Situated sites of migration control: Swiss deportation practices and their relational materiality in prisons, hospitals, and airports

Deportation is often studied in the context of research into the administration, enforcement, and control of immigration, with researchers highlighting the violent effect on deportees and studying the various actors involved in the deportation process. This contribution adds to the growing literature on deportation infrastructures by emphasizing the inseparability of deportation procedures from the specific sites in which they unfold, as well as highlighting the analytical interest and political agency of such spaces. My socio-material approach applies a rather classical understanding of infrastructure, asking what three specific deportation sites – prisons, hospitals, and airports – can tell us about deportation procedures as a technology of immigration enforcement. Using Switzerland as a case study, this paper analyses deportation procedures, including the role of human and non-human actors, paying particular attention to the situatedness and relationality of deportation infrastructures. The socio-material analysis of the architecture of the three sites under discussion ultimately exposes deportation as violent statecraft.

1 Introduction

Noncitizens¹ are frequently deported from prisons or from specialized detention facilities that house individuals with precarious immigration status, including refused asylum seekers or people awaiting deportation at the end of a prison sentence. Noncitizens are also deported from temporary housing facilities, including former hotels, asylum centres, and medical facilities, as well as from private homes (Jerrems et al., 2023; Saadi et al., 2017). While some of these sites are physically designed to facilitate immigration enforcement (e.g. through surveillance or physical confinement), others are repurposed by immigration authorities. As this paper argues, the physical design of such sites can have an impact on deportation proceedings, affecting the different actors involved and sometimes complexifying the work of state agents. Between its appropriation by immigration authorities and an agency that facilitates the resistance of diverse actors through various material and spatial configurations, immigration-control infrastructure is embedded in a complex and dynamic socio-material field. This paper explores the significance of deportation sites for our understanding of immigration-control procedures, examining them as situated spaces of control and resistance and as relational technologies of immigration enforcement.

The role of space and infrastructure has been discussed across a broad swathe of migration-studies scholarship, with particular attention being paid to legal and carceral geographies and their articulation of the intricate relation between law and space (Könönen, 2020; Pickering, 2014; Hiemstra, 2013); research has highlighted, for instance, how detention facilities are often located in peripheral spaces where people can more easily “vanish” (Chak, 2017). By a similar token, the study of infrastructure, including both material spaces and digital technologies, has gained momentum within border and deportation studies (Dijstelbloem, 2021; Walters, 2017). Scholars have for example shown how infrastructure can facilitate noncitizens' resistance to border controls (Amelung et al., 2020). I build on this existing scholarship to study spaces of immigration enforcement as sites of negotiation, resistance, violence, and state power (Walters, 2016); my aim is to contribute to an expanded understanding of such sites through the analysis of infrastructure. In doing so I engage with concepts of socio-materiality (Hultin, 2019), actor–network theory (Latour, 2005), assemblage theory (Müller, 2015), and feminist materialism (Bennett, 2010), all of which draw attention to the intricate nature of human and non-human interactions.

From this point of departure, the analysis that follows focuses on ad hoc sites of deportation (such as detention and prison facilities) that make use of the state's existing coercive apparatus, as well as on less conventional spaces where removals take place, such as hospitals or asylum accommodation. While “classic” deportation sites tend to already be linked to enforcement, thus offering an infrastructure that is familiar to the state agents charged with implementing deportations, “newer” sites of deportation, even those that have a long history with removals (Saadi et al., 2017), were not designed with immigration enforcement or the penal system in mind; their architecture is thus at least partly different. The state's capacity to deport individuals from hospitals, psychiatric facilities, and asylum accommodation underlines the extent to which such practices have been normalized and testifies to the expansion of immigration enforcement. Yet such sites also pose challenges to the deportation regime and open up spaces from which it can be contested. They become (new) nodes through which states perform their power; the possibilities of resistance they offer derive from their not yet being (or being unable to be) entirely adapted to their function in deportations. An analysis of (local) deportation infrastructure thus discloses potential ruptures in state power, as well as contestations over the use of space and the challenges that such space can be pose to all actors involved. Looking at the variety of sites can help to understand how such spaces are shaped by removal practices and vice versa, as well as how individuals working within these infrastructures either adapt or are already attuned to the work of migration enforcement. By looking closely at the situatedness of deportations, such an analysis can also explore the relationships between such spaces, disclosing a relationality enacted not only at the local level, but also transnationally.

My analysis is based on ethnographic fieldwork conducted in Switzerland between 2015–2017. I present observations and interviews conducted with police officers and immigration officials in charge of planning and implementing the removal of noncitizens, as well as with staff at the three types of deportation site under study. The next section introduces relevant literature at the intersection of infrastructure, socio-materiality, and immigration enforcement; this is followed by a brief contextualization of the Swiss migration regime and a note on my method of data collection. My analysis then discusses the three sites before the paper concludes with a final discussion.

2 Migration and deportation infrastructure

New techniques of migration governance are evolving as dispersed assertions of sovereign power merge with increasingly biopolitical modes of controlling individual mobility. (Mountz and Hiemstra, 2014:385)

Deportation, one of the main fields of migration governance, is undertaken with the aim of enforcing such controls and putting such assertions of power into practice through the removal of unwanted noncitizens. As a process it is characterized by a mixture of planning and unpredictability for both state actors and noncitizens, given that deportation procedures are entangled in (trans)national and local negotiations involving diverse sets of actors and are happening in various places.

Deportation is therefore increasingly studied from a socio-material point of view, with researchers showing how removal procedures are embedded in a dynamic, political–technical assemblage (Blue, 2015). The past decades have witnessed a growth in scholarly interest in infrastructure and socio-materiality within migration studies (e.g. Ho and Hatfield, 2011). This has included scholarship on infrastructural Europeanism, an approach that analyses how material structures support administrative agency, facilitating continuity in the face of policy gaps (Pelizza and Loschi, 2023). Socio-material studies have also explored the role of immigration enforcement technologies, their dehumanizing effects on detainees (Radziwinowiczówna, 2017), and the increasing digitalization of such spaces of control (Leese et al., 2021). Prior research into the relational materiality of immigration enforcement technologies and infrastructures more broadly has highlighted the extent to which they are co-constitutive of migrants' everyday realities and practices (Oginni, 2022). Infrastructures can constrain agency; but they also open up space for contestation and participation, shaping the agency of migrants in their struggle for recognition (Amelung et al., 2020).

Such socio-material approaches in migration studies have their origins in classical works on assemblage (Deleuze and Guattari, 1987) and socio-material research that highlights the connectivity of human interactions and objects, spaces, or other kinds of infrastructure (Müller, 2015). They highlight socio-material interdependencies (Hultin, 2019) and the quasi-agency of “things” (Latour, 2005; Bennett, 2010), allowing for the exploration of complex sites of oppression and resistance. As such, a socio-material lens allows us to “trace how disparate elements – human and non-human – emerge in webs of activity, and become linked into assemblages that collectively exert power and generate knowledge” (Fenwick et al., 2012:6).

Approaching deportation as a specific field in migration studies from a socio-material angle therefore offers an understanding of how space and agency emerge as the result of human and non-human relations entangled in differentiated power dynamics that, for example, facilitate or hinder certain practices (e.g. the carrying out of a deportation or resistance to one). Previous scholarship in this area has explored how sites of removal have multiplied – especially during the Covid-19 pandemic – to include hotels (Saadi et al., 2017), “army barracks, hospital wards, makeshift evacuation tents or pods, suburban apartments, and more” (Jerrems et al., 2023:1). As such, the deportation industry increasingly engages private providers in both organizing (e.g. planning post-removal support in a person's “home country”) and executing deportations (Sullivan and Zayas, 2013). A socio-material lens has been used to critique the “weaponization” of various infrastructures, materials, and technologies that have been (re)appropriated by state actors and turned into technologies of border violence (Keady-Tabbal and Mann, 2023). Deportation practices increasingly involve two distinct dimensions: one set of infrastructure is involved in forced removals, while another works to seduce, harry, or convince unwanted noncitizens into returning; deportations are thus enacted through both hard and soft power (Leerkes et al., 2017). Yet a focus on infrastructure can also help uncover how resistance towards restrictive immigration policies can take material form, e.g. the writing of e-mails and customer complaints to businesses involved in deportation or the physical picketing of their headquarters (Hintjens et al., 2011). William Walters (2017, 2018) has extensively researched both the macro- and the microphysics of deportation, highlighting in particular the role of aviation infrastructure and the various actors that seek to implement or contest deportation via the air. Material – often mundane – things and technologies can, when deployed by migrants, very easily emerge as security threats, highlighting the tension between different actors (see Campesi, 2015, on how the stuffing in pillows can be used in suicide attempts). Attending to such structures or materials, this line of scholarship underscores the relational aspect of materiality (Barad, 2007) – how it can influence and shape interactions (Sørensen, 2007) and disclose power inequalities, for example, when noncitizens face uncertain waiting periods and apparently random transfers between detention facilities prior to deportation (Hiemstra, 2013). In this regard deportation practices, much like detention facilities, can become chaotic sites (Hiemstra, 2013) while – as I argue below – at the same time being subject to a high degree of structure and organization and embedded in regulations, laws, and organizational practices. For instance, deportation infrastructure frequently facilitates standardized techniques for the identification and categorization of migrants (Tazzioli, 2017); this underlines how matter and things become social entities with an active role in the generation, stabilization, and reproduction of social order and sociality (Aradau, 2010). This also holds true on a transnational level, where “struggles for the control of space extend beyond and cut across national boundaries” (Soto Bermant, 2017:137), creating division and generating “new sets of relationships between the socio-spatial units they bring into existence” (Soto Bermant, 2017:137).

Despite such research, the “expansion of sophisticated migration and border control infrastructures” (Amelung et al., 2020:587) and its consequences have not been addressed sufficiently from a socio-material angle. Walters (2016) has underlined how, despite the proliferation of deportation spaces and infrastructures such as aircraft, the transport involved in deportation remains something of a black box; this is despite deportation flights being sites of tactics and politics where “deportation [happens] as [a] reiterative performance of bordering” (Walters, 2016:437). We must therefore conceptualize such sites and their infrastructure not simply as instruments used to carry out deportations “but [also] a[s] dense socio-material domain[s] and terrain[s] of struggle” (Walters, 2016:437).

This paper, then, seeks to address the interdependencies between local sites of deportation and deportation procedures, arguing that while immigration enforcement might (re)appropriate sites for deportation, such sites also exert an influence on proceedings, opening up space for resistance. I explore how such processes play out at three sites of removal in Switzerland, some of which are long-established and some of which have only recently been appropriated by immigration enforcement. In discussing the three sites I analyse how infrastructure disciplines individuals by both partitioning and containing mobility (Mezzadra and Neilson, 2013) and supporting it, e.g. during removal procedures. In the interests of enriching my picture of localized tactics I look not only at the design of the components that make up the system – that is, its architecture – but also at these components' relationships. What is crucial is the knowledge produced within the different sites and the power relations that deportation from them enacts. My analysis complements Hiemstra's (2013) picture of chaotic geographies, in which organization and locality play a crucial role.

3 Context and data

In recent years, frequent changes in the law have created a multiplicity of immigration statuses and restrictions on the possibility of claiming asylum in Switzerland. Generally leaning towards the more restrictive end of the spectrum as far as migration governance is concerned, the past decade has also seen public referenda on the question of the “automatic deportation of foreigners who [have] committed certain crimes, regardless of the imposed penalty” (Global Legal Monitor, 2016). As a result, various groups of noncitizens have become deportable; refused asylum seekers and irregularized individuals (e.g. those not applying for asylum but without a legal right to residency) stand out in terms of numbers and are therefore the focus of this contribution.

Switzerland is a federal state divided into 26 cantons, and deportation procedures are implemented by a mixture of national and canton-level bodies. The national authority that decides on asylum applications, the State Secretariat for Migration (SEM), is charged with transferring responsibility for asylum seekers refused at the national level down to cantonal level in cases where responsibility has not already been devolved during the asylum-application process. Once responsibility is transferred, canton-level migration offices and cantonal police units specializing in deportation assume responsibility for implementing removals.² Administrative staff in migration offices arrange meetings with refused asylum seekers during which they inform them about the deportation procedures, including the steps involved. They explain the different levels of force that can be used; these range from unaccompanied deportations through accompanied deportations on regular flights to, finally, so-called “special flights”, where planes are chartered for the sole purpose of deportation (this takes place on both a national or an international scale). Cantons which possess a more specialized administrative structure have specialized police units charged with implementing removals (migration offices do not implement removals themselves but rather delegate the task); officers in such units have received training as so-called “flight escorts”. These specialist units sometimes carry out functions usually performed by the migration office, such as meeting with individuals to inform them about the various steps in the deportation process. The same police units also provide support to smaller cantons which undertake fewer deportations (usually because they play host to a smaller number of foreign nationals per se). They also take part in Frontex-led deportations and other transnational removals and undergo regular training (e.g. in restraint practices). There is also a national database of police officers who have expressed an interest in undertaking such training or who are currently available to assist with deportation. Flight booking is done by a national-level entity that is part of the SEM – swissREPAT.

Over the past decade the impact of deportation procedures undertaken in various locations has made the local news in Switzerland. Stories have included that of a woman who was deported from an emergency shelter despite being in an advanced stage of pregnancy (Fanconi, 2017) and of a young man who jumped from the window of a transit centre in an attempt to escape a deportation that had been initiated with little warning (Peternell, 2016). The Canton of Zurich deported a suicidal patient from an asylum centre just days after he had left a psychiatric clinic, despite psychiatric and other medical certificates testifying to his condition. Such certificates are not binding unless issued by OSEARA (Roth, 2017), a company contracted to undertake medical evaluations and assessments of “transportability” in deportation cases. There have also been cases of individuals being arrested for deportation from churches despite having received church asylum (Läubli, 2016; SRF, 2019). Yet there is little publicly available information on deportations, and no statistics exist on the number of deportations taking place from particular sites; it is left to researchers, activist networks, or the occasional news report to inform the public about such procedures.³

The following ethnographic data were collected between 2015 and 2017 in several Swiss immigration enforcement agencies. They include interviews and observations conducted in one migration office (fieldwork took place over 4 months), one cantonal prison (a day of fieldwork), and one cantonal police unit (6 weeks of fieldwork); interviews were also conducted in a second migration office as well as with the administrator at a hospital and with border police officers processing the cases of noncitizens apprehended at airports or on trains. Access was granted after an official request to conduct fieldwork at the described sites, allowing me to follow the everyday work of various departments and to engage in informal conversations throughout the working day. Influenced by prior state ethnographies that have taken an interest in the power dynamics within organizations and between state actors and the public (Dubois, 2016; Eule, 2014; Lipsky, 1980), this approach involved “studying up” (Nader, 1972) by following bureaucratic actors during direct contact with “clients of the state”. This was done with the aim of understanding how policy and law are implemented “on the ground”, paying particular attention to administrative struggles and the role of discretion and materiality in everyday bureaucracy (see Borrelli, 2020, for more on data collection, including a critical reflection on positionality).

I have translated data from interviews originally conducted in German; in doing so I have tried to leave sentence structure and the quiddities of speech intact wherever possible. All interviews and observations were anonymized according to the terms of a confidentiality agreement between me and my interlocutors. Since I also had a methodological interest in the way staff made sense of me as a researcher, I left it to them introduce me to third parties, including migrants encountered during removal procedures – who were therefore, at times, not informed about my role. Such encounters – between migrants and immigration authorities – are often brief and emotionally charged, potentially rendering my presence less significant. Conducting fieldwork in this way poses an ethical dilemma, yet the scant methodological scholarship discussing third-party consent has tended to conclude that not doing research in such contexts would mean that the only voices heard would be those of the actors implementing removal procedures; the critical scrutiny that researchers' observations can bring to bear would thereby be precluded (Fluehr-Lobban, 1994; Hallett and Gruner-Domic, 2019).

4 The sites of deportation

Like other countries, Switzerland does not deport all noncitizens who do not have a right to remain. Rather, only a small number are deported, rendering deportation a “performative [action] in that it disciplines the un-deported majority by investing illegalized migrants with the fear of being deported” (Casas-Cortes et al., 2015:84). In recent years there has been significant political will to increase deportations, especially in the wake of an increase in asylum applications in 2015, but such ambitions are not easily realized. Deportation practices are influenced by a vast international infrastructure on a European as well as a global level. Both Dublin deportations within the Schengen Area and deportations to so-called third countries depend – among other things – on the willingness and responsiveness of receiving states. Securing this willingness and responsiveness often involves the negotiation of complex return agreements (Pillet, 2015), which may be more or less formal in nature. Other obstacles to deportation include politically determined deportation exclusion lists, which prevent the deportation of individuals to specific countries. The transnational imbrication of deportation practices and the connections between place, states, and international relations (Geiger and Pécoud, 2010) become more tangible when attention is paid to the ways in which noncitizens can be restrained on deportation flights. Frontex's rules differ from those of nation states (field notes, police unit, 2017), and police officers need to comply with both. One police officer returning from Frontex training showed me a table cross-referencing restraint methods and national regulations and covered in ticks (denoting things permitted in a specific national context) and crosses (field notes, 2017).

Local contexts also influence, in a very practical sense, how deportations can be carried out. Not every canton in Switzerland has local access to specialized detention facilities for noncitizens awaiting deportation (though during my fieldwork, efforts were being made to put in place such a system). When migrants are detained they are therefore likely to be placed in prison facilities, where they are not always housed separately from other prisoners (though they may be held on separate wings or in separate sections). Families may be held in specialized accommodation prior to deportation, though this was not a regular occurrence. People facing deportation who have not been detained may live with their families; others live in state-owned or private accommodation facilities or in emergency shelters.⁴ Some noncitizens may remain in hospital if they are receiving medical treatment. As a result, when a deportation is “in train” (meaning all relevant documents have been obtained, a flight has been booked, and the individual in question assessed as “fit to fly”; see also Malka, 2020), immigration officers and police officers need to factor into their planning the different locations where the prospective deportee might need be held en route to their removal.⁵ In some cases – e.g. where noncitizens are apprehended by force or escorted to a deportation flight but allowed to board independently, or (for those being forcibly removed) where deportees are escorted to the airport and accompanied throughout the entire flight – the logistics of this can be complex. The place of apprehension and its architecture are not the only relevant factors; officers must also consider equipment (e.g. restraint devices) and transportation (vans, cars, trains, and aircraft), making sure that all of this material infrastructure is working properly. This involves not only attention to technical functions (e.g. do the handcuffs work?), but also the planning of routes to avoid traffic (e.g. in order to make a flight in time). Also involved in the process are the airport facilities used to detain individuals before deportation, as well as airport and airline staff and members of the public; a pilot, for example, may refuse to take a “deportee”, while passengers may boycott airlines that undertake deportation charter flights (Walters, 2021).

As such, I follow Walters' (2017) expanded definition of migration infrastructure (based on Xiang and Lindquist, 2014) as “systematically interlinked technologies, institutions and actors that facilitate and condition the forced movement of persons who are subject to deportation measures, or the threat of deportation” (Walters, 2017:2800). Most importantly, I want to highlight the relationality of such infrastructure, for example, how it can influence the outcome of a deportation procedure and how it connects various groups of people who might use the infrastructure differently (e.g. at airports, the same infrastructure might be used by tourists and deportees alike). Importantly, given that sites differ in their accessibility and usage, police officers depend on the cooperation and knowledge of staff working in the different facilities. In the following three subsections I will engage with three spaces of deportation, analysing the negotiations that take place between immigration enforcement staff and individuals working at the three sites while also painting a picture of the infrastructure of the three sites in order to highlight their relational socio-materiality.

4.1 The prison as a classic site of deportation

Prisons or detention facilities cater to the agendas of state agencies in the sense that they are governed by strict regimes subject to a high degree of organization. They are part of the public sector (even when privately run) and operate according to a coercive system that is characterized by the many rules that staff and inmates must follow. Each day is alike, structured by the times allotted for meals, courtyard walks, visiting hours, and prisoners' meetings with lawyers or other officials (e.g. police officers informing deportees of their removal). Prison guards, too, are subject to the tight rhythm of prison life. During my visit every guard except one was wearing a wristwatch and would regularly check the time (field notes, 2017). The actions and whereabouts of detainees are predictable; this is of considerable importance during the deportation process as officers may be required to counteract noncitizens' attempts to harm themselves or state actors while resisting removal.

The prison I visited has one designated wing for people held in so-called deportation detention (Ausschaffungshaft), but this group is not strictly divided from other inmates, who may be housed on the same wing. When the designated wing reaches capacity, noncitizens are also accommodated on other wings. Despite a regimented daily routine that is conducive to deportation planning, such prison facilities are not specifically designed to accommodate noncitizens who might, at most, have flouted an administrative regulation. By contrast with immigration detention facilities, there is often little or no free movement on such wings; due to this inadequate and unsuitable infrastructure, migrants experience their confinement as a punishment.

During my visit I asked one of the guards about the differences between deportees and other inmates. He answered as follows:

Well, they are treated equally, but they are allowed to do different things. For example, we always have to consider who is allowed what. It is often difficult for them to understand that they just have to wait here. They say they are not criminals and ask why they are here. Often they are just here in advance of a visit to their embassy; after their visit, they are free to leave.

In the cases he is referring to, migrants are apprehended; placed in detention to make sure they cannot abscond; and then taken to a meeting at their national embassy to, for example, establish their nationality and to secure travel documents ahead of deportation.

Or they are waiting for identity checks to be completed or for their flight to be booked. (Field notes, prison facility, 2017)

Deportees are therefore a group that present a special challenge to prison guards as they go about their work. The prison environment may facilitate the work of immigration officials, but noncitizens in administrative detention are not a prison's regular “clientele”; as a group they are differentiated from other prisoners, and their very presence can call into question the legitimacy of the prison environment. The guard goes onto explain what he means by needing to think about “who is allowed what”: he is talking about differentiated rights. Those awaiting deportation are allowed 2 h of visits per week and may have direct physical contact with the visitor; other inmates have only 1 h, during which they remain divided from their visitor by a glass wall. Here laws, spaces, and materials (e.g. the presence or lack of a glass wall during visits) interact (see Aradau, 2010). Spaces thus change according to their users; they can be used in a variety of ways and become open to “new clientele”, e.g. noncitizens awaiting deportation or police officers implementing immigration law.

Police officers are generally familiar with prison architecture and can use the environment and its structure to their advantage. On the ground floor of the prison are several rooms in which police officers meet with detainees. Officers arrive at the counter, leave their ID, give the name of the person they wish to see, and then enter the facility without needing to be escorted by staff. They are familiar with the room designated for their use, which is laid out simply: a table, some chairs, and a wall display of forms translated into different languages. A printer is directly outside in the hallway, allowing officers to copy the forms that deportees have signed while keeping an eye on them through the open door. Prison staff escort the noncitizen to the room at the beginning of the meeting and back to their wing at the end of the meeting. A special security cell is kept in the basement to be used in cases of self-harm or other violent behaviour. Next to this cell is another room that is used by the police unit to prepare noncitizens for deportation. Once removed from their (regular or security) cell, the noncitizen is taken to this room, seated in a chair (they may also be restrained), and informed about deportation procedures. In such moments it is not prison guards who accompany noncitizens but rather police officers specifically trained in using force to subdue bodies. Such officers need to be familiar with the spaces in which they are operating, including this cell; it is useful for them to know how the furniture is arranged or which sets of stairs they will need to take to reach which room. Before undertaking a deportation, police officers often rehearse the different steps mentally; they may discuss timings (e.g. which officer will enter which room and when), officers' physical positioning with respect to one another, and the anticipated physical location of noncitizens within the relevant cell. They are familiar with the prison regimen; if a deportation is due to take place early in the morning, they will know (or assume) that the noncitizen will still be in bed. Based on such information, decisions are taken about where officers should be positioned to minimize resistance or self-harm.

In contrast to other sites of deportation – such as private homes or hospitals – prisons are spaces about whose structure and design police officers are highly knowledgeable. While officers highlighted to me the need to consider carefully every aspect of the rooms and other surroundings in which deportations occur, including for contingency planning in case of unforeseen circumstances, they can at least count on knowing the size of a cell; the number of steps and floors they will have to negotiate to get from their vehicle to the cell and back; and the support of prison guards, who maintain a regular dialogue with immigration officials. The controlled environment of the prison is thus conducive to forced removals; its mundane infrastructure serves as a kind of epitome of the state power that penal and detention facilities help maintain.

4.2 Asylum accommodation, hospitals, and psychiatric facilities as less adapted sites of deportation

Prisons function as sites of control, facilitating deportations through, for example, their physical layout and coordination around schedules. Hospitals and asylum-accommodation facilities, on the other hand, are not designed with forced removals in mind, even though they are not necessarily new sites of deportation in a global context. In Switzerland, as previously explained, removals from such sites have been the subject of public debate and are by no means a regular occurrence.

During one deportation from a former country hotel, scenically situated at the peak of a small mountain and originally built for vacationers seeking a weekend escape from city life, the officers involved remark on the views and rural peacefulness of the facility, which is currently in use as asylum accommodation. Driving uphill on a winding road, we pass residents of the facility who are struggling to make their way back to their accommodation; there is no public transport, and they are sweating in the summer heat (field notes, police unit, 2017).

This former vacation space has been appropriated by the immigration regime as a shelter for asylum seekers; it now houses individuals at various stages of the asylum process. Also, a second order of appropriation comes into play when deportations are implemented from such places. Due to a lack of public transport, which diminishes the agency of the those accommodated in asylum accommodation, such locations also become sites of confinement. Residents have little prospect of absconding; and little public attention is directed to such peripheral spaces (see Sheller and Urry, 2000 for a discussion beyond migration studies of the double meanings that spaces can acquire). In such spaces, the scope for noncitizens facing deportation to strategically disrupt the often finely tuned ecologies of removal (Walters, 2021) is extremely limited. By a similar token, supportive groups such as nongovernmental organizations (NGOs) or activists cannot easily gain access to such remote places, further reducing the possibility of resistance. At the same time, such spaces are also unfamiliar to immigration officials; their architecture varies considerably, and it is uncommon for deportations to regularly be initiated from the same site. Unlike prison guards, the staff at hospitals or asylum-accommodation facilities retain considerable agency.

Similar tension can come into play at hospitals and psychiatric facilities. Staff (e.g. in medical clinics) and the public (e.g. demonstrators) may occupy or otherwise compromise infrastructures; by doing so they may seek to contest individual decisions or the broader deportation infrastructure, as well as challenge their own complicity in the deportation regime. Combined with individual acts of resistance (more likely in some spaces than in others), public scrutiny creates uncertainty for the officers involved in implementing deportations. Such personnel depend heavily on the support of nongovernmental actors – e.g. clinic staff or the social workers in charge of running asylum facilities – or, at times, local police forces with detailed knowledge of the terrain on which they are operating.

In cases where an OSEARA assessment has concluded that an individual is fit to fly, their being an inpatient in a hospital is no impediment to their removal. The hospital where I witnessed deportation proceedings in action is located on the outskirts of a large Swiss city. It is a typical hospital building. A large entranceway with sliding doors leads to a reception area. Beyond reception, various corridors and elevators conduct visitors to the different departments. Here, collaboration between hospital staff and police officers extends to noncitizen patients being conducted to a room very close to the hospital entrance. They are taken to this room by hospital staff under the pretence of a medical examination; it looks like any other examination room. Here police officers await the noncitizen, who is promptly informed that they are to be deported. Restraint may be used to ensure their compliance. This may include not only handcuffing, but also full-body restraint, which involves an individual being forced into a seated position and carried by officers (see Angst, 2011, for a video mimicking the force used in a deportation procedure).

Deportations from such places capture the gaze of many non-migration-related actors, including members of the public. Patients, hospital visitors, medical staff, and administrators can all end up as witnesses to a deportation, including a noncitizen's resistance. Police officers will therefore park their vehicle – often a van – close to the hospital entrance and seek to swiftly manoeuvre the noncitizen being detained from the small room into the car (often through the back door of the van). Between the room and car, the deportee can nonetheless be heard and seen. The officers tell me that they try to arrange for deportations to take place early in the morning, when there will be fewer people (e.g. visitors) around who might get in the way physically or witness the deportation proceedings. Hospital staff appreciate this precaution because, as a hospital secretary told me (interview, 2017), they tend to feel uneasy about their complicity with immigration enforcement. Yet – the secretary explained – they have no choice but to collaborate due to the agreement that is in place between the police unit and the hospital. According to this agreement, staff are not allowed to warn patients about their removal date. Given the implications of this for the patient–physician relationship (and also for trust), a quick removal is in the interests of the hospital. Hospital staff thus become outsourced arms of immigration enforcement, while also feeling uneasy about the deportation practices and technologies to which they bear witness. As in a prison, the routinized environment of the hospital facilitates deportations.

Yet hospitals and clinics can also turn into spaces of resistance, however fragile. During a weekly meeting of the police unit, one officer informed his colleagues about a case that had caused confusion. He had been overseeing the deportation of a young man who walked on crutches. The young man had opted to leave “voluntarily” and was due to be accompanied to an airport of a bigger city, from where he wanted to continue his journey unaccompanied:

He walks slowly on crutches and had previously been admitted to [name of clinic]. Now we wanted to pick him up for an accompanied deportation; we went with a support worker and the police to his place in [city name], but at the agreed pickup time he was not there. His room was empty, and the question arose of whether he might not be back at the clinic. (Field notes, 2017)

In this case, moving between places becomes a strategy through which noncitizens seek to prolong their stay and avoid deportation; private homes and hospitals are converted into spaces of evasion, and hiding is employed as a strategy. In a certain sense, migrants who use hospitals as places to hide in order to avoid deportation end up mimicking state strategies that have been termed “chaotic geographies” (Hiemstra, 2013). Rather than states making use of disorganized bureaucratic procedures to induce a sense of chaos or suddenly transferring individuals between detention facilities as a way of asserting their power, noncitizens move between places to make it more difficult for officials to ascertain their whereabouts. In this case, the failure of the police officers to find the individual in question allowed him to at least defer his removal. The case also underlines a lack of communication between immigration enforcement and non-migration-related sites of deportation. State actors told me about migrants' strategic use of hospitalization in cases where a “partial deportation” was being considered, i.e. where one member of a family – often a husband and father – is deported in the hope that his family will follow more voluntarily.

“We always have the option to partially deport”, says the head of the police unit. However, one officer has decided against this in one of his cases because of the poor health of the mother:

There's a risk that she will be committed to a clinic – a psychiatric ward – and then we will have to take care of the children.

One of the children was subsequently hospitalized. The team speculates about whether the hospitalization was a coincidence. One officer explains,

This is the way it usually goes: for example, the older child might have harmed the younger one in order to stop the deportation.

The unit head interjects,

Well, to me it seems like a complete coincidence in this case. Although we still don't not know what the child is ill with. We've just been told that the child absolutely cannot fly. (Field notes, 2017)

Deportations from sites that are not classic spaces of enforcement therefore become implicated in a discourse on secrecy. It is in the interests of immigration officials to conceal such removals, but they have no choice but to engage with the multiple actors that have the power to influence deportation proceedings. Undertaking deportations from such sites creates new challenges for state agents, as well as facilitates the resistance strategies of migrants. This underlines the tightly imbricated relationship that binds together such spaces and the human actors that circulate within them.

As a result of this imbrication, such sites come under increased scrutiny. Prisons and detention facilities can more easily be “weaponized” to enforce deportations because enforcement is embedded in such structures, with their fixed schedules and tightly surveilled inmates. By contrast, places such as hospitals, asylum-accommodation facilities, hotels, and private homes operate according to looser regimes; they also play host to many more people who might become obstacles to a “smooth removal” (rasche Anhaltung), as officers often call it (field notes, 2015–2017).⁶

4.3 The airport – a relational and segregated site (of deportation)

The police deportation unit is in charge of the repatriation of people arriving in Switzerland from abroad as well as extradition from Switzerland to other countries … . Most such missions [Aufträge] are implemented via air and are subject to very tight security regulations. Flight escorts will provide support on the ground in [airport name], during the arrest [Anhaltung] of deportees in regional prisons, and when deportees are being escorted to airports in Switzerland. (Extract from a manual for flight escorts, 2013; own translation)

Airports serve as a junction or central node in all deportation procedures; they are the nexus through which deportees pass, whether they are apprehended at home or in hospital or brought from prison. As such they can come to resemble parallel worlds in which several realities exist simultaneously. Tourists are classed and differentiated, with some enjoying lounges, priority boarding, and superior food while the majority are excluded from such privileges. At the same time we also see airports becoming “laboratories for new strategies of both technological and social control” (Salter, 2008:xi) as they facilitate the high degree of surveillance and policing required to differentiate between unwanted and wanted migrants (so-called “smart borders”; see Salter, 2004). Such social control is enacted in a site that is situated within a nation state, yet it operates at the margins of the law (Salter, 2008). Such sites contribute to the creation of irregularity.

Certain ironic similarities exist between otherwise highly differentiated groups of passengers. When it comes to movement, airports have two fast lanes – one for the wealthy and one for the deportable (Walters, 2002). Deportees are driven to the airport in regular cars or vans. Upon arrival they are conducted to their flight along special corridors that are invisible to other passengers; a certain similarity to priority boarding suggests itself. During one deportation, we arrive at one of Switzerland's larger airports and follow a route through its premises that is entirely different to that taken by “regular” passengers. We go in through a special entrance, outside of which our car is allowed to park directly. In this parallel world, parking spaces are located closer to entrances; indeed, an entire structure that I did not know existed opens out in front of us. Tourists, students, deportees, and commuters share the airport space just as they share aeroplanes (Walters, 2017), but their access to such spaces is differentiated. The structure that opens out in front of me has been created especially for the transport of deportable individuals, yet many actors interact with it. The building contains holding cells where noncitizens await removal. On one occasion, a detainee asks to smoke and is allowed to do so inside his cell. The rules of this “parallel” airport are similar but not the same. Luggage is checked and conveyed via luggage belts. From this space migrants are escorted onto either regular passenger flights or special flights (field notes, 2017).

What links prisons, detention facilities, hospitals, and airports is the structured setting, which includes arrival and departure times (e.g. visitor hours, flight schedules, meal times) and the power the authorities enjoy to restrict people's movement or keep them in specific areas (or cells). Airports are crucial infrastructure points for the confinement and observation of noncitizens awaiting deportation. Cells need to be secured, potential risks assessed, and luggage examined. The sites I have described are thus not only transformed by the practice of deportation; they also exert an influence on deportation practices. They reveal, moreover, how an expanded deportation infrastructure has come to permeate everyday spaces inside nation states and beyond.

Being separated from other patients at the hospital, from tourists at airports, and from other inmates or detainees in prisons and detention facilities “increases deportees' vulnerability to violence and abuse at the hands of security actors” (Walters, 2021:10; Fekete, 2011). Alyssa Sambor (2021) discusses the spatial and structural tactics deployed by Immigration and Customs Enforcement (ICE) in the USA to conceal and expand already vast infrastructural networks of immigration control. Sambor focuses on the use of flights, arguing that aeroplanes have become mobile sites of migrant detention – spaces both obscure and hyper-visible. As deportation enacts the border, aviation infrastructure becomes a materialized text of enforcement rhetoric and anti-immigration discourse (Sambor, 2021).

Considering these three exemplary sites of infrastructure involved in deportation using a socio-material lens helps to show how spaces and architectures, on the one hand, are appropriated by immigration enforcement and, on the other, influence how removals are implemented. Key factors include the way such spaces are built; their accessibility for various actors; and their relationships, which are always mediated by the “actor in common” in all efforts to remove noncitizens: immigration officials. Such spaces therefore enact a relationality in respect of one another whilst retaining a situated role. Looking at the path traced by removals from private accommodation or asylum housing via hospitals and/or prisons to the airport (and, in some cases, back again) lays bare the “jagged” trajectory of the deportation regime. Deportation is a highly networked phenomenon, not only at the local level but also in respect of the part it plays in global processes of exclusion. Technologies and infrastructures have facilitated the contraction of space, rendering the global local (Müller, 2015), with airports serving as the connective tissue between nation states and translational sites.

5 Discussion and conclusions

In summary, “[d]eportation infrastructure” broadly encompasses “the systematically interlinked technologies, institutions, and actors that facilitate and condition the forced movement of persons who are subject to deportation or the threat of deportation” (Sambor, 2021:79; see also Walters, 2017). It consists of complex practices of coercive removal (Walters, 2021), all of which need infrastructure in order to function (Blue, 2015). Various scholars have made use of actor–network theory and the concept of assemblage to advance an understanding of the active and interactive roles played by human and non-human actors in shaping the field (for an in-depth discussion of both concepts, see Müller, 2015). Such approaches help capture the multilayered and multiscalar role of the spaces, buildings, and tools involved in the deportation regime in structuring and controlling migration (see Lin et al., 2017). Importantly, “[a]ll these technologies have to be understood in the context of the ideological regime that controls them” (Mattern, 2018; see also Ticktin, 2016). Deportation infrastructure is therefore political in nature, as well as affecting and informing the ways in which immobility is created (Lin et al., 2017).

This paper set out to study infrastructure and infrastructural sites involved in deportations with the aim of arguing for their significance for our understanding of deportation and as technologies of immigration enforcement writ large. The way in which such sites materially shape deportation procedures (e.g. through their location, structure, and the agency of the actors involved at each site) makes them situated. Yet there is also a relational aspect: on the one hand, infrastructure is bound to a network through the common aim of removing noncitizens and the requirement for engagement with immigration enforcement. On the other hand, since all three sites pre-exist their use in deportations, they each offer opportunities for resistance. Adopting a socio-material lens, this analysis has shown how techniques and devices not only can become political through their use (i.e. their instrumentalization) but also may have social-control potential built into their design. Contesting social control therefore means contesting the very development of such sites (e.g. detention facilities; see Müller, 2015, more generally).

This in-depth examination of prisons, hospitals and psychiatric wards, asylum-accommodation facilities, and airports has shown how, although the structure and organization of such sites pre-exist their use in deportations, they may nonetheless serve the goals that police officers and immigration officials are aiming for. Scholars writing in other contexts have stressed the lack of order and high level of discretion characteristic of the deportation regime (Kalir, 2014; Walters, 2017). Yet these two dimensions are not mutually exclusive. Infrastructures can be both mutable and dynamic and highly structured and regimented: both aspects are instrumentalized by enforcement authorities to gain the upper hand in the struggle over removals. Immigration officials' desire to enact orderly and predictable removals and the use of infrastructure by noncitizens and other actors to resist deportations render such sites both chaotic and organized; in laying bare these two orders of weaponization, my analysis enriches the existing picture of how localized tactics can shape efforts to control migrants' movement.

Bodies and interactions, then, are shaped by socio-material practices and spaces of deportation, influencing noncitizens and bureaucrats as well as the other actors that become embroiled in such processes. A focus on infrastructure allows for an understanding of how different sites are repurposed and appropriated by migration governance and noncitizens alike. This approach “de-centers the state within migration” (Walters, 2017:6); it does not seek to ignore the state's power but rather to allow the analytic gaze to additionally take in non-state actors, as well as spaces, practices, and material not directly related to deportation. The ways in which actors relate to the tools and spaces of deportation and interact with one another through them shape immigration enforcement and its outcomes.

Looking at different sites has allowed me to trace both how spaces are shaped by removal practices and how individuals working in such infrastructures either adapt or are already attuned to the work of immigration enforcement. It is worth noting, however, that the three sites under analysis in this paper are subject to a high degree of administrative structure as well as being, in the main, state-run. The influence exerted on deportation procedures by private homes remains to be studied.

Materiality nonetheless emerges as a key element in such practices. Looking at infrastructure may also help in understanding why discussion of the highly violent and disruptive nature of deportation tends to be neglected in favour of the analysis of technologies. Given the large number of actors, sites, and tools involved, current debates tend to centre on how to improve deportations in order to make them more “humane”. Such conversations obscure the key question of the rightfulness of such state practices.