Partnership policing and the dynamics of administrative growth

ABSTRACT The current article reports findings from a research project on partnership policing in Stockholm, Sweden, to investigate how partnership policing strategies translate into social action. Consideration is given to the ways in which police officers and city employees produce chains of administrative tasks as they navigate their institutional environment and strive to produce legitimacy for partnership policing. More broadly, the findings suggest that the inner mechanisms of a partnership approach to policing are shaped by the self-referential (Eigendynamik) character of administration. The article discusses implications for partnership policing and for the broader literature on policing.


Introduction
For well over three decades, partnership policing has been a seminal strategy to tackle crime and urban safety problems across the globe (Crawford 1999, Garland 2001, Brogden and Preeti 2005, Fleming and Wood 2006, Jacobs 2010, Delpeuch and Ross 2020).Existing scholarship has significantly advanced our understanding of the drivers of this strategy (Johnston 1998, Garland 2001), and various dimensions that promote (e.g.trust) and impede (divergent organisational cultures and resources) collaborative processes (Fleming and Rhodes 2005, Fleming 2006, O'Neill and McCarthy 2014, Crawford and Cunningham 2015).However, as Quéro and Dupont (2019) assert, less is known about the organisation of and relationships between actors in policing partnerships and about policing networks more broadly.Further, despite recent advances in research into the importance and impact of the mundane practices of policing (Jacobs 2010, Fassin 2013, Crawford and Hutchinson 2016, Youansamouth 2020, Sausdal 2021, Nøkleberg 2022), there is a dearth of research into the everyday realities and experiences of partnership policing.
Against this backdrop, the current article sets out to explore how a partnership approach 1 to crime prevention and urban safety translates into social action by actors 'on the ground'.Drawing on empirical data from the lived realities of partnership policing in Stockholm, the article argues that Swedish policing partnership symbolises more generally how, at the practical, everyday level, one of the main outcomes of partnership policing is the production of seemingly endless chains of meetings and documents (administration).Although the administrative dimension of policing is well-documented in scholarship into police culture (see, for example, Ericson and Haggerty 1997, Van Maanen 2010, Fassin 2013), the article explores a neglected driver in these processesnamely, the nature of administration itself, whereby the participants' need for order and management of disorder sustained existing administrative processes and initiated new ones.Accordingly, the main thrust of partnership working revolves around the governance and administration of inter-and intra-organisational relationships.In short, the article argues that partnership policing is largely an administrative exercise, suggesting that the 'appeal' (Crawford 1999) to partnershipsinadvertentlyis also an appeal to administration.In unpacking these social processes, the article argues that bottom-up as well as top-down processes fuel the proliferation of administration in Western societies variously associated with new public management, the audit society, and regulatory changes (Power 1997, Hood et al. 1998, Levi-Faur and Jordana 2005, Hood and Dixon 2015).
The remainder of this article is organised into six sections.The next two sections outline previous research and influential theories into the expansion of administration; thereafter, the data and methods on which the article is based are described.The following two sections cover the partnership structure and the processes of administrative growth involved in the social translation of policies into practices.The last section presents conclusions.

The partnership approach to policing
In the 1980s and 1990s, the stubbornly rising crime levels of the 1960s and 1970s coupled with a rhetoric advocating that 'nothing works' ushered in two major shiftsor, perhaps more accurately, spurred the re-emergence of previous thinking (Johnston 1992, O'Malley andHutchinson 2007) in policing and crime policy.Whilst the discontent with the traditional policing model provoked a re-orientation of policy towards a penal regime that is increasingly more punitive and dramatic (Garland 2001), it also drove the emergence of policing arrangements centred on a plurality of actors and partnerships directed towards preventive, risk-oriented schemas (Garland 2001, Johnston andShearing 2003).As research suggests, the rise of partnership policing is firmly connected to global structural shifts that have occurred across the globe in late modern society (Johnston 1998, Garland 2001, O'Malley and Hutchinson 2007), including the shift away from government to governance whereby 'state-centred forms of social control' are coupled with 'forms of regulation developed and implemented at local levels by local governments' (Lee and Herborn 2003, p. 26).In general, this shift has seen the rise of partnerships, not only in policing but in many different context and sectors, including but not limited to education, health care, elderly care, and social work (see, for example, Danermark and Kullberg 1999, Fimreite and Laegreid 2009, Jacobs 2010).
Despite the popularity of the partnership approach to policing, it still 'has a lot to prove in terms of its capacity to sustain what is envisioned' (Alstam and Forkby 2022, p. 3).One problem is that research and policy are 'informed by normative assumptions about the efficacy of partnership and its desirability as a model for governance' (Jacobs 2010, p. 928).Although there are notable exceptions (O'Neill and McCarthy 2014), research into partnership working repeatedly stresses how the fraught relationship between the police and its new partners hampers the ability to produce desirable results (Rosenbaum 1994, Hale et al. 2005, Huxham and Vangen 2005, Crawford and Cunningham 2015, Crawford and L'Hoiry 2019, Forkby 2020).Indeed, as Youansamouth (2020, p. 13) notes, partnership working itself often turns into 'a wicked problem'.Whereas partnership policing has received extensive attention amongst criminologists and police scholars, there has been less focus on the experiences of partnership working (Jacobs 2010) and almost no research into how the lived realities of collaborating relate to public administration.Perhaps, this is because 'police studies remain isolated from administrative studies' (Fleming 2008).
Although Sweden and the other Scandinavian countries for many decades were regarded as exceptions in terms of criminal justice and penal policy, scholars in diverse fields have proclaimed the 'decline' of 'Scandinavian exceptionalism' (see, for example, Pratt 2008, Høigård 2011, Pierre 2015).Rather than being characterised by exceptionalism, Sweden has followed the general development outlined in the criminological literature.For example, despite a notable crime drop in the 1990s (Nilsson et al. 2017), Sweden has adopted increasingly more punitive policies (Tham 2022), and more resources have been allocated to the police and the criminal justice system (Uhnoo and Hansen Löfstrand 2018).Moreover, current crime policies are becoming increasingly punitive, particularly as a response to the past years' increase in gang-related activities.At the same time, research indicates that the police' capability to govern crime by itself is limited (Wikström and Torstensson 1999).Thus, like its counterparts elsewhere (see, for example, Brogden and Preeti 2005, Crawford 2009, Diphoorn and Berg 2014, Delpeuch and Ross 2020), the Swedish police are today expected to form partnerships with local authorities, private business, and voluntary forces (Johansson 2014, Uhnoo and Hansen Löfstrand 2018).
Mirroring many of the reforms adopted in other countries (i.e. the Crime and Disorder Act in England and Wales and Contrats locaux de sécurité in France), Sweden embraced the partnership approach in the mid-1990s (Hörnqvist 2001, Uhnoo andHansen Löfstrand 2018).In general, the unfolding of a partnership approach to policing across countries and jurisdictions has been characterised by important differences (Brogden and Preeti 2005, Crawford 2009, Delpeuch and Ross 2020).For instance, whereas the Crime and Disorder Act in England and Wales (1998) was intimately connected to situational crime prevention and placed a statutory responsibility on the police and local authorities to engage in partnerships, France mainly stressed social interventions.Sweden, and the other Nordic countries, took something of a middle ground as social prevention became integrated with situational approaches on a voluntary basis (Takala 2005, Johansson 2014, Tham 2022).
Thus, although the partnership model is characterised by diversity in its directions and scope, it is generally accepted that it challenges traditional assumptions about crime, its causes, and the range of police activities (Marnoch et al. 2014), as well as 'many bureaucratic assumptions about professional expertise, specialisation, and disciplinary boundaries' (Crawford and Evans 2017, p. 800).In Sweden, this development has spurred worries about a more 'unprofessional approach to crime prevention' characterised by 'slogans' and a 'let a thousand flowers bloom' attitude towards policing and crime prevention (Wikström and Torstensson 1999, p. 460).Importantly, the main outcome of partnership policing has not been a reduction of crime.Rather, as Hörnqvist (2001, p. 55) writes with respect to Sweden, the partnership model 'was not directed towards those who engage in criminality but towards those who are intended to prevent crime'.Relatedly, in a context where the police and local authorities engage in partnerships because it has become the 'true' form for crime prevention, there is evidence that expertise about 'how to execute collaboration' is being formed instead of effective and meaningful crime prevention strategies (Alstam and Forkby 2022).
Accordingly, it is useful to focus on partnership working from the outlook of professional practices and the 'everyday experiences, activities and expectations' of those responsible for translating partnership policies into social practices (Crawford and L'Hoiry 2019, p. 58).Thus, partnership policing should be approached as 'meaning in action' (Bevir and Rhodes 2006, p. 1).Essentially, what is at stake here is the question of what professionals do when they collaborate.This resonates with Mariana Valverde's (2011, p. 5) argument that 'all that we can know about security is what people do in its name, and that therefore our focus should be on practices of governance that in fact appeal to "security"'.That is, security is not simply lower instances of crime or a less fearful population but the system of governance 'in which organisations struggle to make complex problems manageable' (Jacobs 2010).This includes the ways partnership policing is organised and 'called into being, made visible, and responded to' through administrative processes and practices (Ericson and Haggerty 1997).As a caveat, the article primarily engages with the literature on policing and public administration in Western societies.Although there is an important scholarly literature into policing outside the Western setting (see, for example, Johnston and Shearing 2003, Brogden and Preeti 2005, Diphoorn and Berg 2014, Diphoorn 2016), the scholarship into public administration, as Johnsen (2019) points out, remain focused on the Western, Anglo-American, context.Mirroring this limitation in public administration scholarship, the ideas that inform the current article's theoretical framework are grounded in research into public administration in Europe and Anglophone contexts.

The self-reinforcing force of administration
The proliferation of the partnership approach to policing should not merely be considered as emerging from central government.While legal changes and the creation of a national and institutional frameworks have been critical in the promotion of partnership policing, policing strategies also disseminate and are shaped from the bottom-up (Pollitt andBouckaert 2009, Marnoch et al. 2014).As Marnoch et al. (2014, p. 308) note with respect to their empirical case, 'internally driven professionalisation and specialisation were frequently the main policy drivers'.Analogously, policing involves self-referential processes whereby police officer's create and respond to their own (organisationally) defined categories and typologies (Manning 1982(Manning , 2004)).Indeed, such 'enactments' of an organisational and institutional environment involve administration (i.e.documents that classify offenders or communities, meetings where classifications are discussed and decided, the production of statistics and qualitative measurement, etc.).However, administration, too, is self-referential since meetings and documents tend to grow on their own volition, making administrative growth one of the most common, yet overlooked, consequences of the partnership approach to policing.
Suggesting that the administration expands in policing and in Western countries more generally is hardly novel.Key theoretical contributions attribute administrative growth to new public management (Hood and Dixon 2015), advanced capitalism (Rose 1999), audit society (Power 1997), and regulatory capitalism (Levi-Faur and Jordana 2005).In accentuating programmatic top-down changes, these and allied concepts capture how the police and other professionals spend increasingly more time on administrative tasks, including meetings, reports, communication, measurement, audits, and statistics (Forssell and Ivarsson Westerberg 2014).Nonetheless, as already suggested, people and groups are not merely instructed to administer more; they themselves promote administrative tasks.Ironically, then, though policing agents may complain about meetings and the production of documents (see, for example, Ericson and Haggerty 1997, Van Maanen 2010, Fassin 2013), they embrace such tasks and willingly act as vehicles in the creation and reproduction of infinite circles of administration (Åkerström et al. 2020, 2021).
Whilst there has been a growing public and institutional demand for accountability and risk-management, which are important drivers of administrative growth (Ericson and Haggerty 1997), the idea that administration grows on its own volition is based on the sociology of Georg Simmel.Famously, Simmel (1957) applied Eigendynamik to various social phenomena (such as fashion, money, and the city) to unpack how these aspects of modern society evolve as an effect of the indefinite, pendulum movement between imitation and innovation.Key to partnership policing is people's desire for order (structure, coordination, etc.), which makes administration grow on its own force.For example, the creation of a joint youth crime project group, a new or revised policing strategy, or routines epitomises efforts to attain order (clarity, transparency, and coordination); however, as the order is an ongoing accomplishment (cf.Czarniawska and Sevón 1996), these efforts simultaneously introduce ambiguity and new circumstances (disorder) that must be managed.
One important effect of the Eigendynamik of administration is that administration take precedence over the core task of the police and local authorities (cf.Power 1997, Hjärpe 2019), that is criminal investigation, rehabilitation, crime prevention, and interaction with citizens/clients.Relatedly, meetings and documents have been identified as salient features of the participants' need for order and control in joint work across organisations (cf.Hall 2012).Likewise, and with direct relevance for partnership policing, Åkerström et al. (2021, p. 61) stress that there is a danger that meetings 'becomes proof of "actual" collaboration' and operate in of themselves as a 'token of quality', which shifts the focus 'from actual collaboration to procedural and technical concerns around the form and its associated meetings, which reproduces and strengthens the administration society'.
In sum, administration grows as a result of top-down programmes since, for instance, the police and local authorities are subjected to scrutiny and audits and are required to engage in collaboration with other partners.Nevertheless, following a Simmelean perspective, the expansion of administration observed in Western countries may also be examined as a process that operates according to its own forcespecifically, as a result of the dynamic interplay between people's need for order and freedom.In what follows, attention will be directed towards the ways the self-reinforcing force of administration shapes partnership policing.

Data and methods
The research reported in this article is part of a larger project analysing the structure and organisation of the formal collaboration agreement between the City of Stockholm and the Swedish Police Authority (signed in 2019)hereafter referred to as the Stockholm partnership. 2Specifically, the project aims to analyse how collaboration is organised and how mutual problems are identified and translated into social action by 'boots on the ground'.With respect to the governance of partnership policing in Sweden, collaboration agreements were first introduced in circular PoA-480-5788/06 titled 'Police-municipal cooperationfor locally based police activities throughout the country' (2008).And since 2008, collaboration agreements have been implemented on a national scale, as per 2013 nearly 90% of Sweden's municipalities have signed partnerships with The Swedish Police Authority.Although formalised collaboration agreements have become a significant instrument in the governance of partnership policing in Sweden and beyond (see, for example, Delpeuch and Ross 2020, Nøkleberg et al. 2022), there is almost no systematic knowledge about how and with what consequences activities are regulated via collaboration agreements.In the Nordic context, one explanation is that scholarship into policingdespite the interest in pluralisation, networks, and partnerships in other contexts (see, for example, Johnston and Shearing 2003, Wood and Shearing 2007, Diphoorn and Berg 2014, van Steden 2017, Whelan 2017)largely remains police-oriented (see, for example, Finstad 2000, Granér 2004, Høigård 2011, Sausdal 2018).However, the past five years have seen the emergence of scholarship that explores various forms of policing initiatives that occur in collaboration with, and beyond, the police (see, for example, Hörnqvist 2015, Engdahl and Larsson 2016, Forkby 2020, Stenström 2020, Nøkleberg 2022).Nonetheless, the Nordic scholarship largely examines collaboration and coordination of activities between actors in loosely coupled security networks (i.e. between insurers and police).In a contrast, the research reported in this article examines a partnership characterised by a high degree of structure, formality, and voluntarism.While voluntarism separates Sweden from jurisdictions where the police and local authorities are given a statutory duty to develop partnerships (i.e. the United Kingdom), the Swedish approach incorporates assumptions found in such contexts, such as a shared leadership approach, the recognition that crime and safety are complex, and that solutions must be multifaceted and involve multiple actors (Hughes and Mclaughlin 2002).
The Stockholm partnership is formalised into 14 agreements (one overarching city-wide agreement and 13 local collaboration agreements).At the city-wide level, the format for the partnership is a two-tiered structure involving (i) a steering group of senior representatives from the City of Stockholm and Police Region Stockholm and (ii) a planning group.The city-district-department level uses a more complex structure involving (i) a local steering group, (ii) a local planning group, (iii) thematic subgroups, and (iv) thematic project groups of all those involved in a specific intervention (i.e.street patrol, youth crime, etc.).Complexity was further enhanced by the fact that the geographical boundaries of the City's district department often transcend several local police districts.Consequently, several city district departments had a local agreement with more than one police district.
During the research project, 19 interviews were conducted with police officers and city employees working within the Stockholm partnership.The total number of participants was 22 as some were interviewed in pairs.See Table 1 for details.
Research participants were selected from two city district departments: one located in the city centre and one suburban district.Interviews were also conducted with professionals operating at a citywide level.The districts were selected to address the diversity of crime and safety problems in the Stockholm area.Although the districts are characterised by different social conditions (i.e. the area located in the city centre has a buzzling nightlife, the suburban district includes two socially disadvantaged communities), the administrative processes examined in the current study were identical.The conclusions, then, are applicable to both districts.The majority of the interviewees were members of the partnership's steering, planning, thematic subgroup, and project groups.However, some research participants, for example, police officers, were not official members of any group, but were included in the study because they were responsible for the implementation of decisions made within the structures of the Stockholm partnership (i.e.targeted community patrols).Since resources were not solely devoted to the Stockholm partnership, all participants had to perform other job duties, such as case management (social workers) and patrols (police officers).Research participants were also involved in collaborative work that was not a part of the Stockholm partnership, such as inspections of restaurants and pubs.
All interviews were recorded and transcribed.The interviews were conducted in Swedish and were translated into English by the author.Fieldwork was conducted between April and November 2022, where data was collected at various meetings, one partnership conference, and collaborative activities.Observational data were recorded into fieldnotes.These data sources were complemented by an analysis of partnership documents (i.e. the formal collaboration agreement, local collaboration agreement, situation analysis, activity plans, etc.).The data were analysed using thematic analysis, which is a method for identifying and analysing patterns of meaning in a qualitative data set that permits the development of rich, detailed, and complex accounts of these data.In practice, the analysis was performed in several steps, involving general coding and more focused coding aimed to develop a more detailed and analytical understanding of the data.The coding process was conducted using NVIVO qualitative analysis software.

Order, disorder, and administrative growth
My job is to convene the steering group and the planning group.In collaboration with the police, I produce decision guidance information to both these groups.Steering group meetings are held regularly twice per term, and we gather the planning group before and after those meetings.
Interviewer: What about the situation plan … Is it your job to write the local situation plan?Interviewee: Yes, it is, and I do it together with the police, but it's difficult, we discussed this issue at the conference last week.Today, we have a template that must be followed, we didn't have that before, so our working methods has evolved little by little.For instance, a situation plan must contain certain headings, so we make sure that these documents include the same information.To be honest, I don't think that this is the case at the moment.Seemingly, some situation plans are sixteen pages long, but one city district department has produced a plan of seventy pages.That's a bit astonishing, I think, […] so there's room for improvement.Specifically, the situation plan must contain manageable and actionable information.The plan should not be wide-ranging because when this document trickle down to one of the project groups, the plan will be too broad for them, they need precise information, and that's the theme for our next conference and workshop sessions with the project groups.What type of information must be included in the situation plans for the project group to do a good job?(Interview 1, city employee) The excerpt serves as an introduction to administrative circles in partnership policing.In particular, four themes are important for the ensuing discussion and analysis.First, the quotation highlights how one meetingin this example, the steering group meetinggenerates new meetings (the pre-and post-meeting held by the planning group, conferences, workshops).Second, the excerpt suggests that partnership policing involves documents about documents (the local situation plan In what now follows, the article explores the empirical findings around the administrative process involved in turning the Stockholm partnership into social action.In so doing, it draws on insights from the notion of Eigendynamik and stresses the link between partnership policing and administrative growth.

Organisational order and administration
Indeed, a key dimension of partnership policing is the interaction between partners which, in a bureaucratic context, means administrative tasks (meetings and documents).As sociological scholarship exposes, there is a multi-layered and self-reinforcing dynamic between meetings and documents (Buckland 1997, Thedvall 2019, Åkerström et al. 2021) that inevitably suggests that a partnership approach fosters administrative growth.In essence, then, the professional's will to create structure and order results in the creation of documents and meetings, which, in turn, generate new spirals of administrative tasks.
This dynamic gains momentum when professionals are expected to target crime and safety problems in a large and diverse city such as Stockholm (with a population that comes just south of one million citizens).For instance, as a part of the partnership agreement, local city departments district was instructed to produce a series of shared documentsincluding, but not limited to, a local collaboration agreement, a local situation plan, an activity plan for each of the 10 joint-priorities, and annual follow-ups of activities.As the City of Stockholm consists of 13 city district departments, and as each city district department must host 10 project groups, the agreement initiated the production of a complex set of documents required and reviewed on a city-wide level.Whilst the demand is to produce local situation plans, these plans must cover 10 joint priorities; thus, each plan entails complex and variegated crime and safety problems (see Table 2).
Essentially, this 'smorgasbord approach' is inclusive and designed with both the police and the city district departments' responsibilities and mandates in mind.As previous research makes clear, the 'smorgasbord approach' is a conflict-avoidance strategy utilised to mitigate differences in goals and mandates between actors in partnership policing (Crawford 1999).Nevertheless, as the Stockholm case suggests, the approach increases the administrative load on the police and its partners and fosters feelings of confusion and uncertainty amongst professionals.Whilst the interview data suggest that some city district departments responded to all priorities while others picked and chose, several interviewees stressed their feelings of irrationality stemming from their perceived lack of administrative action taken by the city-wide level: 'How have they picked them [the priorities]?There is no situation plan, a general situation plan.It's politics.There are certain areas that they want to include' (Interview 6, police officer).Additionally, this notion, in combination with the demand that actors 'on the ground' must produce an analysis of all 10 crime and safety problems, gave rise to ritual administration: Obviously, we cannot write about private security guards, that's not a method, or CCTV, that's a technology, but because they are part of the shared priorities, we must report what we do in these areas.Even if it's not an activity, we report what we do.(Interview 6, police officers) Likewise, several police officers reported that project meetings were held even when the shared priority had not been identified as a problem at the local level.As one police officer explained: We don't have gang desisters in our local police district.Of course, I understand, there is a point in being prepared in the event that we will in the future.Still, we have a gang desistance group because it is mandatory.(Interview 10, police officer).
In societies concerned with risk management, administration in policing grows as a part of external demands and the governance of risk in other institutions (Ericson and Haggerty 1997).External administrative process is also linked to the creation and implementation of partnership policing (i.e.documents, strategies, programmes, and routines in other organisations or departments).In part, the Stockholm partnership came into being via circles of administration that fitted the Stockholm partnership into a wider corpus of bureaucratic texts which, in turn, served to foster legitimacy for crime preventive and safety enhancing actions.Specifically, the Stockholm partnership was represented as a continuation of a process at the national governmental level as it referred directly to the detailed, step-by-step, recommendations issued by the National Council for Crime Prevention (NCCP) and the Swedish Association of Local Authorities and Regions (SALAR).Relatedly, the partnership was linked to the City's and the police' general policing and security strategies, such as the safety strategy (the City) and the police' civic promises.Referencing external strategies and processes, then, emerged as a key mechanism for creating structure, generating legitimacy for the partnership approach, and, conversely, using a partnership approach to policing to gain legitimacy to other strategies and security policies.One important practice for actors enrolled in partnership policing, then, is to create partnership documents with references to extant documents.
Accompanying the production of partnership documents are partnership meetings (Åkerström et al. 2021).As mentioned, the partnership in Stockholm involved steering groups, planning groups, thematic subgroups and project groups for each of the 10 priorities.The individuals engaged in these groups experienced mundane administrative demands as the production of documents, both before and after the meeting, were necessary to coordinate inter-agency collaboration.For instance, as a part of the research, the author attended group meetings called 'place collaboration', which gathered a group of local partners with a stake in crime prevention and safety in a particular geographical space.These meetings required sending invitations, preparing a meeting agenda, creating PowerPoint presentations, and taking notes.After each meeting, post-meeting memos were disseminated, reviewed, and commented on by the participants at the next meeting.Essentially, the interviews suggest that there was a shared lived reality that these taken-for-granted aspects of public administrative life are vital, yet often neglected, dimensions of partnership policing: You might think that collaboration should work by itself.You think if we sit down together, that's enough, we'll collaborate.But simple things such as booking a meeting, managing an agenda, memo-writing or protocols.There's a lot of little things.(Interview 3, city employee) As previous research makes clear, the mundane 'matters' in policing: boredom and the pains of quotidian tasks are, for instance, known to shape the more dramatic aspects of policing (Fassin 2013, Sausdal 2021).While the administrative actions described above are not connected to the drama of policing 'on the street', they show that one fundamental aspect of partnership policing is the creation and management of a rational administrative order.Underpinning this work is the assumption that an effective reaction to crime necessitates a cogent structure that, in practice, is based on a series of administrative tasks.For reasons that will be elaborated on later, maintaining order between actors in partnership policing is an endless endeavour as signs of disorder constantly arise from interactions across organisational boundaries.

Meetings as visual expressions of partnership policing
In an organisational context, meetings often act as a vehicle for building trust and co-ordinating action, yet previous research suggests that meetings are tolerated rather than treasured (see, for example, Kello 2015).Still, research also indicates that many professionals are committed to meetings (Schwartzman 1989(Schwartzman , Åkerström 2019)).The interviewees expressed the view that meetings operate as a pivotal catalyst for producing many of the prerequisites for working partnership outlined in the literature (e.g.trust, communication, and legitimacy) (Fleming and Rhodes 2005, Crawford and Cunningham 2015, Crawford and L'Hoiry 2019).As one police officer commented, '[meetings] bring us closer to each other'.In a similar vein, when asked about the role of meetings, a city employee pointed out 'we [get to] know each other by name and appearance, and that makes it easier to call, it's easier to talk, everything is easier' (Interview 7, police officer).
Nevertheless, there was also the experience that partnership meetings might endanger partnership policing.These feelings are illustrative of Huxham's (2003) research into the ways 'collaborative inertia' proliferate and hamper partnership working.Specifically, professionals from all organisations stressed that partnership working could involve the wrong kind of order, which several interviewees referred to as 'collaboration for the sake of collaboration'.This expression is interesting as it demonstrates the participants' contradictory feelings about the value of these social gatherings.On the one hand, the expression captures the lived experience of ritual meetings from which few actual actions transpire: There is a forum for collaboration that meets twice a year, but that's a chimera.It's exchanging situation plans, and I've tried, you see this immediately, we cannot have lots of meetings where we only exchange situation plans.If something is happening and you cannot do anything about it, you must put it aside […].So we don't attend these meetings anymore […].We cannot invest time in things that ultimately don't reduce crime and increase safety, to just talk about stray dogs, that doesn't cut it.We don't have time for that.(Interview 10, police officer) The above interviewee's disdain for an order where 'you only exchange situation plans' was echoed by several participants.Another police officer said in an interview, We only exchange situation plans with each other.It does not lead anywhere except that we might create a good atmosphere, a trustful atmosphere, and a strong network, and that, there's no reasons to hide the fact that's one positive outcome that comes from us sitting down and talking to each other.(Interview 2, police officer) Relatedly, collaboration for the sake of collaboration in partnership policing was also said to involve meetings where the participants presented their units' remit, legal mandate, and operational processes.For instance, I observed one group that had been responsible for implementing crime preventive and safety enhancing activities in a local area for well over a year.The fieldnotes show how the meeting's main activity was presentations where each participant described their unit and department and their assessment of the current situation in the area.On the other hand, even though project groups were experienced in terms of being what several interviewees referred to as 'nice fika 3 groups' in which little real collaborative action except the actual meeting transpires, there was also a feeling that these meetings were not useless.They could still have beneficial outcomes because they strengthened the social bonds between the participants.In essence, administration in partnership policing may always be warranted.One city co-ordinator echoed this view: You are there [at the partnership meeting] with your competency and experience, but it's not always useful at this meeting, but it might be needed at the next meeting.(Interview 1, city employee) Here, it is crucial to stress that in a partnership policing setting, whether meetings are deemed useful is of less concern.The point is that meetings become visual representations that partnership work is in fact taking place (Åkerström et al. 2021).In essence, without meetings (and documents) there is no partnership.Moreover, when partnership meetings are perceived as 'collaboration for the sake of collaboration', more administrative demands are often imposed; for instance, new meetings are held to discuss why a group is not producing results, which, in turn, may yield revised documentation (i.e.new goals).These processes are often initiated and requested by professionals on the ground because they want to make sense of their, and others', roles and responsibilities in partnership policing.

Administration as a sign of disorder
The Eigendynamik of administration assumes that there is an ongoing interplay between attempts to bring order (structure) and manage disorder in partnership policing.The latter arises from partnership working because when people meet and work together, problems, inconsistencies, and ambiguities emerge from these interactions.That is, documents and meetings are mechanisms intended to bring order and clarity, but they 'themselves generate new issues or situations that appear disorderly' (Åkerström et al. 2021).Consequently, documents (accompanied by meetings) are constantly revised and scrutinised to attain administrative security and order.
The interview data suggest that signs of disorder are viewed in terms of (1) a lack of rationality and (2) a lack of implementation of administrative tasks.In complex policing partnerships consisting of multiple levels, disorder is likely to emerge with respect to the structures and order imposed from the top down (i.e. the priorities, the aim/goal of partnership working, and the evaluation and assessment of partnership working).For example, as already mentioned, 'the smorgasbord approach' fuelled bewilderment among as professionals 'on the ground', who failed to understand how the priorities had been selected and what was the connection between them: The shared priorities are just a hodgepodge of geographical areas, target groups, and methods, which makes it difficult.The chain of governance is difficult.(Interview 3, city employee) Another sign of disorder was connected to the inherent problems of gauging outcomes in policing networks (see, for example, Fleming and Rhodes 2005).Several professionals expressed their concern about the prospects of measuring the results and effects of their efforts.In part, this experience was based on the feeling that the top tier had failed to administer and produce measurable and thus rational standards to relate the activities to: 'The central level, including the central steering group and central planning group haven't delivered goals and indicators, and it's been two and a half years now' (Interview 2, police officer).At an informal discussion, another police officer echoed the same sentiment: 'Ask members of the central steering group how we're doing?I don't think they have an answer' (Stenström, Fieldnote 2022).The main problem, reported by several participants, was the lack of a 'thorough line' between goals, activities, and measurement.This harks back to the general notion that the administration produced by the managerial level was not up to par with local demands.And following the logic of Eigendynamik, this notion operated as a driver for new administrative tasks, such as meetings and conferences where the topic was goals and measurement of partnership working.
Signs of disorder also emerge in the everyday interaction between local actors.Importantly, there were examples of breaches of the informal norm that, in partnership working, all participants must contribute (cf.Stenström et al. 2014).Exercising freedom by, for example, avoiding meetings or failing to hold project meetings resulted in efforts to promote order and discipline via administrative action.It must be noted that the norm that everyone should contribute was largely experienced by the participants as a requirement to be physically present at meetings.At an interview, a city employee noted that partnership working requires everyone's contribution since 'we cannot have a situation where only because you are dissatisfied with one group you don't attend that group and choose to meet with this group at another meeting instead' (Interview 3, city employee).The interviews also suggest that the lack of motivation manifested in the avoidance of meetings was linked to the above-mentioned feeling that there is a lack of order or clarity concerning aims and goals: Importantly, deviance (e.g.not attending meetings) was not exclusively managed as disorder.It is well-documented that the police often dominate partnership working (Jones andNewburn 1998, Crawford 1999) by, for instance, shaping priorities and the agenda.Similarly, the research reported here suggests that police officers enjoyed more freedom compared to, for instance, social workers.In essence, police officers' breach of the partnership order did not foster new administrative circles.Specifically, the institutionalised norm that professionals should attend meetings was often neutralised as it was a common understanding that the police have other priorities, such as responding to violent crime and attending public events.However, it was also evident in less dramatised situations; for example, the local police and representatives from the City regularity conducted 'safety walks', which was a collective effort to identify risks and sources of unsafety in a limited geographical space.The following passage is from a safety walk in an industrial area that is currently being rebuilt into a residential area and gentrified.The main point here is that the participants accepted and even expected the police to give priority to extraneous events (Delpeuch and Ross 2020).
We arrived at an open place within the industrial area.'The light is switched on at the top floor', said one representative.'It is our tenant.They're allowed to be there.But the awning is in bad shape', said the landlord.'Okay, those are awnings', said one police officer.While the rest of the group moved along the building, the landlord climbed onto a pallet, opened a door, and entered the building.A police officer scrambled and followed the landlord.A man wearing a yellow jacket joined them.From the outside, we could see them advance up the staircase.The rest of the group continued the safety walk […].'Where's the police?' the coordinator asked us.'They found something they had to attend to.They had to "manage' a person"', said one participant.Further down the street, I observe a police buss heading towards the building.'They were at least here at the beginning of the walk.That's always the case.They must make priorities.They're rarely with us the whole time', said one participant when the safety walk was completed.(Stenström, Fieldnotes, 2022) Indeed, crime and urban safety are complex problems that necessitate a holistic approach.A positive outcome of a partnership approach to policing requires both structure and innovation.However, as has been argued throughout this article, the dynamic between these forces will inevitably drive administrative growth.The challenge for professionals and policymakers is to avoid the risk of promoting ritualistic and unnecessary administration that will be of little use both for the public and for the police and its partners.

Conclusion
The shift from a model in which the police were responsible for crime prevention to a model characterised by partnerships and networks is well-documented (see, for example, Johnston 1998, Garland 2001).One expression of this transition is that partnership policing has become a widely accepted approach to crime prevention, referred to by Crawford and Cunningham (2015) as a 'mantra' that permeate the discourse of crime prevention across jurisdictions (see, for example, Fleming 2006).However, there is limited knowledge about the actual practices of policing partnerships and policing networks (cf.Quéro and Dupont 2019).Based on an empirical study of the everyday experiences of actors involved in the Stockholm partnership, this article contends that partnership policing is mainly an administrative exercise whereby the focus of partnership policing revolves around the management and oversight of inter-and intra-organizational relationships.Specifically, this article shows that one crucial dimension of the Stockholm partnership is administrative spirals of meetings, meetings about meetings, the production of documents and revisions of existing documents.Consequently, the particulars of administration take precedence over the wider advantages of partnership policing, such as maintaining security, reducing crime, or harnessing other positive effects.Moreover, administrative processes are driven by participants' need for order, and the interplay between order and disorder propels the growth of the administration in partnership policing.This self-referential force of administration, then, operates within and beyond the framework determined by local authorities and the police, and plays a fundamental role in creating and sustaining the notion that there, in fact, is a partnership.However, while administrative processes are crucial in communicating that partnership work is being performed (cf.Åkerström et al. 2021), the article also shows that there is a disjunction between the idea of partnership policing and partnership practices.Although the idea of the Stockholm partnership is to reduce the level of crime and enhance urban safety, actual practices seek to administer, organise, and manage the relationship between the police and their partners (cf.Hörnqvist 2001).
Effective partnership relationships are characterised by mutual goals, joint priorities, reciprocity, and trust (see, for example, Fleming 2006, Crawford and Cunningham 2015, Crawford and L'Hoiry 2019).However, as this article makes clear, deciding mutual goals or building inter-organisational trust necessitates administration.Police and its partners must meet and produce documents, which in turn generates more meetings, instructions, and revised documentation.Hence, the creation of successful partnerships is fraught with challenges since these tasks may cause frustration among the actors on the ground.If not managed properly, ritual administration may cause the actors to lose faith in partnership working.Recent changes in public administration (i.e.New Public Management and Audit Society) are similar in many Western countries, and it can be reasonably assumed that the administrative growth outlined in this article applies to policing beyond the Swedish context.
It should be noted that in police culture, administrative tasks are traditionally defined as administration as 'shit work' (Van Maanen 2010).Nevertheless, many police officers did not express their outwards negativity towards these tasks.Rather, as reported elsewhere (Åkerström et al. 2020), meetings and documents are defined as essential practices.Perhaps, in a context where partnership policing has become the accepted way of preventing crime, expertise about 'how to execute collaboration' is being formed instead of effective and meaningful crime prevention strategies (see, for example, Alstam and Forkby 2022).While police and local authorities have undoubtedly become more professional in managing collaborative relationships, whether this creates effectiveness and efficiency in achieving the desired results remains highly questionable.In sum, this highlights how changes in public administration permeate the cultural and professional norms shared by actors responsible for the implementation of policing strategies.
Finally, this article draws attention to the theoretical utility of considering the Eigendynamik of administration in unpacking the internal mechanisms and dynamics of policing and security networks.Whether a network aims to solve problems, exchange information, or coordinate resources, such efforts connect to and initiate circles of administration.The regulatory transition from command-and-control models to networks, negotiations, and agreements (Fleming and Rhodes 2005) inevitably fuels administrative spirals, as stakeholders must find ways to create trust, reciprocity, and consensus on what joint action to take, how it should be done, and who should do it.Hence, models seeking to make sense of the governance of policing and security networks would benefit from considering how the Eigendynamik of administration structures, facilitates, and shapes network practices.

Table 1 .
Research participants.template).Third, the interviewee notes how conferences (meetings) are connected to the production and revisions of documents.Fourth, we can see that administrative disorder (documents that vary in length, vague instructions) is presented as issues to be corrected in the future.In essence, the quotation highlights how the Eigendynamik of administration permeate and shape partnership policing practices.
. 'Juvenile delinquency' .'The school as a crime scene' .'Violent extremism' .'Unlawful influence' .'Illegal settlements' They [members of a project group] don't see what they bring to the table.'I don't understand this at all!' It's like that.'I don't want to participate because I don't understand my role.I don't understand what I can contribute with.' (Interview 10, police officer)