The politics of patrolling ‘safety guards’ in Sweden: outsourcing, depoliticization, and immunization

ABSTRACT

The contracting of private guards to patrol public spaces has rapidly become a widespread approach to increase public safety and prevent crime in Swedish municipalities. Drawing on interviews and policy materials from three municipalities, this paper examines how private patrolling guards has become a solution to (un)safety in Sweden, and the political implications of this development. The governmentality analysis shows how the rendering of (un)safety as technical and governable depoliticizes safety by detaching it from its social and political connotations. At the same time, in the process of demarcating the space to be ‘guarded,’ safety is problematized as a matter of order and sameness in public space, which (re)produces racialized boundaries between those to be made safe and those considered threatening safety. The study further demonstrates how the outsourcing of responsibility for public safety to the security industry is bringing about a shift in democratic legitimacy, accountability, and the monopoly on the use of force, largely without political contestation in the studied municipalities. The paper concludes by discussing the underlying rationale of this practice of governing (un)safety as informed by a biopolitical logic of immunization: safeguarding and immunizing some, at the expense of those marked as risky ‘others’.

KEYWORDS:

• Public safety
• security guards
• patrolling
• private policing
• governmentality
• depoliticization

Introduction

During the last few decades, the responsibility for safety and crime prevention has increasingly moved toward the private security industry in general, and private guards in particular. In many parts of the world, private guards now outnumber the public police (Provost 2017). This new role of private security can be understood as a part of a broader development toward what Loader (2000) calls ‘plural policing,’ through which new actors, such as private security guards are increasingly involved in visible frontline policing and patrolling (Crawford et al. 2005; Wakefield 2012; Kammersgaard 2021).

In Sweden, municipalities have played a central role in what earlier research has described as a ‘silent’ marketization of policing at the local level (Hansen Löfstrand et al. 2019, 1). Governmental policy promoting the outsourcing of public policing has during this development been absent nationally. Yet over one-third of Sweden’s municipalities, including the three largest cities, are today contracting private guards to patrol public spaces and neighborhoods to increase safety and prevent crime (Hesserud Persson 2016).

Focusing on the new role of private security in public policing, Sweden forms a particularly interesting case to study. While earlier research points to similar developments of private guards patrolling and policing public or semi-public spaces in many parts of the world (Kammersgaard 2021; Saarikkomäki and Alvesalo-Kuusi 2019; Samara 2010; Crawford et al. 2005; Wakefield 2012), it is a new and rapidly growing phenomenon in Sweden. In contrast to most other contexts, the contracted guards in Sweden are patrolling public rather than private spaces and are explicitly assigned a social and crime-prevention mandate by the municipalities. Through the public-private contracting, they are closely connected to the welfare state. In their patrolling, the guards are assigned to work toward increasing public safety, to take a proactive approach and building relations with targeted groups. They work closely with, and report what they monitor to, both the municipality and the police (Brandén 2022).

At the same time, research into the public perceptions of guards in Finland suggests that they do not signal safety to everyone and that their work involves racialized risk profiling (Saarikkomäki and Alvesalo-Kuusi 2019). It also points toward a lower trust in guards compared to the police, and difficulties for citizens to understand the differences in authority between them (cf. Saarikkomäki 2018, 167). In Sweden, private guards only have a few weeks training and are situated in a limited organizational setting compared to the police (Sverne Arvill, Glantz, and Tonneman 2020). Furthermore, private security organizations are governed by business-oriented, rather than democratic, values (Hansen Löfstrand et al. 2019). The public sector in Sweden is for example directed by regulations to secure transparency and accountability that do not apply to private organizations, such as the principle of public access to information (Swedish Ministry of Justice 2020). What is at stake here are thus some of the democratic cornerstones of policing in Sweden, namely: matters of legitimacy, democratic accountability, and the traditionally state-centered monopoly on the use of force (cf. Loader 2000; Hansen Löfstrand et al. 2019). It therefore becomes pertinent to ask: How did the contracting of private guards become a solution to public safety in Swedish municipalities? What (un)safety problems are the patrolling guards imagined to resolve, and how? What are the political implications of this development?

To open up these questions for critical scrutiny, this paper examines the contracting of private patrolling guards as a new form of governing safety in Sweden. As such, this development is understood as part of broader neoliberal shifts in the organization and governing of the Swedish welfare state (Larsson, Letell, and Thörn 2012). These shifts include ongoing processes of marketization and outsourcing of the public sector (Allelin et al. 2021; Peters and Pierre 2018) which here results in a transfer of responsibility for safety toward the private security market. Drawing on a governmentality framework, the analysis starts with the understanding that for (un)safety to be governed, it first has to be delimited and established as a domain that it is possible to act upon by means of governmental intervention, here with the private patrolling guards as a response (Dean 2010; Li 2007). From this perspective, my aim in this paper is to analyze how the issue of (un)safety is given meaning as a political problem and being shaped into an object of government, and the political rationalities that inform this process. The analysis builds on a wide selection of policy materials and interviews with public officials, local politicians, and representatives from the local police and security companies involved in the organization of patrolling guards in three Swedish municipalities.

With Flyvbjerg (2006, 232), I conceptualize this case as paradigmatic, highlighting ‘more general characteristics’ of society. The study provides insights into how current discourses of public safety operate in relation to neoliberal forms of governing within the welfare state, as well as into the contemporary governance of public (un)safety through private forms of policing, which are relevant beyond the context of Sweden. In particular, it illustrates how safety in public space, and the problems articulated as related to unsafety, are rendered governable through the contracting of patrolling guards.

The article is structured into five sections. First, the governing context in Sweden, characterized by neoliberalization, a powerful safety discourse, and a changing role of private guards in policing, is outlined. The framework of governmentality is then introduced, and the main analytical concepts are presented. This is followed by a description of the case, methods and materials. The empirical analysis then shows; how (un)safety is made intelligible, calculable, and governable; how the space to be governed is being demarcated and produced; and how the outsourcing of safety to private security companies becomes a natural and uncontested response to unsafety. To conclude, the political rationalities underlying this form of governing safety are discussed, along with its political implications.

Public safety and private guards in the neoliberalized Swedish welfare state

While Sweden is continuously defined as a social democratic welfare state, offering universal welfare and equality (cf. Esping-Andersen 1990), inequalities and socioeconomic differences have been growing over the last few decades (Ahrne, Stöbler, and Thaning 2021). Previous research connects this development to significant changes in the organization and governing of the Swedish welfare state, characterized by cutbacks and privatizations, market-based logics and neoliberal forms of governance (Allelin et al. 2021; Larsson, Letell, and Thörn 2012). Through these changes, the distinction between the functions and responsibilities of public and private actors has become vague, not least at the municipal level (Montin and Granberg 2014; Öjehag-Pettersson and Granberg 2019).

Alongside this development, crime statistics show an increase in ‘gun homicide’, which is described by the Swedish National Council for Crime Prevention (Brå [The Swedish National Council for Crime Prevention] 2021) as ‘limited to a very specific group – young men involved in criminal milieux in socially disadvantaged areas’. This situation is frequently debated in media and public politics. As part of targeting what are described as ‘gang related crimes’ the Swedish Government (2021, 20) has now proposed new regulations for ‘a more flexible and increased use of order guards’, particularly for maintaining public order and increasing safety. The new law for order guards is planned to come into force on January 1, 2024.

In parallel with this process, a powerful discourse around crime and safety has developed in Sweden, bringing public safety to the top of the political agenda (Olsson and Rönnblom 2020). Through this development, the previously broader understanding of safety [in Swedish trygghet] as social and collective wellbeing, has become narrower and more closely linked to matters of personal security, crime, and immigration (Olsson and Rönnblom 2020; Sager and Mulinari 2018). As safety is regularly measured, ranked, and compared, to be labeled ‘safe’ has also become a competitive factor for municipalities (Brandén and Rönnblom 2019; Listerborn 2016). It is within this context that Swedish municipalities are increasingly ‘buying into safety’ (Hesserud Persson 2016) and spending more than ever on private security services to patrol and monitor public space (Perkiö 2017).

Policing has traditionally been the exclusive responsibility of the state in Sweden and the employment of private guards for policing purposes has consequently been restricted. Nevertheless, private guards are increasingly acknowledged as important actors for increasing public safety in national crime-prevention policy and described as a valuable complement to the police (Swedish Ministry of Justice 2017, 22; Swedish Government (2021, 20). Hansen Löfstrand et al. (2019, 1) has argued that the ‘regulatory void’ that has long characterized the role of private security in policing, has opened up opportunities for the private security industry to lobby for guards as a ‘pragmatic’ solution to municipalities’ increased demand for public safety. In this paper, I understand Swedish municipalities outsourcing of the responsibility for safety as connected to these broader neoliberal transformations of the welfare state and the rationalities of governing behind these transformations.

The guards in focus for this study are employed by private security companies that are contracted by municipalities to patrol public spaces. They fall into two distinct categories, both obligated to wear uniform and allowed to carry handcuffs and a baton, but with different regulations and authority to use force. Like all Swedish citizens Security guards [väktare] may use force in self-defense, and detain someone who is caught committing a crime, but not act on suspicion. They are authorized and controlled by county administrative boards. Order guards [ordningsvakter] have the authority to intervene, detain, and forcefully remove people. They are trained, authorized, controlled and appointed on a three-year basis by the Swedish police. To be re-appointed requires continued training (see Sverne Arvill, Glantz, and Tonneman 2020).

The authority to deploy patrolling order guards currently only applies within demarcated so-called ‘Paragraph 3’ areas, requiring a time-limited approval from the police authorities. This is based on the assessment that there is a specific need and that it is of significant benefit to the public (Sverne Arvill, Glantz, and Tonneman 2020, 43). The spaces being assessed as fulfilling the requirements of this exception to the law are increasing, both in number and size, and now often cover entire city centers (Letterfors 2018). Statistics from incident reports show that the number of arrests made by order guards, rather than the police, in Sweden increased from about 5000 in 2017 to nearly 11,000 in 2018 (Säkerhetsföretagen 2022).

Analyzing the contracting of patrolling guards as a practice of government

To examine how (un)safety is shaped and given meaning as an object of government, to be met with the contracting of patrolling guards as a response, I draw on a governmentality framework. The concept of governmentality refers to a shift in the focus of government from a more sovereign form of rule toward the administration of life (Foucault 2008). This involves the persistent striving of government to secure the well-being (and safety) of the population, being its most vital resource, through different forms of biopolitics (Dean 2010, 125; Li 2007, 270; Foucault 2008). Thus, governmentality also signifies a shift in the ‘how’ of government, operating in more calculating ways through the conduct of conduct. In other words: the attempt to regulate, shape, and change the behavior of a particular object, field, or population, for a variety of ends (Dean 2010, 18; Walters 2012, 11–12). A governmentality analysis focuses on the practices and forms of knowledge underbuilding the governing techniques involved in directing our behaviors, to make visible the often implicit understandings of what is being governed, by whom and why (Dean 2010, 24–28).

Within governmentality studies, governing is understood as a meaning-making process. Through the employment of different governmental techniques, shaping and delimiting the domain to be governed, rendering it calculable and governable, a specific meaning of the problem being governed is being produced (Elden 2007; Lövbrand and Stripple 2011; Öjehag-Pettersson 2020). Governmentality studies focusing on safety and security have for example demonstrated how a range of social and political problems are increasingly governed and constituted as potential risks, which produces a state of anxiety and exception. This in turn legitimizes actions taken to protect the public from these risks, while largely concealing the effects for those being targeted as risks (see, for example, Bigo 2002; Walters 2004; Huysmans 2006).

Drawing on the governmentality literature, I conceptualize the contracting of private patrolling guards as a practice of government (Dean 2010). As such, it forms part of what Dean (2010, 40) calls ‘a regime of practices’ in the government of public safety in Sweden, which relies on certain ideas about what it is that should be governed, through what means, and to what ends. These ideas are in turn understood to be discursively produced by, and part of (re)producing certain truths around (un)safety that shape, limit, and inform ‘the thought’ in practices of government. The governmentality framework provides the analytical tools for analyzing how (un)safety is made thinkable, calculable, actionable, and governable through these established knowledges and truths.

Through Li (2007, 7) I understand government as a problematizing activity, one that involves ‘identifying deficiencies that need to be rectified’. It also involves the process that Li (2007) terms ‘rendering technical’, including a range of practices that contribute to demarcating, defining, and producing the problem or space to be governed, rendering it intelligible, calculable, and amenable to government intervention (cf. Rose 1999).

Accordingly, my analysis will focus on how the domain of (un)safety is being problematized and shaped into an object of government, to be ‘solved’ through the contracting of patrolling guards. I will also make visible how this practice of government is intertwined with certain rationalities, forming the underlying logics that need to be in place for safety to be demarcated and problematized in a particular way (cf. Öjehag-Pettersson 2020, 624). Importantly, the process of governing also involves determining the boundaries between those ‘with the capacity to diagnose deficiencies in others, and those who are subject to expert direction […] a boundary that has to be maintained and that can be challenged’ (Li 2007, 7). As Walters (2004, 249) argues in relation to immigration policy in the UK, dividing practices are core elements of ‘modern governmentality’. In the analysis, I particularly attend to the boundary being drawn between those to be made safe and those considered causing unsafety to others.

I consider the dominating rationale behind maintaining this boundary as operating through ‘a biopolitical frame of immunization: a form of governing that secures the life of some while sacrificing others’ (Swyngedouw and Ernstson 2018, 5; see also Esposito 2008; Lorey 2015). The immunity rationale builds on the notion of a biopolitical split (Neyrat 2010, 110), leaving on the one side the exposed, those constituted as deviant ‘others’, commonly seen as risks. On the other side of this split are the immunized, those included in the safe ‘we,’ being protected from risk through the patrolling guards. As a political rationality, the idea of immunization builds on a promise of managing (un)safety in a manner that permits life to go on untouched for those included in the immunization, while increasing the exposure for those being excluded (Swyngedouw and Ernstson 2018, 17).

Methods, materials, and study design

This study is based on a collective case study design (Stake 2000), including interviews and policy materials from three Swedish municipalities. These municipalities all contract patrolling guards for increasing safety, but vary in geographical location, size, political majority, and socio-economic structure. While Municipalities X and Y contract security guards [väktare], lacking the authority to use force, Municipality Z contracts order guards [ordningsvakter], with an enhanced mandate to use force to detain or forcefully eject people. The sample strategy has been to gather a broad selection of materials from different contexts, with the aim of comprehending primarily commonalities, but also nuances and potential contrasts of the phenomenon of interest (Stake 2000).

In Municipality X, sparsely populated and governed by a right-wing majority, the patrolling security guards [väktare] were initiated through a project in 2017, as a ‘visible’ and ‘safety promoting’ measure, with the goal to reduce unsafety among municipal residents (Project description 2017: 5, Municipality X). The project was evaluated as successful and today the municipality plans to contract a private security company on a more long-term basis.

In Municipality Y, a suburb of one of Sweden’s largest cities, patrolling security guards [väktare] have formed part of the municipality’s local safety work for the past decade. The municipality is large, densely populated, and governed by a left-wing political majority. Several areas have been identified by the police as ‘especially vulnerable’¹ (Swedish Police Authority 2017, 10). The guards were assigned by the municipality to work with ‘crime-prevention’, and ‘establish local contacts with youth’ by visiting ‘the municipaly’s schools and youth centers’, to build relations and ‘play down the presence of uniformed staff’ (Public procurement 2017: 3, Municipality Y).

The urban Municipality Z is governed by a left-wing political majority. In 2017, order guards [ordningsvakter], with the authority to use force, were approved to patrol large parts of the municipality’s city center and one area categorized as ‘especially vulnerable’ as part of a temporary project. The patrolling order guards were assigned to ‘build relational dialogue’ and ‘complement the police’ in maintaining public order, through their authority to detain, and forcefully remove people (Project evaluation 2018, Municipality Z). Today, the municipality is seeking to make the patrolling order guards permanent, operating in the whole municipality.

As summarized in Table 1, I have interviewed 23 persons (IP, interview person), involved in the organization of the patrolling guards in the municipalities. The larger number of interviewees in Municipality Z is due to the more comprehensive organization around the patrolling order guards in that municipality.

Table 1. Overview of interviews.

	Municipal civil servants	Local politicians (majority and opposition)	Police officers	Private security managers/CEOs	Total
Municipality X	2	2	1	1	6
Municipality Y	2	2	2	1	7
Municipality Z	4	4	1	1	10
Total	8	8	4	3	23

The semi-structured interviews lasted between 45 min and 2 h and were conducted during 2018 and 2019. A broad selection of policy materials has also served as an important source for understanding the process of governing safety through the patrolling guards involving: project descriptions, evaluations, reports, political decisions, applications for areas to be patrolled by order guards, and public procurement documents. For ethical reasons, any identifiable information from these materials has been excluded. The analysis of both interviews and policy materials focuses on identifying descriptions and underlying understandings of (un)safety and the problems related to it, and how the patrolling guards are being motivated and problematized as a solution to these problems.

Governmentalities of safety in the turn toward private patrolling guards

The empirical analysis is carried out in three steps. I begin by analyzing the shaping of (un)safety into a measurable and calculable object. Then, I turn the analysis toward the demarcation of a space to be ‘guarded’. Third, I connect the rendering of safety as technical to its outsourcing to private security companies as a ‘solution’. Throughout the analysis, I connect this practice of governing safety as intertwined with an underlying rationale of immunization: aimed at protecting some, at the expense of risky ‘others.’ In the analysis, I highlight the contextual differences between the municipalities and the different type of guard being employed, but primarily focus on the dominant features of this form of governing safety and the similarities between the municipalities in this aspect.

Rendering safety technical: (un)safety as measurable and calculable

In Sweden, unsafety is regularly measured through citizen-based surveys asking about experiences of unsafety. Based on these statistics, all three municipalities described a similar problem of either high or rapidly increasing levels of (un)safety, rather than violent crime, as a central motive behind the contracting of private guards to patrol public spaces. As expressed in an evaluation report in Municipality Z, ‘signals of increased unsafety,’ particularly among young women in the city center, provided the motivation for demarcated areas to be patrolled by order guards, while there was no equivalent increase in crime.

The measured (un)safety was described in both policies and interviews as ‘hard to explain,’ as a complex, subjective feeling or experience, separate from risk. Nevertheless, unsafety statistics were frequently articulated as a form of evidence of the problem. As explained by the Mayor of Municipality Z, ‘it’s the feeling that’s important here […] there was an experience that something had happened and there was an increased unsafety […] and we had to deal with that’ (IP7Z). Similarly, politicians, police representatives, and civil servants, primarily in Municipalities X and Z, frequently pointed out that, while they were uncertain about the reasons behind the increased unsafety and did not see it as directly related to risk or crime, it was crucial to take these feelings of unsafety seriously and act upon them.

The measured unsafety itself is, in this manner, being articulated as the main problem to be dealt with through patrolling guards. My interpretation is that it thereby functions as a technology of government, shaping how the problem of (un)safety is thought about and made intelligible. Thinking with Li (2007), these (un)safety statistics also contribute in rendering the problem to be governed technical and enabling the authorities to rule according to the political rationalities at work. In the context of what was described as ‘a major focus’ on public safety in the municipalities, which also included a call for more police officers on the streets, the unsafety statistics become a point of departure for motivating the use of guards. As described by the Mayor of Municipality Y, ‘the rhetoric has also become tougher, we’re talking more about police and security matters now than we did just five or six years ago’ (IP16Y).

This was described as making it impossible for local politicians to avoid acting on signals of increased unsafety: ‘They shoot themselves in the foot if they don’t take the issue of safety seriously,’ said one police officer (IP9Z). Similarly, the Mayor of Municipality X described the turn toward contracting guards as almost inevitable:

First, I felt like, no […] this can’t be a municipal responsibility […]. But then we reasoned back and forth and were like, well, we have to do something, we can’t just let this situation be, we have to deal with this […] it’s about our residents, it’s about our organization and this is probably how we’ll have to work in the future. Whether we like it or not, that’s probably the case. (Mayor, IP4X)

The unsafety statistics thus become a way of not only describing but also constituting a crisis that must be managed (cf. Bigo 2002). They serve as a foundation upon which arguments for the necessity of employing guards are constructed, which legitimize everyday patrolling (cf. Walters 2012, 29). As Rose argues, ‘such numbers do not merely inscribe a preexisting reality. They constitute it’ (Rose 1991, 676).

The patrolling guards in turn generate numbers of their own, making progress both measurable and auditable. The number of interventions made by the order guards was continuously being registered in Municipality Z, and daily reports were produced in all three municipalities detailing the guards’ whereabouts and activities and giving precise descriptions of anything deemed deviant, unsafe or suspicious. The data and incident reports produced by the guards did not include any information about increased safety or less crime due to the guards patrolling. Rather, they provided information about the time, location, and character of the various incidents that the guards intervened in. This was described by both the municipalities and police authorities as being used to prioritize the allocation of resources for safety and crime prevention more efficiently. In municipality Z, the argument that the patrolling guards was a successful response was primarily based on the large number of interventions they had registered, as exemplified below:

During the trial period, the guards registered 2,871 interventions and activities of various kinds. The most common have been various forms of reprimands and ejections, which together make up about 80% of the activities.

(Evaluation report, Municipality Z, 2018, 6)

In this manner, the various data and reports produced by the guards were regularly described as measures of ‘success’ in the municipalities, contributing to produce quick and visible results and increasing efficiency in the local safety work. As technologies of government, the measurable results produced by the guards also contribute to shape the issue of (un)safety into a calculable and auditable object (cf. Li 2007; Rose 1991; Walters 2012). ‘Calculation is central’ to government, writes Li (2007, 6), and ‘requires, in turn, that the processes to be governed be characterized in technical terms. Only then can specific interventions be devised.’ The measurable and calculable here becomes a privileged form of knowledge, characterizing both the understanding of the problem and the response. As I have shown, the calculable results produced by the guards also serve to reduce the expressed uncertainty around the meaning of the measured (un)safety and around the impact of the patrolling guards as a response. The rendering of safety as technical thereby contributes to conceal the contingency of both that which is governed and the presented solution.

Marking the spaces and subjects to be ‘guarded’: safety as order and sameness

The process of rendering safety technical (Li 2007) - making it intelligible, calculable, and amenable to government intervention – also includes the demarcation of a space to be governed. In Municipality Z, this involved the very concrete delimitation of three demarcated areas to be patrolled by order guards, who were authorized by the police to use force within the boundaries of these demarcations. In all municipalities, it also included identified ‘hot spots’ assigned to be regularly visited by the guards, often based on reports of disorder, specific crimes, or ‘signs of unsafety.’ These areas included whole city centers, but also specific public squares, central travel points, parks, schools, and youth centers. Rather than completely enclosing these areas, they are, in Walters (2012, 35) words, ‘premised on a certain vector of openness’ toward the flow of people, goods and commerce, while the ‘risks’ of this openness are monitored by the patrolling guards.

The problems that were being linked to these ‘unsafe’ public spaces, in both interviews and policy documents, were wide-ranging and varying in character. They included specific crimes, such as muggings, vandalism, sexual harassment, and drug dealing. However, they also involved various social concerns that were described as becoming increasingly visible in public spaces such as homelessness, ‘begging,’ and certain groups (often young) being loud, acting out, or just hanging out, in public space. The common denominator was that these concerns were considered causing unsafety to ‘the public,’ rather than being criminal acts. In Municipalities X and Z, the guards were described as a response to a ‘changed public space,’ being connected to new groups of immigrants, ‘aggressive beggars,’ and certain crimes (Evaluation report, Municipality Z, 2018, 3). The police officer interviewed in Municipality Z said:

I think we got in over our heads when we opened the borders as much as we did in 2015 and 2016, that we could not really keep up with the development, because it brought a lot of unsafety. (Police officer, IP9Z)

Correspondingly, the security manager in Municipality Z linked the increased unsafety to a profound change of scenery in public space, with ‘different gatherings in large groups making people feel unsafe,’ following ‘the rapid increase of refugees’ in 2015 (IP2Z). Also in Municipality X, the increased unsafety was connected to new groups of immigrants, as the mayor described it:

I believe that the perceived unsafety among our residents in 2015–16 was largely related to the fact that we got—we were a big recipient when the refugees came, and it became a completely different society […] then you started feeling unsafe about going out in the evenings, because our new Swedes, they might have gathered in groups and be talking in a different language, and sometimes they might be loud. (Mayor, IP4X)

Categorizations such as ‘new Swedes’ being used here not only result in distinguishing these subjects, and behaviors such as ‘gathering in groups,’ ‘being loud,’ and ‘speaking in a different language,’ as non-Swedish (Boréus 2005), but also as unsafe. Through the drawing of these boundaries, the universal, ‘Swedish’ norm of safety against which these behaviors are being valued also becomes visible. Immigration was not described as necessarily related to increased crime or violence in Municipality X, but to the gathering of ‘newly arrived immigrants,’ often men, who were being loud and acting differently, ‘which we have a hard time accepting’, explained a police officer (IP5X).

In Municipality Z, large groups of unaccompanied migrant minors denied residence permits or awaiting decisions, were for example described as a group causing unsafety. While this situation was said to require other efforts as well, the patrolling guards were described as a ‘necessary action’ for the municipality to take to deal with an ‘unsustainable situation,’ (Mayor, IP7Z). The guards were described by local politicians and civil servants in Municipality X and Z as a powerful response to this situation, maintaining order and signaling safety to the public, reassuring people that the authorities were there. In effect, the figure of the immigrant, together with other groups described as ‘disturbing elements’ in public space, are being construed as problems of unsafety, to be made orderly by means of the patrolling guards. This constituting of (un)safety as a problem of order simultaneously legitimizes and reinforces the idea of safety as being threatened by immigration and a ‘changed public space.’

In consequence, safety becomes represented as closely connected to the familiar and recognizable, to ’us’—the white and ‘Swedish’—that is being disrupted by ‘them,’ the unfamiliar, unrecognizable, and threatening, racialized other (Ahmed 2000; Hübinette and Lundström 2014; Sager and Mulinari 2018). This resembles what Bigo has described as a ‘governmentality of unease’ characterizing the securitization of immigration (Bigo 2002). I read this unease as linked to what Hübinette and Lundström (2014) describe as a ‘white nation in crisis’ following the loss of the ‘hegemonic whiteness’ that historically characterized Swedish society. As particularly women’s unsafety is being targeted, a postcolonial idea of predominately male guards ‘protecting’ (white) women, and (Swedish) gender equality, from ‘immigrant’ men also underlies this practice of governing safety (cf. Sager and Mulinari 2018).

From a governmentality perspective, this construct of ‘the immigrant’ as a figure of disorder and threat to safety can be understood as a dividing practice. As such, I read it as shaped by an underlying rationale of immunization aimed at protecting some, at the expense of risky ‘others’ (Lorey 2015, 44; Walters 2004, 249). By positioning the threat to safety in the racialized and immigrant ‘other’, the problem is at the same time being individualized and differentiated from ‘us’. This logic is not new but connected to a long colonial history in Sweden through which ‘the immigrant’ has formed a constitutive outside to the ‘Swedish’ citizen, and the imagined white, homogeneous, safe, and gender-equal Swedish nation-state (Hübinette and Lundström 2014; Sager and Mulinari 2018).

In all three municipalities the guards are assigned a social responsibility, to build relationships with at-risk groups and defuse tensions. Their social role was particularly emphasized in the patrolling of certain residential areas, generally those housing a significant population with migrant background, being marked as ‘especially vulnerable’ and unsafe. In Municipality Y, high levels of unsafety in general, rather than a rapid increase, motivated the security guards as part of a ‘larger concept’ explained the mayor (IP16Y). Here, it also involved more police resources following the classification of several areas in the municipality as ‘especially vulnerable.’ In Municipality Z, the safety coordinator explained that:

Based on the marking of the area as especially vulnerable […] politicians have decided that safety guards are needed there […] that was decided before the results of the measured safety. (Safety Coordinator, IP8Z)

Unsafety was in these ‘especially vulnerable areas’ in Municipalities Z and Y, linked to certain violent ‘gang’ or drug related crimes, visible in the public, but also to socio-economic challenges, unemployment, segregation, and a lack of confidence in the judicial system. As a response, ‘visible guards’ were in these areas described as a way of ‘building relations’ with young people and visiting schools and youth centers, to ‘increase trust’ in the authorities and to ‘defuse the presence of uniformed staff’ (Public Procurement, Municipality Y, 2017, 3). At the same time, guards were commonly described to have a ‘deterring’ and crime-prevention effect, and were assigned to engage in practices of monitoring, surveillance, and control. This dual role of guards, both caring and deterrent, demonstrates the complex task they are ascribed to manage various societal concerns. It also makes visible that patrolling guards are not meant to improve safety for everyone.

The spatial construction of these areas as ‘especially vulnerable’ and ‘unsafe’, not only marks the domain for governmental intervention but also demarcates these spaces as deviating from the ‘ordinary.’ By establishing the complex criminal, but primarily social, challenges described in these areas as problems of (un)safety, my interpretation is that they are being externalized as problems brought in by immigration, of individual failures, as problems stemming from the areas marked as unsafe, detached from ‘us.’ In the process of making these problems amenable to government intervention, they are being individualized rather than treated as collective, societal failures. In effect, the responsibility for change is being transferred from the field of politics to the individuals who act in a ‘disorderly’ manner, commit certain crimes, lack confidence in the authorities, that are homeless, hang out in loud groups, or ‘beg’ for money. By being identified as primarily problems of (un)safety, they are made both governable and ‘solvable’ through the means of patrolling guards (Rose 1991; Li 2007). When constituted as problems of (un)safety and order it is the symptoms of these issues that are identified as in need of change, rather than their causes. Furthermore, these symptoms only become read as problematic insofar as they are interpreted as causing unsafety for others.

What is consistently left unproblematized is why these issues are produced as problems of (un)safety in the first place, and what the consequences are for those being constituted as causes of unsafety, as risky or disorderly. Thus, the exclusion of racialized, precarized and deviant others from the very idea of public safety is also silenced. This silencing is made possible, I argue, because the rendering of (un)safety as a technical issue effectively conceals these core lines of conflict, and the boundaries being drawn between those to be made safe and those considered as causing unsafety to others. It thus offers a consensus-based and depoliticized (Mouffe 2005) representation of safety as ‘a common good’ to be efficiently ‘managed’ by outsourcing the responsibility for safety to private security companies.

Outsourcing safety: private guards as a neoliberal response to a welfare state in decline

In all three municipalities, the use of patrolling guards was motivated in relation to cutbacks in various welfare services, while a lack of police resources was only mentioned in Municipality Z. Here, the security manager explained that ‘the police themselves’ had expressed the need for order guards ‘because they had such a lack of resources, especially during the summer’ (IP2Z). In the other municipalities, the guards’ new role was for example linked to fewer field social workers and described as resources for building relations with young people and disorderly or ‘risky’ individuals. In Municipality X, the guards social focus was described as filling a void caused by a temporarily closed youth center and the withdrawal of field social workers:

We don’t have any field social workers out in the town anymore […] they removed them a couple of years ago. And when I talk with the security guards, which I have done regularly during this project, I’ve gained a very good impression of how they work with young people and communicate with them. So, I think that’s the most important part. (Security Coordinator, IP5X)

Civil servants and politicians in all three municipalities underlined that the guards were not meant to replace the police or social services. As the Security Manager in Municipality X explained, ‘this is still a guard project, it is not a social effort in any way’ (IP5X). However, as exemplified in the quote above, the guards’ new social role was often expressed as filling the gap of cutbacks of welfare services. The CEO of a security company described that, ‘we don’t work as police officers, but sometimes have to do that, we are not social workers, but sometimes have to be that, and I guess that’s what often comes with too much responsibility’ (IP19X).

While the contracting of patrolling guards for increasing safety started as a temporary measure, it was turning into a long-term solution in all three municipalities. The guards were described as becoming a familiar element in public spaces and in certain neighborhoods, and as an ‘invaluable’ and ‘cost-efficient’ resource for the police and the municipalities:

I would love to see more field social workers out around the town too, in addition to security guards. But the gain here is that we can make it into a larger package, with both monitoring and building social relationships […] then we sort of get a better deal. (Security Coordinator, IP5X)

‘The guards are really there every day’ explains the Mayor in municipality Z (IP7Z), and are consequently often the first person to reach an incident. In relation to their visibility, the order guards in Municipality Z were said to function as ‘the municipality’s public face,’ reassuring people that ‘the authorities are there’ (Security Manager, IP2Z).

From this perspective, the patrolling guards – becoming both indispensable and permanent – are filling the void of a diminishing welfare state and becoming a response to its effects. As underlined by the Mayor of Municipality Y, ‘with increasing inequalities there also comes increased unsafety’ (IP16Y). Many interviewees also expressed concern over the risk that resources would be withdrawn ‘from that youth center or those social efforts, and spent on guards’ (Opposition Politician, IP6Z).

The short-term contracts awarded to the security companies were further described as running the risk of resulting in a lack of continuity and a loss of knowledge and legitimacy (IP16, Mayor, Municipality Y). As noted by Crawford et al. (2005, x), private policing in the UK ‘generates significant turnover and flux in the delivery of community safety’, and risks undermining relations between ‘policing personnel and local communities.’ One of the few interviewees to raise concerns about these aspects was the CEO of the contracted security company in Municipality Z:

I would much rather see the police patrolling than us. Maybe that’s stupid of me to say, but […] the police have a completely different way of working, they have an intelligence service, they can work much more efficiently, they have an even closer cooperation with the social services, which means that they can work more intimately with these problem youths, work through more steps, than we can. (CEO, IP10Z)

As highlighted in the quote above, the outsourcing of safety involves a transfer of responsibility for both policing activities and social efforts to, in this case, order guards, who have limited training and resources. The guards’ new crime-prevention role, especially order guards with the authority to use force, arguably suspends the state’s monopoly on the use of force through what a local politician described as a ‘privatization of the police’ (IP6, Municipality Z). It also contributes to further blurring the boundary between the responsibility of the public and the private sector, making it more difficult for citizens to grasp who is accountable for the guards and their operations and in whose interests they serve (cf. Loader 2000; Crawford et al. 2005). As private businesses making a profit from unsafety, they are inevitably part of ‘a market offering security expertise’ with, as Walters writes, ‘a vested interest in the persistence of the very threat it is supposed to eliminate’ (Walters 2004, 255; cf. also Hansen Löfstrand et al. 2019).

While the altering of the monopoly to use force was described as a problematic effect of the patrolling order guards in Municipality Z, the mayor explained that:

There was, like, no major discussion, at least not on the municipal council, about whether it’s right to give a monopoly on violence to trained police vs. guards with shorter training.

(Mayor, IP7Z)

Rather, the political decision to employ patrolling order guards was characterized by a ‘broad political consensus,’ said the security manager in Municipality Z, underlining that what was being discussed was rather the details of the mission, the ‘exact design, whether it was enough, too little or too much and so on […] but basically all parties agreed that we needed to at least try it out’(IP2Z). In the municipalities, the need for patrolling guards was articulated as a natural, even inevitable, response to unsafety, characterized by a lack of political debate. As stated by the Mayor of Municipality X, ‘we’ve been very much in agreement, so there hasn’t been any politics in this at all because, after all, it’s really just a question of creating safety’ (IP4X). The nonpolitical representation of safety as a virtue, a ‘common end’ that is given rather than debated (Brown 2015, 128), in this manner seems to effectively conceal the political implications of the contracting of patrolling guards. In that sense, the immunization rationale is at work also in relation to the outsourcing of responsibility for safety to private security companies, operating to immunize the current political order.

Concluding discussion: immunity biopolitics at work

The empirical analysis has shown how (un)safety is shaped into an object of government being met by patrolling guards as a response. Despite the contextual differences between the studied municipalities, and the two different categories of guards being contracted, the study demonstrates the similarities in the governmentalities of (un)safety underlying this proposed solution.

First, I showed how the rendering of (un)safety as technical, calculable, and amenable to government intervention simultaneously detaches it from its social and political connotations, largely making the measured levels of unsafety the problem to be addressed. Through this process, the guards form a visible and efficient solution, providing visible and measurable results. Secondly, I demonstrated how, in the demarcation of the space to be ‘guarded,’ safety is problematized as a matter of order and maintaining a homogenous public space. In this process (un)safety problems are largely being linked to risky ‘others,’ detached from ‘us’ and made governable and ‘solvable’ by means of patrolling guards. Thirdly, I demonstrated how guards are becoming a response to a welfare state in decline. Following a technical and depoliticized representation of safety as a ‘common good’, I argued that through the contracting of private guards, democratic dimensions of policing are being altered and exclusions from public spaces are enacted, largely without political contestation.

To understand how this depoliticized notion of safety is maintained, and the implications of the naturalized and racialized exclusions it encompasses, I will conclude by further discussing the underlying rationale of immunization underpinning it (Swyngedouw and Ernstson 2018; Esposito 2008; Lorey 2015). In line with this rationale, the patrolling guards – despite their pronounced social and caring function – contribute to regulate, control and in different ways exclude those constituted as ‘unsafety problems’ from the safe public spaces being produced.

In a broader sense, the rationality of immunization builds on the idea of managing (un)safety in a manner that permits life to go on untouched for those included in the immunization, who are to be made safe, while increasing the exposure for those constituted as causing unsafety (Swyngedouw and Ernstson 2018, 17). It thus entails a promise of safety, and of managing the problems described as linked to unsafety, without political, social, or redistributive change and therefore constrains any threat to the consensus around the current political order.

As Lorey (2015, 44) argues, the perceived risk within contemporary immunity politics comes from within, threatening to collapse social unity and ‘the common consensus’, that sustains the current order. However, by positioning this threat of unsafety at the social margins, stemming from the ‘especially vulnerable’ areas and those marked as ‘other,’ it is being differentiated from ‘us’. The patrolling guards can thereby be presented as a response that protects the safe community from this threat, while leaving the politics leading up to these problems both unproblematized and untouched.

As a political ‘solution’, the patrolling guards are thus protecting the immunized, and securing the economic-political order, from the symptoms of these problems, rather than dealing with their causes. My argument here is that as a practice of government, the patrolling guards forms a ‘solution’ that individualizes and conceals the problems it governs, offering a quick fix to complex societal concerns. Following the immunity rationale, the contracting of patrolling guards can be understood to have become such a widespread response to (un)safety in Sweden precisely for this reason, because it forms a government intervention that avoids dealing with the causes and therefore effectively conceals any ‘challenge to the status quo’ (Li 2007, 7).

In consequence, the contracting of patrolling guards constitutes a depoliticized response to deeply political issues, which risks intensifying the (unsafety) problems that it was designed to address. In its excessive form, as Swyngedouw and Ernstson argue, the ‘immunological biopolitical fantasy’ becomes self-destructive to the point of ‘auto-immunization,’ turning against that from which it set out to immunize the public (Swyngedouw and Ernstson 2018, 5; see also Esposito 2008). As I have shown, the immunizing dynamic reinforces, legitimizes and depoliticizes ‘a constitutive split of the demos’, between those to be made safe and those deemed to be causing unsafety (Swyngedouw and Ernstson 2018, 24). In effect, it risks establishing a permanent state of unsafety due to an ever-increasing exposure for those marked as dangerous others - ‘the poor, immigrants, and undocumented aliens’ (Neyrat 2011, 110) - who are positioned outside the (safe) political community (Lorey 2015, 44–45). And all in the pursuit of sustaining a fantasy of immunization, of maintaining the idea that the social challenges marked as (un)safety problems can be efficiently managed by contracting patrolling guards, while life as we know it can remain untouched.

Acknowledgments

First of all, I would like to thank all of the participants in this study who have contributed with their valuable time and reflections. I also want to thank the two anonymous reviewers as well as my colleagues Josefine Olsson and Elisabeth Olivius for generous engagements in the article and helpful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Marianne and Marcus Wallenberg Foundation [MMW 2012.0211].

Notes on contributors

Jennie Brandén

Jennie Brandén has a PhD in Political Science and currently works as a postdoctoral fellow at the department of Epidemiology and Global Health, Umeå University. Her research is largely situated in governmentality studies and linked to critical studies of gender equality, safety and security. Her research interests involve the politics of violence and safety in Sweden, with a particular analytical focus on how gendered and (post)colonial relations of power operate in and through practices of government. Her recently published dissertation “In the Name of Safety: power, politics and the constitutive effects of local governing practices in Sweden” demonstrates that while safety is increasingly presented as an answer to societal issues in Sweden, the government of safety operates through a de-democratizing dynamic of governmental precarization.

Notes

1. The Swedish Police Authority describes an ‘especially vulnerable area’ as ‘characterized by social problems and a presence of criminality that has resulted in a broad unwillingness to participate in legal processes and difficulties for the police to fulfil their mission. The situation is considered urgent’ (2017, 10).