The obsolescence of detention: Versatility,expendability and plasticity in the field of immigration confinement

Introduction

Actively participating in punishment and society debates, border criminology scholars have persuasively challenged conventional views on penality, stressing that immigration enforcement practices such as migration policing, detention and deportation should be crucially considered in exploring the current penal landscape and examining the scale and intensity of penal coercion (Barker, 2018; Bosworth, 2019a; Bosworth et al., 2018; Franko, 2020; Pickering et al., 2015). The border criminology literature has also compellingly claimed that – at least in some jurisdictions – a dual criminal justice system is gradually consolidating, in the framework of which non-citizen clienteles are targeted by an emerging model of ‘abnormal justice’ (Franko Aas, 2014, Franko 2020), being excluded from legal safeguards and penal welfarism arrangements (Stanley, 2018; Turnbull and Hasselberg, 2017; Ugelvik and Damsa, 2018).

Immigration detention has played a pivotal part in this academic conversation. Only some decades ago, immigration detention was still an insignificant phenomenon across Europe (see Axster, this issue; Young, 2021). This scenario has dramatically changed since the 1990s, when immigration detention emerged or re-appeared in many global north and especially European jurisdictions (García Hernández, 2019; Leerkes and Broeders, 2010; Turnbull, 2017). The expansion of immigration detention drew the attention of the punishment and society scholarship, which begun to scrutinize its policies, practices, functions and impact in the penal landscape in different North Atlantic (Bosworth, 2014; García Hernández, 2014, 2019; Turnbull, 2017) and continental Europe jurisdictions (Broeders, 2010; Campesi, 2013; Esposito et al., 2022; Fernández-Bessa, 2021; Lindberg, 2022). Mary Bosworth's work is critical in this regard, since it provides a granular examination of detention practices in the UK, scrutinising a wide variety of aspects such as the peculiar relationship between staff and detainees, the pains of immigration confinement and the futures of detention practices (Bosworth, 2014, 2019b, 2022, 2023).

Despite the significant contributions of this literature, ¹ comparative, cross-national analyses on detention are still rare (see Canning, 2019; Lindberg, 2022), as they are longitudinal explorations on this changing penal sub-field (see Campesi, 2013; Fernández-Bessa, 2021). Consequently, this scholarship apparently relies on a typological notion of migration detention, which manages to represent neither the diversity of migration confinement across Europe nor the ways in which these practices are manifestly changing.

Policymakers, the public and the academic literature itself (see e.g., Hiemstra, 2019) frequently embrace a notion of immigration detention which is actually a metonymy of the varied forms of immigration containment and the varying roles played by detention practices within immigration enforcement systems. In fact, the perspective describing detention as the widespread form of confinement enforced in closed, relatively opaque facilities geared towards the expeditious deportation of non-citizens is actually a synecdoche. This detention model, which is prevalent in certain jurisdictions, for example, France, the UK and the US, is far from having a universal character. What's more, there are manifest signs that beyond being inhumane (Aiken and Silverman, 2021; García Hernández, 2017) this model is becoming obsolete, not least because it has failed to adequately carry out the various extended border control and post-entry social control assignments it has been tasked with (Kanstroom, 2007). ² In other, relatively underexplored jurisdictions, new and more malleable ways of performing those assignments are emerging (see also Romero, 2022).

This paper aims to explore the various ways in which we may describe the current detention scenario as showing signs of obsolescence. ³ For these purposes, the authors build on their previous analyses on the versatility, expendability and plasticity of detention practices (Ballesteros-Pena, 2022, Brandariz and Fernández-Bessa, 2021; Fernández-Bessa, 2021), keeping an eye on the blind spots of border criminology debates and the underexplored regions, in a sort of southern (border) criminology analysis. ⁴ In developing this analysis, the paper proceeds as follows. The ways in which immigration detention, its missions and its versatility have been theorised by the academic literature are examined in the next section. After an intermezzo exploring whether the coronavirus pandemic operated as an enabler of change in this field, the third section scrutinises the notion of expendability, that is, the decline of detention-centric models of immigration enforcement and the gradual consolidation of deportation systems that rely less and less on confinement practices. The fourth section, in turn, explores the unprecedented plasticity given to immigration detention by the proliferation of new forms of confinement and containment in extended border control areas. Finally, the concluding section recapitulates the lessons to be drawn for punishment and society studies from this analysis on the obsolescence of detention.

Versatility: Which question is detention the answer to?

From a legal viewpoint, immigration detention is an immigration law instrument that aims to serve the purposes of the border control system, and specifically of the deportation regime (Franko Aas, 2014; García Hernández, 2014). In fact, immigration detention is legally considered as an administrative law precautionary measure to be imposed for the preparation of the eventual enforcement of a return order (Campesi, 2013; see also Directive 2008/115/EC of the European Parliament and of the Council of 16 December 2008; Articles 15–18; Return Directive - hereinafter RD).

However, this legal definition does not exhaust the socio-legal and criminological debate on the nature and missions of immigration detention (Campesi and Fabini, 2020; Falsone, 2021). As a matter of fact, a number of influential theses have been elaborated to grasp the aims, both instrumental and symbolic, actually pursued by this immigration enforcement device.

In this regard, the crimmigration thesis (García Hernández, 2018; Stumpf, 2006) argues that the gradual convergence of immigration law and criminal law has enabled the utilisation of immigration law devices such as immigration detention and deportation to deal with criminal offences perpetrated by non-citizens. Crimmigration changes have resulted in the adoption of punitive, criminal law-like instruments and procedures in the field of immigration control (Eagly, 2010; Legomsky, 2007; Sklansky, 2012). This thesis, therefore, concludes that immigration detention is being increasingly used for crime prevention purposes (García Hernández, 2014; Turnbull, 2017).

In contrast to this influential theory, another strand of the literature has scrutinised the purpose of immigration detention by giving prevalence to its role as an institution ancillary to the deportation machine. In this framework, various scholars have highlighted the deterrence aspects of detention policies. Some authors embrace what might be called a ‘general deterrence’ thesis, according to which detention practices aim to dissuade unwanted non-citizens from coming to, and staying in Europe (Bosworth, 2019a; Campesi, 2015; Hiemstra, 2019; Leerkes and Broeders, 2010). By contrast, another body of literature adopts a sort of ‘specific deterrence’ lens, which claims that detention practices aim to forcefully persuade non-citizens ‘to leave’, encouraging them to either collaborate in preparing their forced removal or sign in for so-called ‘voluntary return’ programmes (Hasselberg, 2016; Leerkes and Kox, 2017; Lindberg, 2022; Martínez et al., 2018). In this regard, US scholars have singled out ‘self-deportation’ as the ultimate goal of immigration enforcement policies (Jain, 2019; Provine and Zatz, 2015). Needless to say, these specific deterrence goals require detention conditions to be particularly harsh.

Despite the significant academic effort invested in exploring detention practices from a dissuasion perspective, there is a consistent consensus among scholars working in various national settings that bordered penality devices fail to serve deterrence purposes (García Hernández, 2014; Hasselberg, 2014; Hiemstra, 2019; Leerkes and Kox, 2017).

These deterrence theses seem to assume that immigration detention policies are based on significant institutional capacities guided by a well-planned law enforcement rationale. However, only certain national detention systems show signs of long-lasting consolidation and institutional stability. By contrast, in many jurisdictions, these state control apparatuses are either too tiny or too unstable to play a significant role in developing extensive removal policies, let alone crucially contributing to enable immigration management policies. This critical shortcoming gives credit to these stressing the modest role played by immigration detention practices. Considering the ‘undeportability’ deficit affecting various immigration enforcement systems (Campesi, 2015; Fabini, 2021, 2022; Leerkes and Broeders, 2010), these analyses call into question the allegedly pivotal contribution of detention policies to the deportation apparatus. The Italian case is critical in this regard. In Italy, the detention apparatus has been serving the relatively humble purpose of managing street-level public order, facilitating policing efforts to govern allegedly troublesome non-citizen groups (Campesi and Fabini, 2020; Fabini, 2022).

The teleological failures mentioned so far have led scholars to explore additional immigration detention goals, such as policing membership, strengthening national sovereignty and managing destitute groups (Bosworth, 2019a; Leerkes and Broeders, 2010).

In sum, these diverse academic perspectives suggest that detention systems are far from having a clear and precise mission, but are serving various, site-specific goals and following divergent paths. This means that immigration detention has been – or, has been said to be – the answer to many different questions. In fact, versatility is one of the most distinctive features of immigration detention, which has allowed it to adapt to changing migration flows, new political and social priorities and variations of state capacities.

Intermezzo: The pandemic as an accelerating force of the obsolescence of detention?

In contrast to conventional descriptions of immigration detention, many national detention systems have recurrently proven to be unstable and relatively fragile. One critical occasion demonstrating this lack of robustness was the coronavirus pandemic. State coercion systems did not elude the impact of such a public health event (see also Esposito et al., 2022).

Specifically, the cross-border nature of return measures (Franko, 2020) caused the implications of the coronavirus restrictions to reverberate across the entire European deportation system. Detention systems showed clear signs of frailty as well. Although some countries were reluctant to release non-citizen detainees, many global north countries implemented ambitious release programmes (Brandariz and Fernández-Bessa, 2021), resulting in the number of detainees significantly declining in 2020 in jurisdictions such as Belgium, France, Germany, Greece, the Netherlands, Spain, Sweden and the UK. ⁵ Spain, in turn, closed down all detention facilities over a four-month period in summer 2020 (Ballesteros-Pena, 2022). This decline, though, seems to have been a short-lived intermezzo, especially in countries such as France, Greece, Italy, the Netherlands and the UK, where the annual number of detainees in 2022 was as high as in 2019. However, the detention system was still far from operating at pre-pandemic capacity in 2022 in countries such as Austria, Belgium, Spain and Sweden. ⁶

The pandemic laid bare the relative vulnerability of the immigration detention system. In an unexpected win for the abolitionist playbook, many global north detention systems have been underperforming for a significant period of time, unambiguously showing that migration management policies do not collapse when a wide-ranging detention system is unavailable (see also Bosworth 2022). The implications of this intermezzo for detention practices in Europe still need to be fully explored. However, it should be acknowledged that the detention changes witnessed during the pandemic were not completely unexpected. The innovation and plasticity of containment practices, and even the relative expendability of the detention system, are the outcome of processes that had been ongoing for a while. In this respect, the pandemic may be described as an accelerating force. The signs of obsolescence of the detention system that had already surfaced before this intermezzo are going to be explored in the next two sections.

Expendability: Towards a post-detention-centric deportation system?

The grey literature, and to a certain extent the academic literature on migration control and bordered penality have generally embraced a detention-centric perspective, that is, a viewpoint in which deportation policies crucially rely on, or are even dependent on the consolidation of wide-ranging immigration detention systems.

Certainly, there are reasons to espouse this viewpoint. Large detention systems have proven to be vital to develop expansive detention policies in various jurisdictions. The US immigration enforcement system is a good case in point. Its impressive deportation capacity is supported by an extensive network of detention facilities, which managed to confine more than 300,000 non-citizens per year from 2007 to the onset of the coronavirus pandemic (source: DHS; see García Hernández, 2019). The annual number of detainees in the US immigration enforcement system surpassed the symbolic threshold of half a million for the first time in 2019 (Young, 2021). It is less known that Mexico does not lag far behind (Campos-Delgado, 2021). More than 100,000 non-citizens were detained per year in the Mexican so-called migration stations from 2014 to 2016 and again in 2018 and 2019. In the framework of the severe migration crisis that has been affecting Central American and North American countries in recent years, the number of detainees skyrocketed to more than 307,000 in 2021 and more than 444,000 in 2022 (source: Mexican Department of Migration Policies). ⁷

Detention systems in Europe are far from having such a striking confinement capacity. Still, some national systems are notably extensive. The UK is an obvious reference in this regards (Bosworth, 2014). Although the British detention apparatus has been slightly shrinking in the recent past, on average 28,063 non-citizens annually entered detention in the UK from 2010 to 2019 (source: UK Home Office). ⁸ In this respect, the UK system is second only to that of France (Fischer, 2017). On average, the French detention system confined 48,833 non-citizens per year from 2010 to 2019 (source: La Cimade). ⁹ Specifically, the French detention apparatus has a marked (post-)colonial nature, since 48.1% of the non-citizens detained from 2012 to 2021 were in custody in detention facilities located in the French colonies overseas, most notably in the African island of Maoré/Mayotte. Greece is an additional jurisdiction to be added to this list of capacious detention systems (Iliadou, 2023), because the annual number of detainees peaked at over 25,000 from 2017 to 2019, rising above 33,000 in 2022 (source: Global Detention Project; hereinafter GDP; see also Majcher et al., 2020). ¹⁰

The immigration enforcement model developed in these jurisdictions is consistent with European Commission (hereinafter, EC) views, which have recurrently stressed the pivotal part played by detention policies in consolidating an effective return system in Europe. This perspective is epitomised by the 2015 EU Action Plan on Return (COM(2015) 453 final), which emphasised that ‘to meet their obligation to enforce return, Member States should use detention.’ This detention-centric model of immigration enforcement was unambiguously upheld by the EC Recommendation (EU) 2017/432 of 7 March 2017 on making returns more effective, which claimed that ‘detention can be an essential element for enhancing the effectiveness of the Union's return system.’ This Recommendation and the 2018 Proposal to amend the RD (COM(2018) 634 final) specified the paths to be followed to amplify the scope of EU detention systems. These include the extension of the maximum initial period of detention in certain countries, the authorisation of the detention of minors and families with minors – at least, in certain exceptional cases –, and more generally the need to ‘bring detention capacity in line with actual needs.’

In light of this sympathetic approach to immigration detention, there are apparently good reasons to challenge the very notion of obsolescence of detention. However, the detention-centric model epitomised by these EC documents and the aforementioned national cases is actually a synecdoche that falls very short of representing the wide diversity of immigration enforcement practices witnessed in European countries and elsewhere. GDP data show that many EU jurisdictions – at least, Croatia, the Czech Republic, Estonia, Finland, Ireland, Latvia, Lithuania, Luxembourg, Poland, Portugal, Romania and Slovakia – have relatively tiny immigration detention systems that confine some hundreds or a bit more than one thousand non-citizens per year (see also Majcher et al., 2020). ¹¹ This detention landscape might be seen as the confirmation of EC diagnoses, because many of these countries do not rank very high in deportation practices, either; actually, some of them feature a broad deportation gap (see Eurostat data). Still, some of these jurisdictions excel in deportation terms. Both Poland (Klaus, 2020) and – to a lesser extent – Finland (Könönen, 2022) show that national immigration enforcement systems may develop an extensive deportation system and have a narrow deportation gap without critically relying on widespread detention practices.

Interestingly, in some of the aforementioned countries such as Ireland, detention practices have been in a steady decline. The detention apparatus has also been shrinking in other European countries that are active on the deportation side. The Netherlands (Staring and Timmerman, 2021) and Spain (Fernández-Bessa, 2021) are two critical cases in which the detention system has been notably downsized since the late 2000s. Still, Germany is the most important case to be added to this list. The German detention system provides a suitable template of how a post-detention-centric model of immigration enforcement might look like. In stark contrast to France and the UK, Germany makes a relatively parsimonious use of detention measures. ¹² What is more, the German system has significantly reduced the number of detainees at least since the 2000s, even in a period characterised by rising numbers of irregular newcomers and enforced deportations (source: Eurostat). This inspiring development is the outcome of the convergence of various factors. The Bero and Bouzalmate judgement of the Court of Justice of the European Union (C-473/13) of 17 July 2014 ¹³ compelled the German authorities to put an end to their long-standing practice of confining deportable non-citizens in prison facilities, leading to a sort of logistical crisis in the German detention system (Kreuzkamp, 2022; Majcher et al., 2020). In addition, a set of prominent legal reforms – such as the First (2017) and Second (2019) Acts to Improve the Enforcement of the Obligation to Leave the Country – and policy developments have given top priority to so-called voluntary returns (EMN, 2018; Haberstroh, 2021), resulting in new arrangements in which detention might play a relatively secondary role. In this framework, the German deportation apparatus has been increasingly relying on reception facilities since the mid-2010s (Kreuzkamp, 2022). ¹⁴

Although the long-term consequences of these changes in the German case still remain to be seen and recent developments show mixed signals (ECRE, 2022; Graebsch, 2019, 2022), this national case might prove to be particularly influential in the EU scenario. In contrast to the official rhetoric voiced in the mid-2010s, the EC discourse on detention policies appears to be more cautious in the last years, apparently championing a prudent approach to custodial measures. In the framework of the increasing priority given to voluntary return practices (see the EU strategy on voluntary return and reintegration of 27 April 2021; COM(2021) 120 final), recent pivotal policy plans – including the New Pact on Migration and Asylum of 23 September 2020 (COM(2020) 609 final) – are putting the emphasis on the development of alternatives to detention (see also Slingenberg, 2022). In fact, the main goal in cases of forced return procedures seems to be ‘finding a good balance between increasing the availability and use of effective alternatives to detention, and ensuring that specialised facilities are available and well-equipped when detention is used’ (see the EC Communication on enhancing cooperation on return and readmission of 10 February 2021 (COM(2021) 56 final).

Potential institutional changes aside, the gradual emergence of a post-detention-centric model of immigration enforcement is also the outcome of widespread practices of political contention that have contributed to transform borders ‘from below’ (Segrave and Wonders, 2019). In fact, advocacy groups and non-citizen communities have been constantly challenging immigration detention in many European countries at least since the turn of the century, developing various forms of border activism that have kept the political costs of detention practices high (Fernández-Bessa, 2019). These acts of resistance (Lindberg, 2022) should be credited for having undermined the political legitimacy of the detention apparatus.

Additional blueprints for a post-detention-centric model – actually, for a detention-free model – of immigration enforcement may be found outside Europe. In contrast to the frequently assumed global nature of detention (Macías Rojas and Tazzioli, 2022), immigration detention does not exist either in Chile or elsewhere throughout South America (Ceriani Cernadas, 2017). Paradoxically, though, Chile has been championing a hard-line approach to immigration enforcement since the early 2010s, in the framework of which the number of enforced removals was largely on the rise until the coronavirus outbreak (SJM, 2022). However, Chilean authorities have developed such a system without resorting to detention measures, giving vital preference to the deportation of sentenced non-citizens, who are incarcerated until the day of their repatriation (Brandariz et al., 2018). Indonesia, in turn, has witnessed an intriguing decarceration turn since 2018, when detention practices were abolished. This shift gave rise to a new system of containment based on alternatives to detention, which despite being less inhumane has proven to have significant shortcomings (Missbach, 2017, 2021).

Plasticity: The proliferation of new forms of containment across Europe's southern border

Beyond these cases in which detention has an increasingly secondary role within the immigration enforcement system, there is an additional emerging phenomenon that is signalling the obsolescence of detention. Southern European countries such as Italy, Greece and Spain had long been countries of emigration and it was not until the last decades of the twentieth century that they started to receive migrants (Cachón 2009). As destination countries, they have long been characterised by unstable mobility management policies and poor regulations (López Sala, 2007), certain tolerance to irregular migration and the pivotal role played by the sizeable informal economy sector in providing labour insertion for newcomers (Cachón, 2009; Calavita, 2005), for example, in agriculture.

This scenario, in which labour migration and allegedly lax mobility policies took centre stage, changed with the 2010s surge of asylum. Initially in Italy and later in Greece and Spain, the three Mediterranean countries have been receiving an increasing number of asylum applications. ¹⁵ Furthermore, the three countries have ranked among the top five EU countries in number of accepted applications in recent years, along with Germany and France. ¹⁶ As a result, new practices for managing human mobility have surfaced across Southern Europe. These practices are characterised by the emergence of humanitarian logics and its convergence in varied combinations with punitive and coercive measures, along with their short-lived character, their changing nature and functions and the lack of a formal legal framework. Long-standing forms of detention in closed facilities have given way to a more complex landscape of practices and sites of confinement, characterised by their plasticity. Plasticity (see also Ballesteros-Pena, 2022) refers to the fact that systems of confinement have moved from a stable condition, including a physical infrastructure, legal frameworks and guidelines of functioning to mechanisms of containment characterized by flexibility and fluidity. Plasticity combines two main features: adaptability and resistance. Therefore, containment practices can quickly respond to changing circumstances, such as a temporary increase of arrivals, by creating improvised facilities, frequently having a makeshift nature. As quick as they appear they can also be dismantled. Their logics, functions and the actors involved can also vary depending on the specific spatiotemporal context. At the same time, the practice of confinement, while evolving towards more malleable and blurred mechanisms, is still enduring despite its unstable character.

In Southern Europe, hotspots are the most widespread example of this complex set of confinement and containment settings. The ‘hotspot approach’, launched by the 2015 European Agenda on Migration (COM(2015) 240 final), has the declared aim of supporting EU external border countries facing significant migratory pressure in the initial reception, identification, registration and fingerprinting of asylum seekers and migrants. For this purpose, national authorities work in coordination with EU agencies deployed on the ground. This approach has been declaredly implemented in Greece and Italy, although other EU member states can potentially adopt it. Spain has not formally embraced the hotspot approach; however, from 2017 it has developed its own local version of the hotspot apparatus: the so-called Centres for the Short-term Attention of Foreigners (CATE, for their initials in Spanish).

These new facilities are frequently characterised by a lack of regulatory framework (Barbero, 2021; Campesi, 2018a; Papoutsi et al., 2019). Furthermore, this policy approach lacks a clear definition of what a hotspot ‘actually is and does’ (Pallister-Wilkins, 2020, 992) beyond its vague description as a strategy aimed to reinforce the capacity of member states to deal with external border crises (2015 European Agenda on Migration). In Spain, there is no formal characterisation of CATEs beyond the idea that migrants are detained for the first 72 h upon arrival by sea. Along with them, an increasingly complex typology of migration camps and humanitarian centres surface and resurface in the main areas of arrival, particularly the Canary Islands and the Andalusian coast in southern Spain (Ballesteros-Pena, 2022). In these facilities, we can find undocumented migrants that have received a removal order along with asylum seekers waiting to be transferred to other facilities. This unstable and flexible nature exemplifies the adaptability component of the plasticity condition.

Although hotspots and hotspot-like facilities are understood as allegedly provisional responses to mobility crises (Barbero, 2021; Boza and Pérez, 2019; Campesi, 2018b), the model has quickly generalised (Campesi, 2018b; Papoutsi et al., 2019) with constant adjustments and modifications. In the case of Spain, humanitarian centres run by NGOs in the Canary Islands have changed their targeted populations several times in the last few years. For instance, humanitarian facilities initially deemed to host male newcomers ended up receiving women and children due to the changing demographics of sea arrivals. Additionally, in the smallest Canary Islands, CATEs do not have a stable physical infrastructure, being recurrently established and dismantled in harbours and industrial buildings with substandard conditions. Thus, these new mechanisms of confinement and containment combine their structural condition, since they are increasingly consolidated as an enduring part of the system - thus, resistant - and their adaptable character.

Humanitarian and punitive logics are present in these diverse sites, regardless of their declared nature. Scholars and oversight bodies have documented coercive practices taking place in hotspots and CATEs, such as extended detention periods and the use of force for the collection of biometric data (Campesi, 2018a; Defensor del Pueblo, 2021). Still, the humanitarian logics of care coexist in these facilities with the logics of control of unwanted populations. In fact, the reception nature resulting from their legal role as humanitarian response to increasing arrivals is further strengthened by the prominent part played by NGOs and international organisations in running the hotspot system.

As a matter of fact, the diversity of facilities has resulted in the proliferation of new actors playing different roles: together with NGOs, several international organisations such as a UNHCR are active in Greece, whilst European agencies such as Frontex play a part in the management of hotspots and Spanish CATEs. The implications of this multiplicity of actors are diverse. They can end up creating a ‘surrogate state’ like in Greece (Moschopoulos, 2023), superseding state migration management capacity – temporarily, until they leave the country. Additionally, in Greece but also in Spain, NGOs end up having a conflicting relationship with other actors in the field, such as social movements and the migrants and asylum seekers they are supporting. Frequently, these non-state actors must enforce governmental decisions that run counter to their values and reduce their autonomy of action. Additionally, managers and front-line workers can adopt contradictory approaches in dealing with the situation on the ground. In sum, contrasting perceptions and strategies (Panebianco, 2019) from different state and non-state actors interact and overlap within these facilities.

In this mutable landscape of containment sites, screening processes are also malleable. Hotspots in Italy and Greece perform a process of social sorting separating the ‘deserving’ asylum seekers from the ‘underserving’ newcomers that are channelled into removal procedures and socially and economically segregated (Campesi, 2018a; Heyer, 2022; Papoutsi et al., 2019). This process of categorisation involves a high degree of discretion, which is further increased by the structural ambiguities of the system itself (Heyer, 2022). Similar deficiencies can be found in Spain, where classification processes also take place in CATEs to separate the national groups with high odds of being deported (Moroccans and Algerians), that are redirected to detention facilities, from potentially undeportable groups (Fernández-Bessa, 2021), that are transferred to camps and other humanitarian facilities.

The academic literature has interpreted the emerging archipelago of sites of migration management and enforcement in different ways. Campesi (2018b) identifies different logics at play in the evolution of the reception system in Italy. In his approach, hotspots follow a logic of containment because they detain people in border zones, under an uncertain legal status and in substandard conditions. Challenging the analysis of detention as confinement, Tazzioli (2020) refers to ‘governing migrant's mobility through mobility’ as a technology of governance aimed at directly or indirectly keeping migrants on the move, disrupting their autonomous trajectories and containing them beyond detention (see also Tazzioli, 2018; Tazzioli and Garelli, 2020). Thus, more than just keeping asylum seekers in detention, this technology hinders and disrupts their mobility through different strategies.

This distinction between confinement and containment appears to have significant implications, which deserve to be further scrutinised (Cornelisse, 2022; Macías Rojas and Tazzioli, 2022). These studies, though, overlook that the gradual consolidation of the hotspot archipelago is in itself a revealing signal of the obsolescence of detention. In fact, rather than being smoothly extended (Mattiello et al., 2022), the detention-centric model is being superseded by a plastic apparatus that more effectively reacts to the pressing needs of European mobility policies. Hotspots do not straightforwardly serve any of the goals of the detention system examined in a previous section. In stark contrast to the outspoken detention-centric stance adopted by the EC Recommendation (EU) 2017/432, the hotspot apparatus devaluates the relevance of deportation purposes in countries featuring broad deportation gaps, leading containment, screening and registration goals to take centre stage. This shift cannot be disconnected from a European immigration enforcement scenario in which so-called secondary movements are a critical concern (Gülzau, 2021; Karamanidou and Kasparek, 2022) and intra-EU deportations, namely the return of Dubliners, are garnering increasing traction (Brandariz et al., 2023).

Conclusions: Promises and pitfalls of a changing detention scenario

The versatility of detention practices has surely contributed to the resilience of detention systems, which may pursue diverse goals and meet different sets of expectations. However, detention policies are not set in stone. In fact, the coronavirus pandemic unveiled the relative vulnerability of detention arrangements. This paper shows that significant changes are unfolding in this field.

The successive mobility crises affecting the Mediterranean area lay bare that detention practices have failed to fulfil their extended border control tasks (Kanstroom, 2007), aimed at shielding EU's southern external borders. Against the backdrop of the increasing relevance of asylum phenomena across southern Europe, the detention field has shown its plasticity in this sphere, giving shape to innovative forms of containment, in which detention is intertwined with reception and border control goals meet humanitarian aspirations. Epitomised by the hotspot approach, in this framework screening and sorting purposes take centre stage, rather than deportation goals.

Still, current detention changes are also being fuelled by processes taking place away from external borders. Many European jurisdictions have long used detention measures sparingly. In addition, in a deportation scenario in which so-called voluntary returns are taking centre stage, other countries such as Germany and the Netherlands show a decreasing reliance on detention. In addition, two global south countries as diverse as Chile and Indonesia epitomise two different detention-free models of immigration enforcement, respectively, relying on the prison system and alternatives to detention.

The implications of the exploration of detention changes carried out in this paper for punishment and society conversations on bordered – and ordered (Franko Aas, 2014) – penality are twofold. Initially, it has implications of conceptual nature. The detention cases explored here show the marked diversity of the current detention and immigration enforcement landscape. In addition, this paper challenges the viewpoint according to which the emergence of the archipelago of hotspots and short-term facilities gives rise to a continuum of confinement (Aradau and Tazzioli, 2021; Campesi, 2018b), to a sort of net-widening shift that essentially extends the scope of the detention system (Esposito et al., 2022). Certainly, this perspective has significant value for advocacy purposes, since it puts the spotlight on the gross human rights violations characterising current immigration enforcement policies. However, these advocacy benefits come with conceptual costs. This viewpoint obfuscates the analysis of continuities and discontinuities, of what is new and what is obsolete in the field of immigration enforcement. It blurs the boundaries between detention and additional forms of containment (Macías Rojas and Tazzioli, 2022; see also Cornelisse, 2022; Tazzioli, 2023), thereby overlooking the different missions, actors and practices separating them, and the shifting migration control rationales operating on the ground.

Additionally, this exploration has implications of normative nature that may have an impact on reformist and abolitionist agendas. Punishment and society scholars have decried ‘catastrophic’ perspectives on penality, which risk to overemphasise rising punitiveness and penal excess phenomena (Hamilton, 2014; O’Malley, 2000). In a similar vein, this paper aims to draw attention to the increasing expendability of detention practices and the promising aspects of the current detention scenario. However, the emerging detention landscape has a Janus-faced character. It certainly supplies cognitive and policy tools for detention abolition, which may leverage the national cases in which detention either does not exist or is increasingly dispensable. However, it also has a threatening dimension, represented by the proliferation of public security-based regulations, so-called voluntary returns and harmful alternatives to detention. EC legal proposals such as the 2018 RD proposal show the increasing relevance given to ‘public order’ and ‘national security’ considerations in grounding detention measures (see also EC's 2017 Action Plan on Return; Graebsch, 2019). The Canadian system unambiguously shows the bleak detention scenario to which these regulations may lead (Ballesteros-Pena, 2021; Wallace, this issue). It is well-known as well the many human rights risks posed by so-called voluntary return programmes, which operate as a form of soft deportation (Kalir, 2017). In addition, recent policy changes have not significantly eroded the detainability of non-citizens (De Genova, 2019, 2020), which is enabled by unreformed legal frameworks that still statute extensive detention powers. Moreover, compelling ethnographic studies have shown the harmful nature of various alternatives to detention, especially mobility restrictions and electronic monitoring – which nonetheless do not seem to be widely used in Europe (EMN, 2021) - that seriously affect the psychological and even physical well-being of newcomers (Sanchez-Boe, 2020; Sanchez-Boe and Mainsah, 2021).

In short, the current migration enforcement landscape is not destined to put an end to detention practices. However, it is characterised by the increasing momentum gained by relatively parsimonious forms of immigration enforcement such as so-called voluntary returns and alternatives to detention, as well as by the emergence of post-detention-centric models of enforcement that increasingly rely on reception facilities and prison facilities to prepare deportations. In developing its research agenda on immigration enforcement, punishment and society studies should consider this shift. The current scenario opens new paths to develop an anti-carceral state agenda, which nonetheless should avoid digital confinement and other punitive arrangements prepared to eventually fill the detention vacuum (Missbach, 2021).