The police and ‘the balance’—managing the workload within Swedish investigation units

Professionals within street-level organizations are essential for the delivery of public services to citizens. However, among a number of difﬁcult dilemmas, they have to deal with an extensive workload. The police can be seen as a good example of this; they are expected to solve most crimes, including the so-called mass crimes and the more spectacular cases that make it into media headlines, and often on a continually decreasing budget. A key regulating mechanism for investigation departments in the Swedish police is the so-called balance. The balance can be described as a basket in which they put the cases that there is a desire and potential to work on but not in the immediate term. The purpose of this article is to analyse the balance as a way of rationing the workload within the Swedish police. Working with the balance consists of two processes: limiting and buffering the workload. Limiting is the practice of reducing the work in a situation. Buffering is the process of putting some work on hold to deal with later, of which the article identiﬁes ﬁve kinds; functional, problematic, quasi, progressive , and symbolic buffering. The exploration of ‘the balance’ contributes to our understanding of how street level organizations attempt to defend their professional jurisdictions, their well-being, and their ability to complete their duties.

balance is a complex concept with different meanings. The term 'the balance' originates from the Swedish word 'balansen' and has the same meaning as in English in which it refers to an equilibrium of some kind. In the context of the police, it is sometimes used to refer to all cases regardless of whether they are worked, and at other times only to inactive cases. Questions about the balance of a specific unit can refer to all open cases. However, at other times it can refer to inactive cases, for example: 'I am going to put that case in the balance', meaning that the police will leave it inactivate for some time. In this article, if the balance includes both active and inactive cases it will be referred to as the 'general balance' and only inactive cases as the 'balance'. So, in the balance there will be cases on which there is a desire and potential to work albeit not in the immediate term. Cases are put in the balance that are on hold, waiting for someone to actively take on the investigation. This means that the cases in the balance have not been terminated because the suspected crime is severe enough to investigate and there are leads that can be followed up, for example witnesses to talk to or technical evidence to collect, such as footage from surveillance cameras. On the other hand, cases in the balance are not the highest priority; if they were, they would be dealt with right away.
The purpose of this article was to describe and analyse the balance as a way of rationing the workload of the Swedish police. How is rationing done in a street-level organization? The results show that rationing consists of two kinds of processes, limiting and buffering. The analysis identifies five kinds of buffering; functional, problematic, quasi, progressive, and symbolic buffering. The article is based on a case study of the Swedish police and includes individual and focus group interviews, observations, and document analysis. There are several types of police work including intelligence, patrol, and others, but the focus of this study is on investigations.
The police can be seen as a good example of a street-level organization that has to ration its services, as it has to face dilemmas such as, on the one hand, being expected to solve more headlinegrabbing crimes and, on the other, the so-called mass crimes, often on a continually decreasing budget. The police, as well as other public services, are under pressure to improve services (Gavin 2011). These complexities have led researchers to describe the police as having impossible working tasks or even impossible jobs (Morrell and Currie 2015). However, it is not only about rationing the services but also an important aspect for the police to keep/increase their legitimacy. In practice, it means that they need to keep a high level of procedural justice, treating people with dignity and respect and giving them a chance to talk about their experiences (Worden and McLean 2017).
There are several reasons why it is important to study work overload within street-level organizations. First, it is important to understand how these organizations work, because they can potentially have a significant impact on people's lives. It is important for citizens to understand these processes, both in general and more specifically, as they determine who does and does not receive help (Maynard-Moody and Musheno 2003). Secondly, these processes have a significant impact on the employees of the organizations. High-stress work has been shown to be related to poor mental health in, among other professions, policing (Hansson et al. 2017). As Brodkin (2011) puts it, understanding street-level organizations begins with understanding '. . . [their] underlying ambiguities and inconsistencies.' (7). The exploration of 'the balance' contributes to our understanding of how street-level organizations attempt to defend their professional jurisdictions, their well-being, and their ability to complete their duties.

R E O R G A N I Z A T I O N S , O R G A N I Z E D C R I M E , A N D B E I N G A T T H E C E N T R E O F P U B L I C A T T E N T I O N
The overarching aim of the police can be described as being to 'help and protect people from crime' (Polisen 2020: 2), and one of the more concrete working tasks is to investigate suspected crimes, though the work also contains tasks such as maintaining order, providing advice, issuing citations, answering concern for welfare calls, searching for missing persons, appearing in court, and more. The last 10 years have been a turbulent time for the Swedish police, with them undergoing one of the biggest reorganizations of any Swedish public body and the general workload increasing due to a high number of violent crimes within the context of organized crime, at the same time as being at the centre of public (including political) attention. The reorganization was part of more wide-ranging changes of the police in Western and Northern Europe in the same era (Fyfe, Terpstra and Tops 2013). In the Swedish case, the changes emanated from at least two sources, one of which was general criticism of the state. The criticism came from a state-initiated investigation by The Committee on Public Administration (2008), which had the task of reviewing the organization of the entire state administration. One of the aims was that the state in general needed to put '. . .in place a clearer and more easily grasped structure for state administration in various respects. . . ' (17). This general criticism was directed at, among others, the police, which had a decentralized organizational model under which working arrangements were set regionally. This meant that the working tasks, titles, and organizational structure varied from area to area (See Björk 2018 and Ivarsson Westerberg 2020 for a more in-depth description and analysis of the reform).
In addition to this more general criticism, there were reports of a continuous fall in the rate of cleared cases, despite increased resources, and the media and Attorney General Beatrice Ask started to question what was happening within the police (Sköld and Ekelund 2012). All this led to the initiation of a total reorganization in 2015. An entirely new organization was set and all the employees had to reapply for a position within the new organization. The old organization was replaced by a central national administration (Statskontoret 2018). The reorganization can be seen as part of the changes that are referred to as New Public Management (NPM), with the goal to increase transparency, discretion, and accountability for the managers with more focus on measurable aspects of the police work, like, for example, clearance rates (Baradaran Baughman Forthcoming). The aim of the reorganization can be described as a centralization of the organizational form and decentralization of decisionmaking. Whether this reorganization has been successful has been widely debated (Björk 2018;Hansson 2018;).
The Swedish Police Authority is split into seven regions, each one containing a number of police districts (27 in all, The Swedish Police Authority 2020). Each of these districts is organized with one department for major crimes and one for crimes in close relations, a call centre, and two departments for out-of-hours work. In the interviews, the out-of-hours departments were described as the 'emergency investigators'. In addition, all police districts have a number of local police districts whose responsibilities include police officers patrolling the streets and investigations of the so-called mass crimes, such as thefts. The title of the leading police officer of a department differs, and may be Detective Chief Superintendent if it is a local police district or Superintendent if it is a major crime department. The front-line investigators have the title Detective Investigator and report to a Detective Inspector. Every department also has a number of Prime Investigators who are the highest in command of the cases the department investigates. The Prime Investigators are not responsible for staff, only for cases. As will be shown later in the article, Prime Investigators hold a key position in regulating the workload of a department.
A widely discussed aspect of the workload within the Swedish police is the number of violent crimes within the context of organized crime. For example, the number of cases in which violence has led to deaths tripled between 2012 and 2017, from 11 to 34 cases (BRÅ 2020). In addition, there are a large number of cases in which the victim survived. In particular, several shootings and killings of bystanders have been at the core of the debate, such as a 12-year-old child in August 2020 (Jamshidi, Stigfur and Månsson 2020). Every time these kinds of severe crimes are investigated, the Swedish police set up a special project group using a methodology called PUG investigation with recommendations on how they should work with violent crimes (Polismyndigheten 2017). PUG is a Swedish acronym and stands for "the police methodological support for the investigation of serious violent crimes). Every PUG investigation means that at least 10 experienced investigators are removed from their usual positions to be included in the PUG investigation, but depending on the complexity of the case, it can be considerably more. The increased number of violent crimes within organized crimes has put huge pressure on the police organization in terms of workload. The debate and the police situation have led to political consensus that more resources must be given to the police, and in 2017, the Swedish Minister of Internal Affairs promised to increase the number of employees in the police force by 10,000 (from 30,000) and to establish two new police academies (Jacobson 2017). During 2018-9, the two new police academies were opened: one in Borås (western region) and one in Malmö (southern region). The The police and 'the balance' 3 centralization and homogenization of the regional structure also increased the possibilities of governance and control in line with the more general trend of NPM, including processes such as increased transparency, and greater focus on measurable results, costs, auditing, and control as a means to a higher degree of responsibility (Davis and Bailey 2018;Bergman and Karp 2020).
In some aspects, the Swedish police have some of the traditional characteristics of a semi-profession; they have to apply a combination of theoretical and abstract knowledge with more practical skills. They also have to balance explicit and more tacit knowledge, just as teachers, social workers, and nurses. Like these semi-professions, the Swedish police academies are placed in the Swedish university system. However, they may not have the same status as some of the more classic professions. In terms of jurisdiction, the police can be seen as either very strong or highly contested. They can be seen as strong, as no other professional group makes any successful claims of doing the same work as the police. However, internally the investigators are divided into two main occupational groups: the traditional police with exams from a police academy and the socalled civilian investigators who should have an academic degree in a subject such as law, sociology, psychology, or social work. Civilian investigators are allowed to do most things in an investigation, but there are some exceptions, for example they are not allowed to carry arms or use force. Previous research has described the relationship between these groups as complex (Rice 2020). In the interviews, some investigators who had passed the police academy exams expressed worries that they were going to be reduced to 'walking the streets', while investigators with more academic training would take over responsibility for investigations.
Professionalism as a way of organizing governance is based on giving occupational groups the trust and discretion to negotiate a division of labour based primarily on scientific knowledge. In the debate on possible threats to professionalism, several candidates have been put forward, such as bureaucracy (Larson 1977;Evetts 2014;Björk 2018;Jacobsson, Wallinder and Sein 2020), marketization (Drummond et al. 2000;Freidson 2001), and layperson governance (Liljegren, Höjer and Forkby 2014) or politics (Lacoe and Stein 2018). Added to that is a heavy workload. Having to handle too great a workload reduces discretion, and part of the solution to achieving a reasonable situation is individual or organizational standardization (Tummers 2017). By exploring how the balance works, this paper provides an investigation of how street-level organizations use various organizational means to defend themselves and their professional discretion.
The police can in many ways be seen as a good example of a public welfare organization that has a tendency for being overstretched, as large amount of resources could be spent on every unsolved crime. The police cannot go all in on every crime. Not all crimes can be treated like the murder of Swedish Prime Minister Olof Palme in 1986, a case on which massive amounts of resources were spent to solve the murder without anyone being convicted in a higher court. The case was closed in 2020 because the suspected killer had died (Löfgren 2020).
The Prime Investigator has a key role in managing the workload of all investigation departments. It is a centralized position that all departments have. As one of the Prime Investigators interviewed explained: 'all cases that come to our department must pass me or my college'. The general workflow of an investigating police department starts with the professional judgement of the Prime Investigator, who must decide if a case is (1) within the department's jurisdiction, (2) investigable, and (3) should be investigated. The first question involves many aspects including checking if the crime is correctly labelled and reading the basic information that comes with all cases. The basis for determining which department should investigate a crime is the so-called 'catalogue of crime', informally referred to as the 'bible of crime', in which all types of crimes stated in the law are given an organizational belonging. In addition, if there are opened and ongoing investigations elsewhere involving the prime suspect the case will be passed on.
All investigation departments have at least three kinds of priorities that are either formal or informal. The first is based on the penalty scale, the kind of punishment a specific crime may receive (e.g. imprisonment or not) and the length of the punishment (e.g. a long or short time in prison). The second category is a result of the organizational structure, for example all departments prioritize suspects who have been taken into custody as they are then subject to a short time limit before evidence has to be presented to the court. As much as possible is done to clarify if a suspect in custody is guilty or not. On the one hand, the police want to avoid innocent persons spending time detained and on the other guilty persons being released on a lack of evidence. Other aspects of prioritization in this category include certain time limits that apply to crimes involving children.
If the previous two categories can be seen as formal limitations on how the police prioritize there is also an informal one. This third category may include personal or group agreements that are more ambitious for vulnerable people; examples from the case study all involved 'weak' groups such as children, the elderly, and small-scale businesses. This category is also based on the police officer's discretion and is subject to personal priorities, which also relate to the kinds of investigations the department carries out. In major crimes, murders are prioritized over rape and crimes committed by persons in public services. Rape investigations are generally a priority but less so than murders and attempted murders. For local police departments, cases that involve assaults are high priority whereas theft is an example of a crime with a lower priority.
Street-level organizations One perspective on street-level organizations is the context in which professional work is embedded. Professional work is not conducted in a vacuum; on the contrary, street-level organizations affect how, when, and where professional work is done. Streetlevel organizations can be seen as mediators between the policies and practices of the state and the public (Brodkin 2011). Street-level organizations turn abstract policies into practical action. In that way they can be seen as negotiators between the interests of the state and the citizens. Doing so they face a difficult task: the challenges have been described as vague, ambiguous, and conflicting goals (Mintzberg 1999;Hasenfeldt 2010;Brodkin 2011;Gofen 2014). In the ongoing negotiation between the state and the public, street-level bureaucrats do not only offer a neutral implementation of public policy but also change it to make their workload more manageable, among other things. Policy has to be negotiated with regard to practical constraints related to organizational and other limitations (Hill and Hupe 2007;Liljegren, Höjer andForkby, 2014, 2018). These practical constraints may include time, access to relevant services, and knowledge. In addition, street-level organizations have a strong tendency to face resource shortfalls (Hupe and Buffat 2014;Tummers 2017). As Noordegraaf (2007: 781) stated: 'Cases, clients, costs, and capacities interact in multifaceted ways'.
The theoretical issue is that street-level bureaucracies have a dilemma: they have a strong tendency to be overloaded with work, and none of the traditional ways of solving that problem, such as being more efficient or improving the quality of services, work. So, what is workload? We define workload as '. . .all activities involving employees' time spent on performing professional duties, responsibilities and interests at work, either directly or indirectly.' (Johari, Tan and Zukarnain 2018: 110). Lipsky claims that streetlevel organizations are trapped in a negative cycle in which more successful programmes tend to create a work overload that may be difficult to deal with, whereas unsuccessful ones may be 'rewarded' with a more reasonable workload. Demand tends to meet supply, especially if the service is perceived as being high quality. Increased supply results in both increased expectations of quality from existing service users and increased demand from new users. Streetlevel organizations are highly adaptive to their environment. According to Lipsky (1980), strategies to manage workloads include queuing, reducing service preparedness, increasing bureaucratic systems, or creating barriers to communication. Increased demand and workload complicate the situation (Fader and Dum 2013;Hupe and Buffat 2014). Lipsky (1980: 2) defines the street-level bureaucrat as: '. . .a public employee whose work is characterized by the following three conditions: 1) He is called upon to interact constantly with citizens in the regular course of his job. 2) Although he works within a bureaucratic structure, his independence on the job is fairly extensive.
3) The potential impact on citizens with whom he deals is fairly extensive', although it should be noted that most public employees within the Swedish context are women (Johansson, Dellgran and Höjer 2015). A key aspect of work is that professionals must have discretion to The police and 'the balance' 5 exercise their knowledge (Freidson 2001). Discretion exists when street-level bureaucrats have the flexibility to choose between several options and the police have considerable discretion (Buvik 2016;Johannessen 2019). Street-level bureaucrats interact directly with citizens, while personal discretion gives the police the liberty to make appropriate decisions in many complex life or death situations, but it can also be misused (Beckett 2016; Lehmann Nielsen 2016). Zacka (2017: 14) claims that discretion '. . .raises the spectre of arbitrary treatment, personal domination, bias and corruption.' Studies of discretional aspects of policing include how organizational characteristics affect discretion (Jeffrey 2011), while others describe and analyse factors within specific situations (Fagan and Geller 2015). If discretion can be seen as a part of any professional practice, rule following can be seen as a part of organizational socialization (Oberfield 2010). We know from previous studies that personal beliefs, norms, and values affect decisions at micro-level (Lipsky 1980;Tummers et al. 2015). Discretion also gives street-level bureaucrats the ability to make decisions to manage their workload.

Methodology-the case and case studies
The studied organization is a police district situated in a major town in Sweden. It was chosen for several reasons. First, it was large enough to include both rural and suburban areas with a range of conditions with regard to social problems and crime. Secondly, interest in the research questions was shown by the police district. Thirdly, access was provided to the organization.
Case studies typically involve one or a small number of cases approached with several qualitative research methods that can include interviews, observations, and studies of different types of documents (Swanborn 2010). Stake (1995) distinguishes between several types of case studies. This study can be seen as an instrumental case study '. . .where the case is chosen to gain insight or understanding into something else. . .' and not primarily to understand the case per se. Two main reasons for choosing this case study design often mentioned in the literature are (1) to capture the complexity of the phenomenon to be studied and (2) because it is difficult to separate the social process being studied from its context (Yin 2003;Swanborn 2010). Both these reasons were behind the choice of methodology. In more practical terms, the researchers did not know where within the police work that capacity to limit the work was situated and for that reason this kind of broad data collection strategy was chosen.
The empirical basis for the article includes 25 individual interviews that can be divided into 5 roles: regional police commissioner (1 person), persons in a strategic role within the studied police district (13 persons), Prime Investigator (6 persons), and police inspector (5 persons). In addition, we conducted 2 focus groups with a total of 16 police investigators, 5 observations (major crimes and a local police district), an estimated 50 documents, and numerous informal talks on the way to and from these events. The data were collected during 2017. The main data source was interviews (rather than observations), as the distribution of cases was mostly on an individual basis and more informal rather than in the context of formal group meetings. Documents were used mainly to understand the new organizational structure of the police and some of the work procedures referred to. Observations were conducted to identify processes and settings through which the workload was distributed to street-level investigators. Five 'morning meetings' were attended on Mondays ('when most things happen') and midweek. In one case, a specific meeting was attended to validate a claim made in the course of the interviews that although 'case distribution' meetings were held, no cases were distributed there. The researchers were curious to verify whether this really was the case. As it turned out, the interviewee was correct and no cases were distributed there. No explanation for the discrepancy between the name and content of the meetings could be given. The individual interviews were all semi-structured, based on a thematic structure with the main topic being work overload and how it was dealt with. The first to be interviewed were the police commissioners: the heads for each of the seven departments within the studied police district. Most of the observations were also conducted during this time. A thematic structure was used for the observations based on Patton (1987) and includes themes such as physical and mental environment, planned and unplanned activities, and verbal and non-verbal communication. It was in this phase that the researchers recognized the importance of Prime Investigators, and six persons in this role were interviewed. To gain a more comprehensive understanding of their work, an active observation was conducted in which one of the Prime Investigators went through the inbox explaining each case and related decision to the observer. After interviewing the Prime Investigators, the next step was the senior investigators, all three of whom were interviewed. In the next phase, two focus groups were conducted with street-level investigators. Focus groups are recommended when a researcher has limited knowledge of the setting being studied (Wibeck 2000). To round up the data collection, the two most senior police commissioners in the studied organization were interviewed in addition to the regional police commissioner (which is the next level in the hierarchy). All interviews were recorded on a voice recorder. The researchers listened back to the interviews and transcribed the relevant parts verbatim. Some parts of the interviews were characterized as '. . .explaining the police and the context', and as this was not the focus of the research questions it was decided that these should not be transcribed. Given the research design, no empirical generalizability can be claimed, at best the results may lead to 'concept generalisation', by which the concepts developed in this study may be useful in other contexts.
The first stage of the analysis focused on identifying when and where the workload was limited so data could be collected in the right settings. The more focused analytical process started with the development of a thematic structure with the key mechanisms that could function as a way of rationing the workload (Braun and Clark 2012). A key feature that the analysis has focused on is key metaphors, as metaphors '. . .are a fundamental scheme by which people conceptualize the world.' (Gibbs 2008: 3). It can be noted that the balance is a metaphoric concept. The concept of 'the balance' came from the police themselves and was widely used. Everyone knew what the concept was about and understood and used it. It was part of the culture of the organization that had to be understood for someone to function within it. The different types of balance are a construct from the researchers' analytical work, where a closer analysis of the phenomenon revealed that despite being referred to as one concept, in reality it could be divided into several subcategories. The analytical approach can be described as an abductive process in which the analysis has gone from theory to data and back with the goal of making a contribution to the theoretical understanding of the phenomenon in focus (Alvesson and Sköldberg 2000). To describe the data, we have used a combination of long quotes of what the informants said, with their context, and short quotes in which only the keywords have been used. The purpose of the short quotes was to document the 'conceptualization of a target phenomenon' (Sandelowski 1994: 480), and the purpose of the long quotes was to give a more detailed description of what is studied (Geertz 1973). The project has been approved by a committee for research ethics. Among a number of precautions the project made to ensure the confidentiality of victims, staff and suspects we did not use a voice recorder during the observation where personal details by necessity had to be discussed. By doing so the project avoided keeping any records of the most sensitive material.

Why have a balance and how does it work?
The importance of the balance was indicated early in the case study by a member of the group of strategic managers who said: 'We do not talk that much about work overload, we talk about "the balance." The balance is mentioned in almost every meeting.' So, what is the balance? All incoming cases to an investigation department must be categorized in one of three 'boxes' (Table 1).
At one end of the scale, cases are not accepted into the department's investigation process at all (more on this later), and at the other they are seen as high priorities that must be dealt with right away and no time is wasted in giving them to an investigator to start an investigation. All cases in which a suspect has been taken into custody, and prioritized cases such as murders, aggravated assaults, suspected crimes against a vulnerable group might fall into this category. In between are all the cases that are put on hold within the system. In the second box, cases are put that are treated with a 'wait and see what happens' approach. This is what is known as the balance.
The police and 'the balance' 7 The reasons behind a 'wait and see' approach can be multiple (more on this below).
The balance is an important way to measure the workload of a department, and a balance consisting of several hundred inactive cases is seen as a major problem that has to be dealt with, while a balance with only a small number of cases is described as a sign of a 'healthy' department with a reasonable workload and that is 'on track'. Having a small or no balance is a situation described as: '. . .taking the cases straight from the tap.' In practical terms, that means no cases are put in the balance because of problems related to a lack of resources, such as the number of investigators. The kinds of cases that were put in the balance varied depending on the kind of department the Prime Investigator worked in. For a local police department that investigates small crimes, they may be a burglary, for major crimes, which among other things deal with murders, a case for the balance may be a rape.

Limiting and buffering
Working with the balance involved two different processes: limiting and buffering. Limiting is simply the practice of reducing the workload in a given situation, an example being closing a case due to lack of evidence or referring it to another department so it is neither active nor passive. However, not all crimes that cannot be dealt with right away can be closed. If a crime is severe enough, closing a case may not be possible. In those cases, a better solution may be to put them on hold, that is to buffer them. The concept of buffering has been borrowed from computer science where it is defined as 'pre-loading data', so the data can be accessed more quickly when one of the computer's processing units needs it (Williams 2013: 1). The same process seems to apply to workload. In order for an organization to run smoothly some cases have to be buffered so they are accessible if/when a processing unit (read investigator) is free to start the investigation. One finding in the study was that street-level investigators did not buffer any cases; they were expected to investigate all the cases they had. If a senior investigator wanted to prioritize a case, others were taken away from the investigator's workload. There was only very limited buffering going on at street level. As one investigator put it: 'If you have a good relation and reasonable communication skills you cannot burn out in this organisation.' By this, she meant that as long as the investigators can tell their superiors what their situation is they cannot burn out, as they should not have inactive cases.

The balance and quasi-balance
Despite interviewees referring to the balance as a single concept, either as the general balance, all opened cases, or the balance, only inactive cases, it seems that it can be divided in four subcategories: (1) the quasi-balance; (2) the progressive balance; (3) the symbolic balance; and (4) the real balance. The quasibalance refers to cases that have not been given to an investigator because the Prime Investigator has decided there is no further action to be taken on it at the current time.
Some cases are just waiting in my drawer for something to happen. In their present state there is nothing we can do. But if some technical lead turns out right, we may have a case. And if it doesn't, I will just close it due to lack of evidence.
In practice, it can be that cases are put in the balance that are waiting for some technical evaluation, such as a screening for DNA or a technical examination of a weapon being undertaken by an expert. The quasi-balance is not seen as a major problem for a department because it has already done everything it can on the case. Cases that have become so old that no one can be prosecuted because a statute of limitations has passed are also part of the quasi-balance. Although these cases have not been formally closed, they are closed in practice because nobody could be convicted. The police try to avoid this happening as it is considered shameful. According to one

Denied entry
On hold (balance) Immediate investigation interviewee, it is seen as better to close (or 'kill') a case before the time limit expires as this does not signal a lack of skills or resources in such an obvious way. Allowing the time to expire on a case is seen as a sign of passivity, as opposed to the active option of closing it themselves.
The progressive balance There is another category of cases that are not being actively investigated and are put in the balance; this can be described as the progressive balance. These are cases that are put in the balance to see how they progress. Doing too much on them too early is seen as a bad investment of investigative resources. A recurring opinion was that it is better for some cases to be put on hold before any action is taken. One example was violence between group of friends that was considered to be non-severe: You do not want to call them while emotions are still running high; it is better to wait and call them after a week or two.
The Prime Investigator argued that in minor cases, such as this one, charges are often dropped once emotions have subsided, so actions taken in the early phase of the investigation are at high risk of being wasted time. Cases can also be put in the progressive balance because the Prime Investigator wants to see if or how they progress. A case that pertains to a verbal threat may be put in the balance because it is not enough to start an investigation, but if the threats continue it may become an active case. Instead of terminating the case right away, it is put in the balance to wait and see what happens. If there are no new developments, some weeks/months later the case is terminated.

The symbolic balance
Another subcategory of the balance is the symbolic balance. These are cases that could be closed right away but to which there may be other benefits of waiting and closing at a later date. The symbolic balance includes cases that a Prime Investigator can see right away will have no further investigative actions taken, but that he or she avoids terminating right away out of what is described as concern for the victim. One example of this is housebreaking without any obvious evidence; terminating such an investigation right away is seen as disrespectful. One Prime Investigator explains: In addition, all home burglaries are under my jurisdiction. In those cases, I hold on to them for a month or so out of respect for the victim, but in most cases I can see right away that we won't convict anyone. If we don't take these criminals in the act it is very, very hard to convict anyone. And in those few cases when we do catch them they have already done so much that one more additional burglary won't mean anything anyway.
In the cases described above, it is seen as a better option to wait and terminate them a couple of weeks down the line. It should be noted that there are no formal constraints that prevent a Prime Investigator from closing a case right away and reopening it later if there is a reason to do so. From the perspective of reducing workload, this may also be seen as a limiting strategy to reduce the risk of, for example, complaints and demands for further action compared with closing the case at first sight. This can also be seen as an example of impression management to please citizens, not giving them reasons to be upset. In addition, terminating a case a couple of weeks later may give the impression that more has been done than is really the case. As such a symbolic balance is kept that is directed towards the citizens. Another example of symbolic balance is directed internally towards the organization. A Prime Investigator was asked: Interviewer -Couldn't it be problem if the balance gets too low? Prime Investigator -That is not a problem in urban areas, we always have things to do and the openness of the case system makes it impossible to fake a balance. But in some rural areas they may be centralised if they can't show that they have things to do and those areas are also harder to control.
Here, the Prime Investigator (from an urban district) describes how rural police districts may have The police and 'the balance' 9 both a reason and an opportunity to keep a symbolic balance. A symbolic balance may be kept in relation to both citizens and the organization, giving other members the impression that more is happening than is the reality.
It should be noted that the 'let it wait' policy only applies to crimes that are not seen as severe. Cases that the Prime Investigator regards as a priority are subject to a different procedure. In those cases, the incoming e-file is printed and the Prime Investigator takes it directly to a first-line officer who sees to it that investigative actions are taken, either personally or by giving the case to an investigator under his or her management.
The real, functional, and problematic balance The real balance contains cases that are put on hold because there are no investigators to give the case to at that moment. These cases are not a high priority but are seen as worth investigating further. This is also the buffer zone of the workload. If at a later stage a balance has built up that is considered too high, these are the cases that would be scrutinized to see if they are really worth investigating. However, the real balance consists of at least two categories of cases: the functional and problematic balances. The functional balance means just what the name implies: the cases that Prime Investigators need for the organization to run more effectively. One Prime Investigator explained: I have to keep some cases in my drawer so I have something to give to people [investigators] when they come and ask for more. And that has to be a reasonable amount as the investigators tend to come and ask for more cases at the same time. I don't want to be in a situation where I have nothing to give them. Not all cases in the balance are a problem.
Keeping a functional balance is a way to supply cases that are still worth looking into to all investigators, rather than suddenly having no cases for them to work on. At some stage, the number of cases becomes problematic, but the Prime Investigators' had no exact number at which this was the case. The problematic balance occurs when cases cannot be investigated that, according to professional judgement, are worth investigating. There are several reasons to describe a case as 'worth investigating', it could be that it is seen as severe enough, that the prime suspect belongs to a group that the police wants to keep close attention on (read groups that are considered to commit most crimes) or that the victim belongs to a group that is considered 'vulnerable' such as children or the elderly. In addition, there should be an action worth taking in the case, a lead to follow up on or someone who may be able to provide information. The problematic balance is cases that the department wants to give to an investigator but cannot.

The balance as buffering and limiting
During the study, the balance seemed at first to represent mainly a buffering exercise, but it turned out to be more than that. Putting cases in the balance was also shown to be a way to close cases. The balance was described as a 'compost' where cases were left to 'rot' or 'die'. One Prime Investigator explained: Formally we can't close cases due to a lack of resources. But when the case gets old enough, we can do that because it is no longer possible to investigate the case. Until that happens we put the case in the balance.//What we put in the balance will be easier to close later because the evidence will become weaker. Because if you start the investigation a year later you don't have the same options as earlier and for that reason my opinion about the case might change. Maybe we should have talked to some witnesses? Can he identify the driver one year later? One year later that is much more difficult. And the penalty value will decrease as time passes and that makes it easier for me to close the case. And a suspect might have been convicted for something else, and if I think that the crime would not lead to a higher penalty, I can also close the case.
In this way, cases die in the balance. There are at least four ways in which this happens. First, evidence disappears, for example digital evidence that it may not be possible to collect at a later stage. Secondly, evidence becomes weaker; witnesses may not be able to identify someone a few months after an incident.
Thirdly, the value of a crime, in terms of the penalty it may receive, reduces as times goes on, which also changes the priorities for the police as more severe punishments mean higher priority (and vice versa). Lastly, if a suspect has been convicted of other crimes, the Prime Investigator can close a case if he or she thinks that the added offence will not lead to higher penalties.
Reducing the balance From time to time, when the balance has grown and become a problem, something has to be done. One Prime Investigator explained that over time there is a tendency for the balance to build up, and they had to do what was described as 'case care'.
When the balance builds up, some of the senior staff meet up and go through the balance. The goal is to close as many cases as possible but also to sort out the cases that we can't kill.
The purpose of these meetings is to 'kill' or 'close' cases to make the balance more manageable again. Even though the overarching goal is to reduce the caseload, it was also important to identify cases that should be investigated. As a case cannot be closed on the grounds and there are not sufficient resources to investigate it, a reason has to be constructed that relates to some aspect of the law. The reason for doing case care in groups was that it was considered to be faster than doing so individually. Not only are there more views on how to close a case that can be covered but also on which cases should not be closed. Several calls could be made for each case, but as the purpose of these meetings is to close cases that is the default action; in each case, this action must be justified by a reason relating to the law. These may be specific to the case or use standardized phrasing such as: 'Cannot be expected to be possible to prove.' In this process of closing cases, a Prime Investigator explained that the perfect person for the job was someone who is 'inventive but not sentimental'. You need to be inventive to come up with reasons to close cases and not sentimental as deciding on which victims' cases must be closed is emotionally difficult.
Proactive limiting Some cases were seen as benefitting from being allowed to wait to see how things developed, but that was not always the case. The opposite approach was used in cases in which the suspect was highly active in crime. A Prime Investigator explained it: You don't want to draw the short straw, so for that reason you have to scan the inbox for highly criminal persons so you don't get additional investigations to do. You want to finish those cases as soon as possible if you can.
A general rule is that when there are several opened investigations relating to the same prime suspect, the department with the oldest opened case should also take responsibility for all the others. If the oldest case was registered at a given department, everything else involving the same prime suspect would be transferred there. So, if there was a chance to finish those cases quickly, they would not have to risk receiving additional cases from other regions. Loyalty was mainly exhibited towards the citizens of a department's geographical jurisdiction, and it was considered a waste of resources to work on cases from other geographical/jurisdictional areas.
C O N C L U S I O N As the case study has shown, workload is an important aspect of our understanding of the institutionalization of public professions (Muzio, Brock and Suddaby 2013). Even though some police may face the challenges of boredom, for example when they are doing surveillance (Philips 2016), the general picture in the study was that the investigating police had more to do than they could deal with. Going beyond the case study and analysing the balance from a more general and abstract perspective, the case study has shown that what the literature describes as street-level organizations 'rationing' the workload contains at least two different but intertwined processes. The first, limiting workload, may include not accepting work or using different strategies to reduce it, and the second, buffering workload, keeps certain kinds of work on hold. The work that is buffered is perceived on the one hand as being too important to dismiss but on the other not important enough to The police and 'the balance' 11 deal with immediately. The Swedish police have chosen a different path to other Swedish street-level organizations and that is to centralize buffering. Instead of buffering being done by street-level bureaucrats, it is done by senior staff within the organization. From a structural perspective, this means that the workload should not be treated as a single thing but rather as a multifaced phenomenon ( Table 2).
As the case study shows, the total workload can be divided into two parts: real workload and pseudo workload. The difference between these categories is whether or not the tasks are a burden for the streetlevel worker. If they are a burden, they are part of the real workload and if they are not a burden, they are part of the pseudo workload. The real workload consists of two main categories. The work in progress category has tasks that are actively being done, though not all the tasks in the real workload are work in progress, some are buffered. To some extent the buffering of the real workload can be seen as functional buffering for the organization to run more smoothly, but the workload can also exceed what is functional and become problematic (problematic buffering). The pseudo workload consists of three categories of workload that are not worked on actively: quasi buffering, progressive buffering, and symbolic buffering. There are two categories with a more neutral contribution to workload. These are quasi and progressive buffering and they are closely related. Quasi and progressive workloads may become a burden, but it is a waiting game to see how the cases progress. In both cases, the street-level worker is waiting for something to occur before deciding what is going to happen to the case. One difference between the quasi and progressive balance is that quasi buffering does not have to be managed in the same way. In the case of progressive buffering, the streetlevel worker has to take some action on his or her own to see results, for example if a verbal threat was a one-time occurrence or if it has happened again. Another difference between the quasi workload and the progressive workload is that the progressive workload is informally opened and the investigator has to actively check if and when something happens in the case. Quasi buffering is informally closed during the waiting period. As noted above, the last three categories of buffering (quasi, progressive, and symbolic) are not a burden for the street-level worker; they may even be the opposite, as symbolic buffering gives the street-level worker less to do by lowering the expectations of others by giving the impression that more is being done than is the case in reality. The case study showed examples of this happening towards both the organization and the public. This means that the inactive workload consists of five types of buffering. Limiting was mainly directed towards functional and problematic buffering, as these were the real burden for the organization. In the case study, several strategies were in use, the main ones being weekend group meetings, letting cases die in the balance, and making the right call on whether quick closure of the case was the best option. So, what are the reasons behind these strategies? The researchers' interpretation of the situation is that in practical terms they constitute a trade-off between the ambition to be efficient, reducing workload to a manageable level and maintaining discretion in the professional environment. In that way, it is a trade-off between the aims of three perspectives: the rational perspective (being efficient), the pragmatic perspective (individual and organizational survival), and last but not least the professional perspective (maintaining discretion and avoiding becoming too much of a routinized bureaucrat).
From the rational perspective, it is important to solve cases, help people and do what can be done for the general public and others (Scott 2003). Even though the critical reader may object that it is not efficiency per se that is important but being perceived as efficient, a good starting point for being perceived as efficient is still to actually be so. In addition, it gives the organization space to focus on what is considered most important. From the pragmatic perspective, the stakes are high when it comes to work overload. At the individual level, it is possible to suffer burnout, and at the organizational level overload also causes problems. In terms of the important issue of legitimacy, these kinds of strategies are performed backstage and rarely shown in the open. That way it is possible to reduce workload without risking scrutiny. From the professional perspective, decisionmaking in a professional context demands time to reflect and deliberate on the alternatives at hand (Abbott 1988;Liljegren 2012Liljegren , 2016. One of the key strategies for managing a heavy workload is to standardize decision-making into limited and fixed categories and act more as a bureaucrat than a professional (Tummers et al. 2015). In that way, a heavy workload constitutes an ongoing threat to professionalism and, as can be expected, professionals defend their discretion. In sum, there are strong incentives to act in this way. Even though the literature has paid much attention to both the most valued and dirty work of given professions (Hughes 1963;Abbott 1992;Dick 2005;Löfstrand, Loftus and Loader, 2016), the main challenge in professional power struggles is not for the most valued work, nor is it to hand the dirty work to others. Instead, it is the struggle by professionals to receive reasonable amounts of the work that in essence they enjoy and value.
Lastly, it may be tempting to moralize about how workload is managed, but from an institutional perspective it is important to remember that following all laws, rules and routines would lead to the collapse of the individual. As Lipsky has noted, there is a tendency for street-level workers to 'drop out or burn out ' (1980: xii). The same thing would happen to entire systems if organizations did not manage workloads. For that reason, the management techniques presented here make it possible to uphold the present system. Whether upholding the system is good or bad is a question to which we have no answer.

A C K N O W L E D G E M E N T
The authors would like to thank three anonymous reviewers and the editor for constructive and to the point comments on the manuscripts. The authors would also like to thank a Swedish police district for generously taking part in the project.