Introduction

Disappeared from the shared spaces of social life, dismissed from membership in imagined forms of stable community, they are left to drift in and out of public perception; constructed as organised criminals and moral menaces by the media and by the law, they haunt the margins, circulating more as an imaginary presence than an identifiable threat (Ferrell, 2018: 187).

The quote above is a striking example of how using the spectral as a conceptual metaphor may help us to focus on those who haunt the margins, who are drifting in and out of public perception and constructed as a threat. As del Pilar Blanco and Peeren (2013: 1) argue, such spectral conceptual metaphors ‘perform theoretical work’, becoming an analytical tool which ‘does theory’. Conceptual metaphors furthermore evoke discourse and systems of producing knowledge by their dynamic and comparative nature (Bal, 2010). In this paper, re-reading our previous material and studies, we use the spectral as a conceptual metaphor to provide a reading of the spectre of ‘vulnerable EU citizens’. Important to highlight is that ‘vulnerable EU citizens’ is a categorisation ascribed to Roma or people perceived as being Roma. Here, both old and new discourses, mechanisms and practices are mobilised to legitimise treating the Roma differently to other ‘intra‐EU’ migrants or ‘mobile citizens’ (Baar, 2017), demonstrating an effective othering of this group of people.

In this paper, we explore how the spectre of ‘vulnerable EU citizens’ used in Swedish welfare institutions and policy discourses produce a ‘ghostly’ existence of certain people. We will argue that such narratives, evident in Sweden’s welfare institutions and policy discourses, are sites of necropolitical power, enabling violent state sanctions. We do this by applying a critical, hauntological perspective to explore how the concept of the spectre as well as temporal and spatial displacement can be used to perform ethical critiques of social injustice, exploring the work of the spectre of ‘vulnerable EU citizens’ in Sweden as a contextual example. The use of the spectral as theoretical motifs has been utilised in contemporary social sciences to describe social phenomena that have a ‘ghostly existence’. After the expansion of the European Union in 2007, migration from Eastern European countries to Sweden increased. The so-called freedom of movement granted to EU citizens made it possible for all EU citizens to seek their livelihood beyond the borders of their home countries. Between 2014 and 2015, the divisive and often prejudiced response to poor EU citizens begging for money in the streets in Sweden became an urgent political issue. Since then, the number of visible beggars present on the streets of almost all Swedish towns and cities has declined. Many explanations for this decline have been offered, but the simple reduction in beggars as a common sight outside Sweden’s shops and public buildings does not necessarily mean that the number of people falling under the category of ‘vulnerable EU citizens’ has decreased. They are merely elsewhere, either within or outside Sweden. There are no binding national guidelines on how Swedish authorities and municipalities should deal with the people labelled ‘vulnerable EU citizens’ and, in the absence of a common policy, local authorities, government agencies and non-governmental organisations have taken their own decisions on what actions should be taken (Giritli Nygren & Nyhlén, 2017; Nyhlén, 2020). One often cited explanation of why the people referred to as ‘vulnerable EU citizens’ have become less of a fixture of the urban landscape is a ‘hardening social climate’ (Hansson, 2019). This means that Swedish municipalities increasingly choose to evict EU citizens and prosecute them for setting up illegal camps. In this article, we explore how processes of spectralisation have enabled such actions, exploring how the conceptual metaphor of the spectre may be used to further our theoretical understanding of how the spectre of ‘vulnerable EU citizens’ keeps specific minorities out of sight.

If we are to describe how our previous research on the spectre of ‘vulnerable EU citizens’ can be used to analyse spectralisation processes, we should first define who may fall under the categorisation of a ‘vulnerable EU citizen’. The term ‘vulnerable EU citizen’ is the one stated in the Swedish Government Official Report (SOU, 2016: 6), in which ‘vulnerable EU citizens’ are defined as ‘individuals who are citizens of another EU country without a Swedish residence permit’. In the public discourse however, the term ‘EU migrants’ is more commonly used to refer to ‘vulnerable EU citizens’. To fully understand the situation, this categorisation must be viewed as accommodating (and partially concealing) a long history of racism, persecution and discrimination (Yuval Davies et al. 2017). Involving exclusive practices, this othering may even involve monstrous connotations, framing the Roma as a security issue in terms of being ‘inferior’, ‘criminal’ and ‘dangerous’ (van Baar, 2017: 450). The Roma, for instance, have been characterised as ‘ethno‐tourists’, ‘asylum adventurers’ (Vašečka & Vašečka, 2003), ‘poverty migrants’ and ‘intrusive beggars’ (Benedik, 2010). Vincze (2014) calls European Roma ‘an ethnic minority without a territory’. We thereby argue, in line with Teodorescu and Molina (2021), that it is not possible to approach any issue involving people from the Roma population without considering racial relations and racism.

Racism against Roma has been studied by scholars for a long time. Roma presence in Europe is marked by the stigma created during long periods of racialisation and de-territorialisation, including the current EU treatment of European Roma as an ethnic minority (Vincze, 2014). Public records on Roma presence in Sweden from the sixteenth and seventeenth centuries demonstrate how laws once entitled landowners to kill Roma trespassing on their lands. More recently, entry bans put in place between 1914 and 1945 forbade non-Swedish Roma from entering the country, and Roma families were not allowed to settle in any Swedish municipality for longer than 2 weeks. Forced sterilisation was also performed on Swedish Roma between 1934 and 1976 (Teodorescu & Molina, 2021; Hansson 2019; SOU, 2016). Furthermore, the Swedish handling of ‘vulnerable EU migrants’ had emancipatory traits in the post-war era up until 2015. For example, Roma first received access to Sweden’s welfare and school system in the 1960s (Montesino, 2012) when they were recognised as one of five national minorities (Teodorescu & Molina, 2021). However, one of the most explicit expressions of renewed anti-Roma racism reappeared in Sweden in the shape of illegal police registration of Roma ethnic identity. Importantly, the ongoing criminalisation of Roma through the framing of ‘vulnerable EU citizens’ as a public order as well as a sanitation issue happened simultaneously as the release of a much noted ‘White Paper’ (DS, 2014: 8) by the right-wing coalition government which described state-sanctioned abuse and rights violations committed against Roma and Travelers during the long 1900s. The publication in itself was framed as a decisive break with the long history of state racism towards Roma and Travelers (Persdotter 2019). The framing of harsher politics and criminalisation of, for example, begging as the morally correct thing to do are ‘rationalised stories' used by states—in this case, the Swedish state—‘to avoid moral consequences, in which monstrous punishments are the only logical responses to monstrosity’ (Higgins & Swarz, 2018: 92). Yuval Davies et al. (2018) point to how we can see the emergence of a frame that posits the Roma as a people that exist everywhere but belong nowhere. This process has been going on for a long time, and the Roma have been effectively banished from the imagined communities of European nations (Andersson, 1982), including Sweden. As such, the use of spectral narratives pertaining to marginalised groups may function as strategies of national border defence, saving Sweden for ‘the Swedes’ and constructing state violence as a justified response to the deterioration of the nation.

In the following sections, we engage with spectral conceptual metaphors and the methodology of excavating ghosts. This is followed by the theoretical themes we can deploy from re-reading our previous material and studies.

Research Context—Material and Previous Studies

The previously gathered material and previously published studies we re-examined for the purpose of this paper were all originally gathered within the same research project. In the project, inspired by anthropological policy analysis, a rich and diverse material was gathered, including participatory observations, interviews, conversations, policy documents, official reports, law rulings, newspaper articles and field notes. The research project aimed to follow a welfare project, located in one of the counties in Northern Sweden, working with social inclusion and empowerment in order to create a better future for so called vulnerable EU citizens staying temporarily in the northern parts of Sweden. The project followed and interviewed people working with ‘vulnerable EU citizens’.

Based on this large empirical, partly ethnographic material we have published several texts, described in more detail below. In Normalising Welfare Boundaries - A feminist analysis of Swedish municipalities’ handling of vulnerable EU citizens (Giritli Nygren & Nyhlén, 2017), the aim was to study how officials and politicians normalise certain beliefs with regard to the (im)possibility of providing support to socially excluded EU citizens residing in their municipalities. Drawing on interviews with nine officials and six politicians from different municipalities where they were asked to reflect on their municipality’s handling of vulnerable EU citizens, the analysis showed that they most commonly referred to the delimitative and prohibitive aspects of their work (Giritli Nygren & Nyhlén, 2017). When re-reading the result from a ghost ethnographic perspective, it is possible to interpret the informants’ comparisons between the category of ‘EU migrants’ and the categories they have at hand not only as a search for order but as a process of making them uncanny. As a concept originally defined by Freud (1919/2001), the uncanny has been defined as the familiar becoming strange or unfamiliar, the blurring of boundaries, the unknowable and that which disturbs and disrupts (Fiddler, 2013). The uncanny both separates and unsettles borders while simultaneously repressing that which threatens those boundaries.

In the study [A Place to Stay Without Being Chased Away. Negotiations of Risk and Social Security in the Case of ‘EU Migrants’ in Sweden] (Giritli Nygren, 2016; Giritli Nygren et al., 2019), a discursive reading of articles on EU migrants from the local newspaper was undertaken. The overall aim of the study was to use the public debate about EU migrants in the local media to discuss how links between welfare, social security and migration are constituted and produced. The reading illuminated how EU migrants fall through the cracks in the system, where humanitarian concerns were superseded by security ones, based on the notions of goodwill or the common good implicit in the welfare discourse, therbey exclusion and discrimination were legitimised. When thinking of these results through the lens of the spectral, it is possible to interpret the media stories as a kind of gothic narrative of the marginalised, which according to Mbembe (2003a; b) are an important site for necropolitical power: ‘where the narrative framework is written through mechanisms of concealing and illuminating monstrosity’ (Higgins and Swartz, 2018: 104). A large amount of the articles were dealing with the EU migrants’ housing situation, framed as an acute crisis, especially for the children living in the migrant camp. The housing issue, discursively situated as a crisis, was thus something that had to be addressed, posing the question of who should be held accountable, how things should be managed—meaning the risks that the EU migrants presented to the local authority, the community and public order, as well as their own vulnerable position. The source material did not only report about the EU migrants’ circumstances, but did so in subjective, value-charged terms such as dangerous and ‘disgraceful’, thus reproducing ideas about what is, and is not, reasonable housing.

In We should call them our friends” – Negotiations on Welfare and Social Security Entitlements for Displaced EU Citizens in Sweden (Nyhlén, 2020), the aim was to analyse the understanding of displaced EU citizens’ rights by taking a closer look at the ways in which discourses on welfare operate within the nexus of humanitarian care, social care and migration. By focusing on the ways in which discourses on welfare operate, it was possible to outline the role benevolent violence played in relation to project work and social entitlement. The study draws on participatory observations, interviews, conversations, policy documents, official reports, law rulings and field notes. The material is from the perspective of the people working with displaced citizens and not the citizens themselves. In total, 21 different persons, including the project leader, the local project workers (both employees and voluntary workers), social workers, politicians and white-collar workers at different administrative levels, were interviewed, some of them several times. Participant observations as well as long-term association with the project workers in their everyday setting (Agar, 1996) were also conducted. By using this approach, we were able to observe the activities of the governing elite and the underlying logic behind their actions.

What prompted the re-reading of these texts was the recognition and the realisation on our part that these previously published texts participated in hiding the systematic violence against Roma by becoming a part of the discourse on vulnerable EU migrants. As these previously published texts became part of the very problem we had identified, we decided to re-read these texts using a more critical perspective that would enable us to excavate ideas and findings that were previously unseen.

For the purpose of this article, all of this material has been re-analysed, as well as our own previously published articles on this material. Through a re-reading of our own previous material and studies regarding the representation and management of the so-called vulnerable EU citizens in different contexts—a welfare institutional setting (Giritli Nygren & Nyhlén, 2017; Ek et al., 2018), local media (Giritli Nygren, 2016; Giritli Nygren et al., 2019) and within contemporary Swedish policy discourse (Nyhlén, 2020)—we will explore the process of how these individuals are spectralised by current welfare institutions and policy discourses and what this does to the (im)possibility of social justice. The existence of the ‘unseen within the already seen’ is inspired by the method of ghost ethnography. As such, inspirited by ghost ethnography, this study will explore traces of spectralisation, unseen in previously published and gathered materials, to determine what these past textual spectres, collapsed in the present text, may demand of our future (Derrida, 1994). We have turned back to both the empirical material to re-read and re-contextualise the findings, as well as conducted a re-reading of the published material. The re-reading, conducted in order to excavate ghosts, was done to see what was unseen in the previous studies.

The Spectral Turn

The use of the spectral as conceptual metaphors has been part of a ‘spectral turn’ of the fields of sociology, and more recently criminology (Del Pilar Blanco & Peeren, 2013; Fiddler et al., 2022). This changing shift of focus is often seen as a reaction, and perhaps remedy, to ‘our alleged age of amnesia’ (Del Pilar Blanco & Peeren, 2013: 16), giving renewed focus to how the past lingers in the present, and makes demands on our future. This turn towards spectrality is often traced back to Derrida’s notion of hauntology in Spectres of Marx (1994), where the concept of the spectre becomes symbolic of the ‘ultimate disjointedness of ontology, history, inheritance, materiality and ideology’ (Del Pilar Blanco & Peeren, 2013: 7). As such, hauntology speaks of the ephemerality of our present, as it relates to our past, and the future this prompts. In the case of ‘vulnerable EU migrants’, this is evident in the past, but also in ongoing racism against Roma people that is prevalent in current policy practices in Sweden. As Fiddler et al (2022) points out, this may also refer to instances of transgenerational and colonial trauma. As haunting has been described as a ‘generalised social phenomenon’ (Gordon, 2008: 7), we would argue that all of late modernity and its encompassed texts essentially are haunted, containing hidden traces of the no longer and not yet.

The concept of hauntology and the notion of the ghost are also intrinsically related to the uncanny and the abject. Not only in the sense that the figure of the ghost itself is uncanny, as someone (or something) familiar which returns in an unfamiliar, re-animated form, but also in the sense that the dissolution of distinctions, the disruption and dislocation of time and space, rendering something familiar unfamiliar, evokes the uncanny. As something that disarranges categories and disorders categorisation, the uncanny is contagious; it unfamiliarises the familiar with contact (Fiddler, 2013). While the uncanny was first conceived in relation to individual psychopathology, the use of this concept has been extended to include effects and experiences on the societal level, including cultural experiences and the affliction of late modernity itself (cf. Fiddler, 2013; Fiddler et al., 2022; Linneman & Turner, 2022). The abject, which Kristeva (1982: 4) defines as that which ‘disturbs identity, system, order. What does not respect borders, positions, rules. The in-between, the ambiguous, the composite’, is in its very essence uncanny. As a function in which humanity expels the impurity of life, the abject must be ejected and removed in order to maintain borders and protect identity.

The spectral and hauntology as a methodology are also connected in the theoretical concept of Necropolitics (Mbembe, 2003a, b). Drawing on Foucault’s concept of biopower, Mbembe (2003b: 1) explores spectrality in political contexts in non-western, colonialised countries, which establishes; ‘extreme forms of human life, death worlds, forms of social existence in which vast populations are subjected to conditions of life that confer upon them the status of living dead (ghosts)’. Mbembe’s (2003a) notion of necropolitics implies the presence of political violence towards a particular group through constriction: the target group is deprived of the opportunity or freedom to improve their hazardous or miserable conditions.

In line with Derrida’s (1994) notion of hauntology, we plan to study the spectral traces of the past in the present, along with what demands this past makes on our futures. While conceptual metaphors of spectres and hauntings may appear as mere linguistic exercises, these motifs offer important theoretical tools. As Kindynis (2018: 39) states, ‘such motifs furnish us with the theoretical language necessary to explicate how memory and trauma become inscribed literally, symbolically, affectively and atmospherically in space and place’. The metaphorical concept of the ghost is already symptomatic of a fluidity of time relating to the ideas of hauntology, as a figure who returns, repeats and is resurrected from a past to make an impact on the future to come. The ghosts may then be conceptualised as those who have been rendered invisible, shadowed or silenced ‘in a present haunted by both the past and the future’ (Fiddler et al., 2022: 5). Trapped in the ‘looming shadows of the “no longer” and “not yet”’ (Fiddler et al., 2022: 5), their time is out of joint. In this spectral turn of the field, an increasing number of scholars have begun to explore haunting as a social phenomenon. As such, using a critical hauntological lens provides a unique perspective since it engages with a disease of temporality, exploring issues and phenomena which hover between absence and presence and where linear time is disrupted (Fiddler et al., 2024). In contrast with other perspectives exploring the othering or displacement of groups of people, the use of a critical, hauntological perspective therefore enables us to temporally (de)contextualise this othering, exploring its temporal disjointedness; tracing the past inheritance as well as the future indebtedness in current policies and research findings. Using ghosts as conceptual metaphors also enables innovative and subversive methodologies to explore different forms of violence and harm as this temporal disjointedness provides an opening of meaning, allowing us to ‘see what we feel haunting us’ (Fiddler et al., 2024: 2, emphasis in original). As we draw on Derrida’s understanding of justice, this fluid temporality, unique to hauntological perspective, where different social phenomena are understood through a coalescing of past, present and future, is central to ensure the achievement of justice (Fiddler et al., 2024). We argue that this approach is suitable in the study of ‘vulnerable EU migrants’ since they are examples of subjects being spectralised, having been marginalised and disavowed by different forms of dispossession and exploitation but are yet persistently present in their absence, silence or invisibility, what Derrida (1994: 157) refers to as ‘invisible visibility’.

(Re)Excavating Ghosts in Texts Using a Hauntological Lens—A Methodological Approach

A kind of ethnography of absence, an anthropology of people, and places and things that have been removed, deleted and abandoned to the flows of time and space. (Armstrong, 2010: 243)

The conceptualisation of social ghosts and analyses of the processes of spectralisation have been used to describe marginalised groups who must ‘learn to live as ghosts, as apparitions, as spectres: on the edge of social life, but never quite visible within it’ (Ferrell, 2015, cited in Kindynis, 2018: 30). Relating to ideas of social death (Cacho, 2012), a condition ‘lacking hope and opportunities for agency where they are not treated as fully human or grievable’ (Elsrud, 2020: 500) and subject to what Ferrell (2015, 2016) terms ‘ghost ethnography’, people like refugees and asylum seekers, who are ‘denied a social or political identity’ and ‘are perceived as less than fully human’ (Kindynis, 2018: 30), may be studied through the lens of ‘excavating absence’ (Farrell, 2015). By focusing on the excluded, on the non-present and not-quite-there, haunting the margins of our everyday lives, ghost ethnography endeavours to unearth a ‘politics of absence’, ‘trac[ing] the ghosts of exclusion, the women or immigrants or homeless folks never allowed in’ (Ferrell, 2016: 227).

As such, the discussion of the Roma through the spectralisation of ‘vulnerable EU citizens’ is intrinsically related to issues of space and place. As space is constantly being re-written and re-constructed, we must look to the interstitial, the in-between and the marginal, as well as the ‘disjointed, uncanny, other, no-longer or not-quite-there absent presences that “haunt” the margins of everyday life’ (Kindynis, 2018: 29, emphasis in original). The spatial is also almost intrinsically linked to marginalised people who are not only forced-out of and forced-in to certain spaces by the enactment of state power, but who must constantly navigate and negotiate the risks of space and being out of place in their daily lives. Following Fiddler (2019), who developed a spatial hauntology by which a space’s violent histories may leave spectral traces, haunting city spaces and everyday life, we believe that the ways in which the Swedish state has handled ‘vulnerable EU citizens’ is also haunted by the violent treatment of Roma in the past. Similarly, we employ this different way of seeing when (re)examining our own previously gathered material and studies, exploring what is there that we initially did not see, what is not there, what may have been there but is there no longer and what is not yet there (cf. Fisher, 2002). As such, this article also aims to illuminate how spectres and spectralising forces may appear, disappear and re-appear in the research process by interrogating previously unseen understandings of our texts; reanimating them and rendering them visible. This process, which draws on Derrida’s (1994) notion of ‘speaking with ghosts’, does not only excavate ghosts that were previously unnoticed, present in their absence, but also explores the ethical, social and political critique these ghosts are enabling. By looking for spectres in our own previous material, we therefore turn this gaze inwards, critically interrogating how our own previous work may have perpetuated a spectralising power that enabled social injustice, and prompting a change.

We employ the spectre as a conceptual metaphor in reconsidering our relationship to different aspects of ‘otherness’. Here, the spectre represents the return of the repressed, haunting the living and urgently calling for attention or justice. Spaces of absence, such as the abandoned informal settlements mentioned by our respondents, may also become spaces of social death; ‘the killing ground on which citizenship and social visibility are exterminated’ (Ferrell, 2018: 191). By excavating unseen narratives of ‘vulnerable EU citizens’ in already ‘seen’ texts, teasing out hidden ‘ghost texts’ (Armstrong, 2010), this study takes inspiration from ghost ethnography. This sensitive way of ‘seeing’ has proved useful when excavating hidden texts-within-texts, as well as studying how state institutions produce the vulnerable and the marginalised. Barker (2017) argues that it is evident in Sweden how control of the mobile poor is often driven by the needs and demands of the welfare state itself rather than the people in need. As such, by focusing on the redacted, on the ‘spaces and bodies lost to the occulting power of the violence of the state’ (McClanahan & Linneman, 2018: 520), ghost ethnography invokes a more sensitive and immersive kind of ‘seeing’; one of seeing beyond the readily visible to the in-visible and un-seen.

By using these hauntological, theoretical motifs, we focus on exploring how people and institutions in positions of power, including policy makers, policy discourses, and media reports, spectralise so called ‘vulnerable EU citizens’. We are using the spectral as a conceptual metaphor to provide ethical critique of the social injustice created by these social welfare institutions. It is therefore important to note that we are not exploring the spectrality of ‘vulnerable EU citizens’ from the EU migrants’ own perspectives, but exploring how transformative discourses from positions of power construct narratives around EU migrants as social ghosts and the effects this might have. As such, the EU migrants themselves might have completely different experiences of these issues.

The re-reading of our studies based on newspaper articles and media reports (Giritli Nygren, 2016; Giritli Nygren et al., 2019), inspired by ghost ethnography, revealed discursive links between welfare, social security and migration using the public debate around EU migrant camps. When drawing on ghost ethnography, we also did discursive readings of articles on ‘vulnerable EU citizens’ and/or EU migrants (often used interchangeably, although they are very different) from a local newspaper in the county, exploring how different materials spectralise so called ‘vulnerable EU citizens’. When re-reading the local media publications regarding the public debate around this group of Roma, we focused on the findings of our previous publications exploring how the links between welfare, social security and migration are constituted and produced, but the re-reading of these texts also revealed that these findings in themselves became part of reproducing these notions. While in our previous interview studies, focusing on how officials and politicians normalise certain beliefs regarding the (im)possibility of providing support to ‘vulnerable EU citizens’ in their municipalities (Giritli Nygren & Nyhlén, 2017), we found the social exclusion of the inhabitants of the informal settlement problematic, the re-reading of these studies also revealed that how the respondents identified the informal settlements themselves was problematic as well. Using ghost ethnography, we excavated the hidden text within this text showing this as actual violence against Roma. Similarly, while our previously published text exploring the project’s entire ethnographic material (Nyhlén, 2020) demonstrated the role of benevolent violence (Barker 2017) played in relation to project work and social entitlement, the ghostly re-reading of this article also revealed how benevolent violence, for the Roma, entails ‘protecting them from their own livelihood’ (Barker, 2017: 121, see also Nyhlén, 2020) which enabled us to provide ethical critique of the social injustice created by the Swedish State.

In the process of excavating hidden texts-within-texts and re-examining our previous material, three main narratives concerning ‘vulnerable EU citizens’ emerged, seeing the spectralisation; rendering the known unknown; out of sight and site; and haunting the margins. In the following sections, we will explore these three themes in turn.

Rendering the Known Unknown

The late-modern uncanny figure is characterised by their rootlessness. They are the vagabond re-imagined for a period of ontological insecurity. (Fiddler, 2013: 284)

When we re-examined how the participants working with the so called ‘vulnerable EU citizens’ described their work, it became apparent that they were using the public discourse term of ‘EU migrants’ rather than the official term ‘vulnerable EU citizens’. Furthermore, the terminology was an important topic for discussion; to discern what people we were actually talking about. The participants compared ‘EU migrants’ with categories like refugees, tourists and guest workers to prove the impossibility of understanding the group through any of the existing and available categories. A reference that was not mentioned by the participants, however, was to the Roma people as a European minority. As such, both the terms ‘vulnerable EU citizen’ and ‘EU migrant’ serve to render the Roma population invisible, de-contextualising this particular group from a colonial history of repression in Europe. The label of the public discourse, ‘EU migrants’, is also a problematic one, being a categorisation that transforms citizens into migrants, but without the associated rights, such as the opportunity to seek asylum (Giritli Nygren & Nyhlén, 2017). Consequently, through the spectralising power of ‘vulnerable EU citizen’, EU migrants can never achieve refugee status. Furthermore, the participants also juxtaposed ‘vulnerable EU citizens’ with Swedish citizens when interpreting legislative regulations regarding rights, which reinforced stigmatisation as well as differences between these categories:

If the problem would grow bigger [if more EU migrants would come to the municipality], and we would need to take further measures like providing shelter, at least my intentions are not to create some sort of side-track for EU citizens; then it [the shelter] would be for everyone who’s going through a rough time. [Interview with civil servant]

These comparisons construct and cement the spectre of ‘vulnerable EU citizens’ as an uncanny category, as something that can both shift in meaning as well as hide haunting traces of colonial or state violence. Taking this perspective, what we previously read as ‘searching for order’ (Giritli Nygren & Nyhlén, 2017) can instead be interpreted as an act of blurring, through which the uncannily deconstructed distinction between ‘vulnerable EU citizens’ and other known categories upholds and reinforces differentiation and stigmatisation. Governmental practices, in most states, have contributed to this considerable blurring of the boundaries between the categories of ‘citizens’, ‘migrants’ and ‘refugees’, in which eviction, forced mobility and differential inclusion are used to strongly stimulate ‘voluntary’ return without enacting more costly deportation procedures (cf. Baar, 2017). The comparison between ‘vulnerable EU citizens’ and other vulnerable groups further works to render the EU migrants, and the Roma population within this group, invisible in their visibility (Derrida, 1994) as their perceived vulnerability is downplayed and positioned as equal to or even below Swedish citizens ‘going through a rough time’.

What to call them and how they should be understood was also an area of media contestation where views differ distinctly based on ideological divides. Whatever the ideological position taken in the texts, the point of departure often focused on the fact that these people have very few established rights according to European and Swedish legislation and guidelines. Where such rights are established, they are often described as being vague and sometimes contradictory. It is obvious that how they are labelled is also an instrument to distinguish them from other groups and categories, although there are also examples where these distinctions are questioned:

It matters not what we call them; beggars, Roma, EU migrants or Romanians. They are people who have made their way to Sweden and Sundsvall in order to scrape together money on which they and their families can survive… These people are citizens of the EU and are covered by the right of free movement. When here, they are also subject to our laws and regulations. Yet they still fall between stools. Nobody wishes to take responsibility. (ST 12.02.2015)

Depending on how the categorisation or construction of the group was carried out, the officials and politicians we interviewed often sought support from other categories to legitimise their attitude towards the marginalised group. In situations where ‘vulnerable EU citizens’ vulnerability was foregrounded, reference was often made to dealings with other vulnerable groups, such as homeless Swedish citizens or asylum seekers. In contrast, when the emphasis was on the municipality’s areas of responsibility and ‘vulnerability’, categories such as ‘tourists’ and ‘guest workers’ were referenced. This indicates that when categorisation is in focus, the issue of vulnerability gets pushed to the background, and rights associated with citizenship become more prominent.

For instance, our findings demonstrate that in the everyday lives of officials and politicians, national and Swedish citizenship take precedence; these are continuously normalised and contrasted with the limitations imposed on those of European citizenship. This indicates that Swedish citizenship forms the basis for the invocation of any human rights, rather than European citizenship. As part of this ‘othering’, certain behaviours associated with specific groups become constructed as abnormal when associated with ‘vulnerable EU citizens’, as in the quote below, when a civil servant makes a joke about the needs of ‘vulnerable EU citizens’ by comparing them to ‘normal’ Europeans:

[…] I was kidding about this the other day with the national coordinator for vulnerable EU migrants, I said what if a group of Danish people would come, the Danish people needs shelter, the Danish people have set up an informal camp, what should we do with the Danish people? (Interview with civil servant)

This discourse constructs ‘vulnerable EU citizens’ as ‘less’ than Swedish people, and even less than ‘normal’ Europeans. By engaging in such narratives, ‘vulnerable EU citizens’ are not only othered and rendered inferior, but this construction also serves to feed into a narrative where these people are seen as ‘less than fully human’ (Kindynis, 2018: 30). By blurring the boundaries between known categories, ‘vulnerable EU citizens’ are decontextualised. As in the quote above, where the ethnic identity of Danish people is mentioned, the ethnic identity of Roma people is turned into an abstract ‘vulnerable EU citizen’. They are positioned as not only out of place but as out of time, removed from the legacy of the displaced realities that many of them face on a daily basis. By constructing ‘vulnerable EU citizens’ as uncanny, disarranging categories and disordering categorisation, they are not only made unknown, strange and different but are also ‘concealed and kept out of sight’ (Freud, 1919: 224–225, from Kindynis, 2018); invisible in their visibility, present but not seen.

Out of Sight and Site

Constructed as threats to social order and public decency, they haunt the lives of the more secure, moving through the margins to seek shelter or beg for money, leaving vague reminders of unauthorised presence in alleyways, on walls, along migrant trails. (Ferrell, 2018: 181)

One of the most pressing issues regarding the people who fall under the spectre of ‘vulnerable EU citizens’ in Sweden relates to their positioning in space and place, most notably their housing or living situation in informal settlements or camps. Public debates around the informal settlements have raised several issues, such as the question of unauthorised camping, homelessness and the responsibility of Swedish authorities to assist ‘vulnerable EU citizens’. Upon re-reading the material (Nyhlén, 2020), utilising the spectral as a theoretical motif, these debates take on a different light. The informal settlements inhabited by this group of people have been framed in Sweden as an ‘acute crisis’ and something that needs to ‘be addressed’ (Nyhlén, 2020); this framing not only enables evictions but constructs these evictions as the responsible thing to do. As exampled by the quote below, help is even framed as something that increases vulnerability;

We may risk ending up in a “curling-situation” where we are giving help that is pacifying the people in need. Yes, and the services we give create more migrants. (Interview with volunteer worker)

The removal of ‘vulnerable EU citizens’ from sight and site, while simultaneously narrating these evictions as the morally correct action, resonates with the spectral. Forcing these individuals to move-out-and-in of certain places, while simultaneously justifying these acts of state violence, ‘storying’ these acts as ‘righteous’ (Higgins and Swartz, 2018) highlights the ephemeral position of the ‘vulnerable EU citizens’. Similarly, both the participants in the project as well as media narratives describe the informal settlements as ‘unacceptable’, ‘harmful’ and ‘bad’ (Nyhlén, 2020: 13, 17). Through the spectralising power of the term ‘vulnerable EU citizens’, their very presence in public spaces was constructed as the main problem (particularly when connected to begging), solved only by eviction and removal. This spectralising power works to construct these people in abject terms; they must be banished to protect society’s values and maintain its boundaries. By constructing the eviction of people from informal camps as morally right, saving them from their vulnerability, moral consequences are avoided; ‘monstrous punishments are the only logical responses to monstrosity’ (Higgins and Swartz, 2018: 92).

By analysing the closing of the settlements through a perspective framed by the spectral that banishes a certain way of life (cf. Lind & Persdotter, 2017), the evictions of marginalised groups may be understood as strategies for maintaining border defence. The ‘unacceptable’ housing situation is discussed in terms of risk, not only for the ‘vulnerable EU citizens’ themselves but for the local community and public order (Nyhlén, 2020). In this way, state violence is constructed as a justified response to the threatened deterioration of the nation. Here, the ‘other’, in the shape of the ‘vulnerable EU citizen’, is constructed not only as threatening and abject, but as inhuman. They are framed as individuals who frighten because they ‘disrespect borders and thus threaten individual, collective or national identity’ (Fredriksson, 2019: 264), and they are made monstrous, ‘threatening social homogeneity by rendering society insecure’ (Neocleous, 2005: 134). As such, they are rendered into what Mbembe call ‘wandering subjects’; trapped in ghostly impermanence, made to move on, forced to ‘escape from himself each time and allow himself to be carried away by flux and accidents’ (Mbembe, 2003a, b: 23). The spectre of ‘vulnerable EU citizens’, renders them into ghostly figures, denied a social or political identity, as spectres without any rights.

According to our previous studies (Giritli Nygren et al., 2019), the underlying discourse facilitating this shift in Sweden is the national welfare discourse; it is the node that defines both the 'normal' and the 'other'. Within the welfare discourse, beggars who have accepted help, in the sense that they are either doing something that could be considered work as a way to get off the streets or taking part in organised empowering initiatives, are held up as positive examples—as ‘good’ and ‘desirable’ individuals:

These people naturally require a sensible job and a roof over their heads. Perhaps they could be offered some form of cleaning or clearing. There is much that needs doing even in our area and any of such opportunities should be investigated. It cannot be sensible to sit cup in hand for eight hours a day. (ST 07.05.2015)

This construction of ‘vulnerable EU citizens’ in the public narratives seeks to justify their banishment from public space. Their very presence indicates an acceptance of ‘bad’ and ‘unacceptable’ ways of life, so they must be evicted and removed to achieve progress and restore normality (Nyhlén, 2020). Once these vulnerable populations are out of sight (and site), forced off the map by violent state sanctions, the problem is considered ‘solved’ (Nyhlén, 2020). Or as one of the civil servants frames it;

The project has been good for the municipality since the evictions have decreased the number of EU migrants ...I think there are fewer who have come since the camp was demolished and it is considerably fewer that are here. Therefore, the issue is resolved now (Office Manager for Municipal Administration).

However, as we discovered in our re-analysis of our material, the eviction of so called vulnerable EU citizens leaves spectral traces, hiding that this is yet another example of evicting and displacing Roma people, forcing them into the margins.

Haunting the Margins

[They] learn to live as ghosts, as apparitions, as spectres: on the edge of social life, but never quite visible within it. (Ferrell, 2015)

When re-reading our studies and material using the ghost as a conceptual metaphor, spectral narratives pertaining to the ‘disgraceful’ living situations of so called vulnerable EU citizens, highlighting their ephemeral and impermanent societal position, become apparent. The visibility of poverty is constructed by narratives in the local media as something ‘other’, strange and uncanny, not to be associated with Sweden or the Swedish welfare state (Giritli Nygren, 2016). Increasing levels of visible poverty and begging are thus articulated as a sign of the ‘end’ of Swedish society.

In Sweden, we have long been spared from beggars on our streets. We are now seeing something which can be compared to an explosion. This is naturally uncomfortable and, not least, unexpected. Sweden has long been renowned as a model of solidarity and the welfare state. A country where we take care of one another; a country where we are unusually receptive to influences from the outside world… In a globalised world, Sweden must also be prepared for change. This need not, however, mean that we compromise on key principles such as compassion and solidarity with our fellow human beings. (ST 14.08.2014)

The quote above illustrates how the spectralising power of the term ‘vulnerable EU citizens’ renders the situation ‘new’; as something inherently un-Swedish, as if the Swedish State has not had a long history of racism against Roma people. These ghostly narratives, which frame the presence of ‘vulnerable EU citizens’ in disrupting, uncanny terms and as ‘less-than-fully human’ (Kindynis, 2018: 30), not only force them out of certain spaces through the enactment of state power but also deny them a social and political identity (cf. Ferrell, 2015). The ‘vulnerable EU citizens’ do not belong; they are not part of ‘Swedish’ society and they themselves are constructed as a risk of permanent poverty in this society (Giritli Nygren, 2016). Per the spectral lens, these citizens are constructed as socially dead; they are ghosts haunting the margins of social and public life, denied rights, displaced and dispossessed. Their invisibility even extends into the avisual (Lippit, 2005); there to be seen, yet only seen in absence; apparent, yet not apprehended. As Peeren (2014) argues, this type of (a)visuality reduces vulnerable individuals such as migrant workers or ‘vulnerable EU citizens’ to mere bodies; bodies which, as either subjected to labour or to various forms of violence, tend to be construed as abject (Wills, 2007).

This politico-cultural production of people who ‘never achieve, in the eyes of others, the status of the living’ (Holland, 2000: 15) serves to actively maintain state power (Linneman et al., 2014: 207). By declaring certain individuals as ‘socially dead’, state violence against these individuals is legitimised, thus reinforcing state power. This renders the ghostly narratives of ‘vulnerable EU citizens’ an important site for necropolitical power, a constriction that can be operationalised through political action but also through inaction. As Mbembe (2003a, b: 21) describes, people subjected to necropolitics are kept in a ‘state of injury’, a ‘phantom-like world of horrors and intense cruelty and profanity’ and reduced to shadows. Suffering therefore becomes a political technology, where certain groups are exposed to conditions in which they are ‘kept alive but in a state of injury’ (Mbembe, 2003a, b: 21). Framing the Roma population as ‘vulnerable EU citizens’ is also a strategy to continuously keep an ethnic minority in a perpetual state of injury.

Within this spectral, necropolitical system of domination, haunted by the ghost of colonialism, conditions that are ‘obscene, vulgar and grotesque’ (Mbembe, 1992: 1) are sanctioned for political ends. The closing of temporary camps and the eviction of their residents may be regarded not only as (post)colonial practices, but as acts of necropolitical state violence. Similarly, exorcising the ‘social ghosts’ from their ‘disgraceful’ living situations and forcing avisuality upon them may also be regarded as such. Upon their eviction from the examined camps on the first of October 2016, the ‘vulnerable EU citizens’ in question had to relocate elsewhere; as long as they remained hidden and confined to society’s margins, the problem was labelled as solved and no further support interventions were taken (Giritli Nygren, 2016; Nyhlén, 2020). This forced this group to be complicit in their own exclusion; they had to enforce their own invisibility to survive their social death as ghosts (cf. Ferrell, 2018). As Ferrell (2018: 186–197) states, ‘all involved conspire in the construction of ghosts, in not seeing those who are not to be seen’. They are forced to haunt the margins of society and everyday life, invisible and excluded, risking eviction and displacement if their spectral presence becomes too apparent.

No matter how much one would like to help those in need, it is impossible to ignore a development that sees people living and sleeping wherever they like… property rights must also be protected. (ST 03.07.2015)

We have demonstrated that the spectral narratives around ‘vulnerable EU citizens’ evident in Sweden’s current welfare institutions and policy discourses are sites of necropolitical power that enable violent state sanctions and social injustice, for instance in the name of property rights. Necropolitics as it is manifest in the Swedish deportation regime, i.e. closing borders, hindering adults and children to seek refuge, denying access to healthcare and schooling and closing spontaneous camps, amounts not only to the violation of human rights, but more broadly, points to a general consensus among politicians and the public that some human lives are worth less than others. The colonial occupation of land when establishing state sovereignty and the implementation of the liberal idea of citizenship rights involve the production of cultural imaginaries (Mbembe, 2003a, b: 25). Such imaginaries ‘gave meaning to the enactment of differential rights to differing categories of people for different purposes within the same space’ (2003: 25).

Furthermore, we have showed that the construction of ‘vulnerable EU citizens’ spectralises Roma discursively and through actions taken by people in power and is the very mechanism that enables this necropolitical power. By exploring how dominant discourses and people in power construct the spectre of ‘vulnerable EU citizens’ using a hauntological perspective, we have been able to draw attention to this process of ghostification of ‘Roma people’. This should, following Membe, not be understood as a new phenomenon or contra liberal western values, but as a logical contemporary expression of historically embedded colonial/modern, racially hierarchical worldviews which have their roots in the colonial inventiveness inherent in the idea of liberal state sovereignty. In the context of colonialism ‘sovereignty means the capacity to define who matters and who does not, who is disposable and who is not’ (2001: 27), and ultimately, necropower works towards ‘the creation of death-worlds, new and unique forms of social existence in which vast populations are subjected to conditions of life conferring upon them the status of living dead’ (2001: 40).

These spectral narratives that enable and justify necropolitical state violence can thus be read as ‘modern methods of racism’ (Higgins and Swartz, 2018: 92) that regulate the distribution of death and enable the murderous functions of the state. As such, it is important to identify and prevent such narratives in the context of the Swedish welfare discourse. Not only do such narratives pose risks of further discrimination and abuse against an already vulnerable group, but they also constitute a serious threat to any possibility of social justice.

Ghostification: Exorcising the Spectres

In the aftermath of enforced absence, in the afterlife of social death, ghosts drift away to circulate along the shadow margins of social life. (Ferrell, 2018: 194)

This study has employed the ghost as a conceptual tool to explore the narratives surrounding ‘vulnerable EU citizens’ within Sweden’s current welfare institutions and policy discourses. By excavating examples from previously gathered material and already published texts, we have explored the unseen within the already seen to critically examine how and with what ethical implications ‘vulnerable EU citizens’ are constructed in social welfare narratives. We have explored the role of the spectral and hauntological in defining our relationships to different aspects of ‘otherness’ and in representing the return of the repressed to demand attention or justice. We argue that ‘vulnerable EU citizens’ not only are constructed as uncanny and abject but also as social ghosts, denied a social and political identity and forced to haunt the margins of societal life.

Unlike other perspectives exploring the vulnerability of marginalised groups, the use of the hauntological and the spectral as conceptual metaphors has enabled a new, more sensitive way of ‘seeing’ which has generated a deeper, more critical understanding of how welfare institutions in Sweden legitimise state violence against these groups, as well as how our own previous research has been a part of this spectralisation by the use of concepts such as ‘vulnerable EU migrants’. In this way, the use of this methodology and theoretical lens have enabled us to explore how ghosts appear, disappear and re-appear in the research process. In addition, as the dominant discourses by the people and institutions in power are analysed through the lens of the spectral, we have not only been able to trace the ghosts of (post)colonial practices, but also revealed narratives which render these marginalised groups into voiceless spectres; social ghosts that are made uncanny and abject, invisible in their visibility, and even framed in inhuman terms, as bodies without any inscribed rights. As this process of ‘ghostification’ of ‘vulnerable EU citizens’ is the very mechanism that enables the use of necropolitical power by these welfare institutions, the use of the spectral as a theoretical tool has thus enabled us to examine new dimensions of violent state practices that has previously remained unseen. As such, the spectral narratives surrounding ‘vulnerable EU citizens’ excavated in the current study emphasise the need for a certain sensitivity or ‘sight’ when analysing marginalised and ‘othered’ groups. These narratives also emphasise a dis-ease of temporality fundamental to understanding the treatment and experience of the people encompassed by the term ‘vulnerable EU citizens’ in Sweden, unique to a critical, hauntological perspective. In light of rising punitive populism, we predict that evictions and camp closures, the violent criminalisation of begging and the closure of shelters are all enabled and legitimised by the spectre of ‘vulnerable EU citizens’. As such, it is this positioning, this process of ghostification of their humanity, drawn from a past colonial inheritance, which enables their exorcism, leading to haunting futures.

When conceptualising individuals as social ghosts, it is furthermore important to explore the nature of the social death that led to this construction (Ferrell, 2018). As Ferrell (2018) argues, a social system based on economic inequality and social exclusion ‘systematically withdraws from its marginalised members the lifeblood of citizenship’ (p. 185). In this study, we also identify Sweden’s national welfare discourse as the node that facilitates the spectral narratives surrounding ‘vulnerable EU citizens’, constructing them as social ghosts. The necropolitical sanctions forced upon these groups are justified on the basis of morality and welfare and are claimed to protect ‘vulnerable EU citizens’ from ‘disgraceful’ and ‘unacceptable’ living conditions (Giritli Nygren, 2016; Nyhlén, 2020) that are constructed as distinctly ‘un-Swedish’ in nature.

By shifting the perspective to the spectrality of Swedish welfare narratives, we have not only found that ‘vulnerable EU citizens’ have been constructed as socially dead ‘ghosts’ but also that a post-mortem of the causes of this social death reveals violent and harmful effects of the Swedish national welfare discourse. Moreover, this discourse is identified as a site for necropolitical power, enabled by the ghostification of so called ‘vulnerable EU citizens’. The lingering violent effects of this discourse are an important subject for future study, as well as how the effects can be mitigated. As this study has also explored the spectral in a Derridean sense, exploring the ghostly traces left behind, haunting our present, we must also explore what these hauntings demand of our future (Derrida, 1994). Our research indicates that these violent state sanctions not only force the already marginalised to haunt the margins of society, but they also force the haunting of welfare institutions themselves; the ‘social ghosts’ and the process of ghostification leave spectral traces of violence that urgently call for redemption and justice. Using a hauntological lens, we have seen the ghosts of the no longer; the lingering spectre of the colonial history rendered invisible, still felt in our presence. We have also perceived the spectres of the not yet; the possibility of the Roma population to claim their minority rights, as well as the looming ghost of the continued erasure of the Roma population off the social map, as if they were never here at all. While exorcising these spectres from sight and site may render so-called ‘vulnerable EU citizens’ invisible as well as unliving, the inherent violence of these actions casts ‘perpetual shadows in the present’ (Fiddler et al., 2022: 5), echoing the violence of social injustices of the past. As Del Pilar Blanco and Peeren (2013) argue, the use of the spectral metaphor on a structural and global scale, ‘spectropolitics’, may emerge as a site of resistance and potential change, where we can be prompted to (re)imagine a future radically different from the one charted out by ghosts of past and present. To live with ghosts, as Derrida (1994) suggested, rather than to exorcise them, encourages a (re)conceptualisation of the not yet as it relates to the no longer (Fisher, 2002), providing a deeper understanding of disjointedness and enabling ethical, social and political change. As our present is simultaneously haunted by the threat of future violence, it will be an important task for future researchers and policymakers not only to answer that call, but to restore the ‘lifeblood of citizenship’ (Ferrell, 2018: 185) to the individuals who have been rendered the social ghosts of our time.