who will be expelled. In addition, prisoners without children in Denmark will be prioritised over those with children for speedy deportation (Danish Government et al., 2021, p. 7). Finally, those convicted of terrorism, known war criminals, the terminally ill, and those who need treatment for mental disorders, will not be transferred to the Gjilan prison (Law no. 893, 2022, article 17, 4).

The previous collaborations contained no formal agreement for the transfer of foreign national prisoners. However, based on information retrieved from Tilburg prison in August 2015, more than half (52 percent) of the transferred prisoners transferred from Belgium had an unknown residence status (Vanhouche, 2022). While there was no formal aim to transfer non-Belgian nationals to Tilburg prison, the data show that this group was disproportionally selected for a transfer. Despite the Dutch government’s express preference for Norway not to solely house foreign nationals in Norgerhaven prison, a large proportion of the transferred prisoners were not naturalised Norwegians. It is worth acknowledging that some foreign national prisoners in the Norwegian system opted for the transfer, because they would have access to Skype calls with friends and family and longer telephone time (Liebling et al., 2021, p. 49). Similarly, some prisoners transferred to Tilburg preferred it there to the Belgian prison they had come from.

While the official discourse in the previous collaborations did not focus on the transfer of (deportable) foreign national prisoners to transnational prisons, the reality showed that this group was specifically selected for transfer. In the Belgian case, this was legitimised with reference to the limited ties this group had to Belgium, and the lack of reintegration measures they would need relative to those returning to Belgium. The Danish proposal takes this a step further with clear and very explicit rhetoric that justifies the transfer only of those due to be deported. This group of prisoners has become increasingly vulnerable to exclusion in what is best described as a “continuing transformation from human prisoner to commodity” (Pakes & Holt, 2017, p. 91). In addition, it is a far-reaching example of the Janus-faced penal policies in the Nordic countries (Barker, 2012). In this respect, certain groups of people are denied individual rights because an ethno-cultural conception of citizenship is used to justify how one group deserves an inferior prison regime (Todd-Kvam, 2019). Additionally, Barker & Smith (2021) use the term penal nationalism to highlight how criminal justice tools are used in Denmark to protect national interests. While their research mainly focuses on the way in which migration is punished as an offence, our research focuses on how foreign nationals convicted of a criminal offence are being transferred and exposed to an inferior prison regime. It is no surprise that some prisoners justifiably refuse to be transferred to an organisationally uncertain transnational prison.

The very fact that Belgian law was declared applicable in a penitentiary on Dutch soil made the agreement between Belgium and the Netherlands in 2009 somewhat peculiar. While the federal government considered its applicability unproblematic by acknowledging Tilburg prison as an appendage of the Belgian prison of Wortel, the Flemish community who organised social and welfare services for Flemish prisons in Belgium (including Wortel prison) did not provide an equivalent offer in Tilburg prison because it was not located within an area of its jurisdiction. While the Dutch staff consequently decided to develop a similar offer, it became clear that some important services remained absent.

A further concern related to the problem of jurisdiction in cases where a crime was committed in prison. The Belgian Council of State stated this concern as follows:

This means that, in addition to being transferred to another country against their will, new infractions could lead to a new conviction and consequent punishment in the receiving country. These measures would be applied and executed according to all laws of the receiving country. In the case of forced transfers, where the person in question never consented to reside in the receiving country, these cases were considered especially problematic.

These and other cases relating to equal treatment and the logical feasibility of applying Belgian rules in the Dutch penitentiary raised concerns. For example, according to the Belgian Prison Act, prisoners have the right to legal assistance during the hearings for disciplinary sanctions. However, prisoners testified that they found it embarrassing to ask their lawyers to travel all the way to the Netherlands to attend. Their right to legal aid was thus not officially refused, but the distance and related costs constituted a de facto denial of representation.

In the Norwegian agreement, Dutch criminal law and procedural legislation was applied exclusively if an inmate died or a criminal act was committed in Norgerhaven Prison. This elicited a strong critique from the Norwegian Civil Ombudsman because Norwegian authorities were unable to investigate or prosecute matters if inmates were to be subjected to torture or to inhuman or degrading treatment or punishment in the prison (Sivilombudsmannen, 2016, p. 22).

The uncertain legal structures surrounding the policy of transnational prisons ensured not only that equal treatment could not be guaranteed, but also gave rise to an environment that posed a tangible risk to prisoners’ human rights. Along with issues questioning the external legitimacy of these prisons, the examples highlight additional issues with regard to their internal legitimacy.

Our focus on the legal and penological issues raised by the trend toward transnational prisons in Europe makes a valuable contribution to the field of comparative penology. In our comparison, we have stressed three major findings. First, we explored the relationship between the growth of transnational prisons in the context of expansionist and reductionist approaches to prison overcrowding. We have shown that variants of expansionist policies were pursued. Secondly, the explicit policy of selecting deportable foreign national prisoners for transfer in the Danish case is less anomalous than both the rhetoric and aims of previous cases suggested. Thirdly, the above reflects the growing synergy not only between the goals and narratives of immigration policies, but also the way that immigration becomes a case of ‘policy creep’ in the organisation of criminal justice. The intertwining of criminal law and immigration law is what Stumpf (2006) has dubbed crimmigration. Prison legislation and the legislation on sentence implementation can be considered as a third field of law that interacts with criminal law and immigration law. Consequently, and in line with the work of Barker & Smith (2021), we have shown that the claims made in comparative penology regarding the exceptional nature of penal policies in Denmark should be questioned. The Gjilan case shows how penal policies aim to ‘protect’ the Danish welfare state and highlight the exclusionary practices of the welfare states that are often overlooked in comparative penology.

With regard to the former, we considered the use of transnational prisons in each case as reflective of an expansionist approach to prison overcrowding. Prison overcrowding is an increasingly widespread problem in several European countries, generating substandard, and in many cases inhumane conditions of incarceration, which affects both staff members and incarcerated individuals. Whilst the Council of Europe, as the continent’s leading human rights organisation, recommends reductionist prison policies (Snacken, 2006), we have shown that in these three cases expansionist approaches were dominant when the policy of renting capacity abroad was framed and implemented. Nevertheless, the experiences of Belgium and Norway strongly suggest that renting prison cells can only serve as a short-term solution within an expansionist framework. In the Belgian case, the use of Tilburg prison only served as a temporary measure to relieve immediate overcrowding. The strategy, however, has failed to remedy the underlying sources of the overcrowding problem and has proved an expensive option. Moreover, 2020 European prison statistics showed that Belgium still has the third highest prison density in Europe (Aebi & Tiago, 2021), with 117.2 inmates per 100 places. The rent of prison cells abroad looks likely to follow a similar pattern and raise equally critical questions about the Danish case.

Our second important finding concerns the selection criteria that are used to transfer people to a transnational prison. In each case, either indirectly (Tilburg and Norgerhaven prison) or directly (Gjilan), transnational prisons target (deportable) foreign national prisoners. In 2013, when Belgium was renting Tilburg prison, research found that:

While the transfer policy in Tilburg prison was not initially designed to send deportable foreign nationals abroad, in practice, the criteria for transfers disproportionately impacted (deportable) foreign nationals, which De Ridder convincingly linked to the crimmigration debate, and has documented as an example of the percolation of migration law enforcement into penal decision-making and sentence implementation in Belgium. Almost a decade later, the Danish policy, with its explicit intention to transfer irregular migrants to a transnational prison located in Kosovo, puts these transfers more than ever at the centre of the crimmigration debate. Moreover, the Danish proposal makes the Janus-faced nature of prison policies – previously claimed to characterise Nordic countries – more visible than in the Norgerhaven case. The Danish selection criteria can be seen as an explicit example of how Nordic countries are characterised by what Barker calls a Janus-faced society:

This dovetails neatly with our third finding, which strongly suggests a bifurcation in policy direction whereby humane punishments are reserved for nationals, and less desirable systems meet the response to the perceived challenge of foreign nationals (Ugelvik & Damsa, 2018). In the case of the latter, they are conspicuously treated as second-class prisoners who are undeserving of proper conditions of imprisonment. Consequently, the treatment of deportable foreign national prisoners shows the extent to which immigration policies and penal policies increasingly intertwine in Europe, resulting in their differential treatment. Border penologists who have highlighted the existence of a parallel penal system for foreign nationals that aims to achieve migration control instead of penal goals such as reintegration (Aas, 2014), such as Brouwer (2020), have called these specific prisons ‘crimmigration prisons’. While the term ‘crimmigration’ prison is mainly used to cover ‘prison-like environments’ that lock up deportable foreign nationals, a criminal offence – which is not linked to a violation of migration law committed by the people that will be transferred to Gjilan prison – appears to serve as a legitimation to create a particular questionable ‘crimmigration prison’. This specific prison is not only located in a remote area (which is often the case), but even outside the Danish borders and without the foreign national officers that are typically present in these places. This instigates questions and concerns relating to the conditions, legal protection and quality of life that will be present in this institution.

In the Tilburg and Norgerhaven cases, this move towards ‘crimmigration prisons’ was an indirect outcome, but in the case of Gjilan it is direct and explicit. Consequently, we believe that the current narrative by Danish politicians to reduce prison overcrowding echoes the policy trajectories in the antecedent cases above: they are just as much a reflection of wider European trends in immigration policy as they are a penological one. The agreement between Denmark and Kosovo thus exists at the intersection of processes of commodification of prisoners, crimmigration and the Scandinavian Janus-faced society. It results in a highly complex development that deserves critical attention from Danish and European scholars and control bodies such as the Ombudsman and CPT once the transfers begin. Moreover, empirical research on the imprisoned person’s experiences in this place will provide us with insights on how these macro-level policies impact daily life in prison.

The authors wish to thank Fraser King for careful reading and constructive comments and the reviewers for their valuable comments.