“It’s about Living Like Everyone Else”: Dichotomies of Housing Support in Swedish Mental Health Care

The deinstitutionalization of psychiatric care has not only altered the living conditions for people with severe mental illness but has also greatly affected social services staff. In the Mental Health Act launched by the Swedish government in 1995, a new kind of service called ‘housing support’ and a new occupational group, ‘housing support workers,’ was introduced. However, housing support does not currently operate under any specific guidelines regarding the content of the service. This study explores housing support at local level in various municipalities of one Swedish county. The data is based on discussion with three focus groups: care managers, managers for home and community‐based support, and housing sup‐ porter workers. The perspective of institutional logics as a specific set of frames that creates a standard for what should or could be done, or alternately what cannot be questioned, is applied to analyze the constructed meaning of housing sup‐ port. The meaning of housing support is constructed through three dichotomies: process and product, independence and dependence, and flexibility and structure. These dichotomies can be understood as dilemmas inherent in the work and organizing of housing support. With no clear guidelines, the levels of organizational and professional discretion create a space for local flexibility but may also contribute to tremendous differences in defining and implementing housing support. We discuss the potential consequences for housing support users implied by the identified discrepancies.


Introduction
This article focuses on housing support (HS) and its meaning in the context of deinstitutionalized mental health care in Sweden. Here, we investigate questions about HS and severe mental illness (SMI) by focusing on contested issues of professional identity, the constructions of meaning of HS, and the various logics that inform HS's organization and provision at local levels. The aim is to shed light on the intrinsic complexity of providing home-based social support to persons with mental illness living independently in the community. This goal will be accomplished by using the lenses of institutional logics, which is a theoretical framework that identifies and accounts for the many and various social influences on institutions.

Housing Support
HS was introduced in Sweden in the 1990s, and it is an important social welfare response to the goals of normalization and social integration for a number of vulnerable groups. Broadly speaking, HS is meant to support users in their daily lives (which includes both home life and life outside the home), and ideally it should be achieved via an establishment of relations with those users (Andersson & Gustafsson, 2019;National Board of Health and Welfare, 2010). Yet, with no guidelines nor policies regulating its provision or organization, HS remains surrounded by organizational and professional uncertainty as well as a lack of clarity about contacts and relations between housing support workers (HSWs) and users. In addressing these uncertainties, this article brings to the fore the understandings of, and meanings assigned to, HS by the HSWs themselves, their immediate team leaders, and managers of community-based mental health support.
HS has emerged as essential for enabling the everyday life and providing non-clinical support to people with SMI who are living independently (Brolin et al., 2018;Shepherd & Meehan, 2013;Shepherd et al., 2014). This population of people is faced with the task of managing their social lives, interacting with others, and establishing their home lives in the context of a community. However, SMI very often impedes social abilities, and without appropriate support, independent living may turn into life in loneliness and social isolation.
The uniqueness of HS in the context of independent living is related to the fact that HS turns the home into a site of professional intervention (Gonzalez & Andvig, 2015)-it actively uses the private space of a user for a professional purpose (Juhila et al., 2016). The context of professional work performed in private spaces is largely built on the quality of relations existing between HSWs and users, and creates special circumstances for negotiating own positions, roles, and performing various tasks. HSWs face numerous emotional and bureaucratic challenges (Ericsson & Bengtsson Tops, 2014;Ericsson et al., 2016) involving, among other things, a need to balance duties and relations to clients (Shepherd et al., 2014) while simultaneously negotiating their own position in relation to other mental health care professionals (Shepherd, 2019).
Along with most western countries, in the latter half of the twentieth century, Sweden reorganized psychiatric care according to the principles of deinstitutional-ization, which meant closing the old mental hospitals and replacing them with smaller units and open care. In Sweden, the process started in the mid-1970s with a nationwide implementation of Community Mental Health centres with outpatient units as a complement to psychiatric hospitalisation. Ideally, every psychiatric clinic would be responsible for all in-and outpatient treatment within a defined catchment area (sectorisation). The aim was to improve the living conditions of persons with SMI and facilitate the transformation from a patient to a person, and for that person to become an active citizen in society (National Board of Health andWelfare, 1970, 1980). However, the pace was slow, and evaluations showed that persons with SMI did not necessarily benefit from open care's lack of support for social needs relating to daily life matters (Stefansson & Hansson, 2001). In 1988-1989, a survey was conducted by Statistics Sweden with the aim of investigating the living conditions of the Swedish population. In this survey it was found that people with SMI had living conditions far worse than average for the Swedish population as a whole, and significantly lower than for example groups with physical disabilities (Prop, 1993). As a response to the result of the survey, the Swedish Government commissioned a parliamentary committee to make proposals for a reformation of psychiatry and psychiatric care in Sweden, which resulted in the 1995 Psychiatric Care Reform. The reform clarified the responsibilities of social services and psychiatry. Social services would be responsible for providing support to persons with SMI in questions of housing, employment and everyday life, thereby establishing the conditions needed for integration into society. The task of the county psychiatry council would be to develop psychiatric treatments and prevent psychiatric illnesses.

Institutional Logics
To explore the emerging meanings attached to HS, we take of the perspective of institutional logics. The concept of institutional logics (first introduced by Friedland & Alford, 1991) expanded the field of institutional theory by drawing attention to societal influences on institutions, the ways institutions change, and the role individual actors play in that process of change (Johansen & Waldorff, 2017). Institutional logics are commonly used to help understand contemporary institutions and observing these logics at work can be used to "represent frames of reference that condition actor's choices for sensemaking, the vocabulary they use to motivate action, and their sense of identity" (Thornton et al., 2012, p. 2). With its focus on material practices and symbols, this new perspective has brought heightened awareness of the making of institutions in practice and their exposure to both external forces and internal processes of interpretation. Rather than seeing institutions as closed and finished entities, the perspective of institutional logics brings forward the everchanging character of institutions that are transformed in the course of everyday practices.
Essential to the understanding of institutional logics and their functions is the notion that multiple logics guide institutions and organizational behaviors. These multiple institutional logics may entail very different, and sometimes conflicting, directions for institutions and institutional actors (for instance, differences between family and market logics; see Martin et al., 2017). Institutional complexity, which increases along with the number of institutional logics and their degrees of incompatibility, can be seen as the result of multiple institutional logics (Greenwood et al., 2011). Multiple and divergent institutional logics may lead to tensions, but at the same time, they may also provide a scope for creative solutions (Martin et al., 2017). For these reasons, the perspective of institutional logics is especially relevant for analyzing the contradictions and dichotomies surrounding the complex institutional setting of HS and the various logics that need to be effectively managed at the frontline of practice (Lipsky, 2010).
Institutional logics is also committed to exploring the local embeddedness and enactments of these logics by various institutional actors (e.g., Currie & Spyridonidis, 2016;McPherson & Sauder, 2013;Pallas et al., 2016). Thus, while conditioning choices and behavior, institutional logics are also "somewhat elastic, being sensitive to local actors' capacities and motives to actively and continuously interpret and enact their different parts" (Pallas et al., 2016(Pallas et al., , p. 1680. Institutional logics can be applied in various way on the ground, depending on the situational constraints, the actors involved, and the positions of those actors in the specific situation. In a way, the more problematic the institutional logics and its various elements, the greater the level of engagement and adaptation of institutional logic to a particular situation or circumstance (Pallas et al., 2016).
A good example of how different institutional logics can stem from actors and local context, and how similar logics can translate to different outcomes, was presented by McPherson and Sauder (2013). Their micro-study of drug court proceedings showed that, especially in the context of contest or conflict, different logics originating from the same institution can be employed by different actors to achieve different goals. Conversely, they also demonstrated that any particular institutional logic may be used differentially depending on who applies it, which means that the same logic may be used to serve different purposes. That variation in application of institutional logic at the ground level reflects situational constrains as well as actors and their positioning within given situation. In a way, the more problematic the institutional logics and its various elements, the greater the level of engagement and adaptation of institutional logic to local circumstances (Pallas et al., 2016). Crucially, as Pallas et al. (2016) emphasize, it is the active process of local translation that brings institutional logics to live and thereby, testifies to various enactments and consequences of the same logics.
The perspective of institutional logics brings forward not only the dynamic side of institutions, but it also pays tribute to the individual and collective agency of institutional actors. In the context of welfare professionals, agency is often conceptualized in terms of professional discretion, which encompasses how professionals make judgements and decisions, and interpret policies, as they perform their work on the ground. Discretion typically involves structural and epistemic dimensions (Molander, 2016): While the epistemic dimension concerns the actor's reasoning regarding preferred courses of action (which may vary from case to case), the structural dimension concerns the overall legal, institutional, and organizational frames that delimit the boundaries of professional conduct. The structural dimension of discretion reflects the influence of institutional logic on decisions and judgements made by professionals who, through their agency, actively respond to the various institutional logics (Garrow & Grusky, 2013).
Each institutional logic provides a set of assumptions about what should or could be done, or about what cannot be questioned, and each logic therefore simultaneously enables and constrains agency. One such logical framework relevant to HS includes the ideas of deinstitutionalization and normalization that were formalized with the Psychiatric Care Reform. The reform clearly promoted the notion that people with SMI would be able to enjoy 'a normal life' in communities just like everyone else. These ideas and new or altered organizations reflected the gradually changing perceptions of disability and the social status of persons with disabilities. Instead of 'patients,' people with SMI and other disabilities became increasingly regarded as (active) citizens (Lindqvist et al., 2012;Lindqvist & Sépulchre, 2016). How and whether this frame of deinstitutionalization is visible and enacted in the practice of HS has yet to be explored.
The positions and roles of HSWs can be also considered thorough the perspective of other social welfare professionals (like case and care managers) and the various institutional logics that affect them. For example, as welfare workers, HSWs might be affected by the overarching logic of bureaucracy that can potentially constrain their occupational role (professional logic; see Freidson, 2001) and the ways they would prefer to engage with people with SMI. At the same time, welfare professionals (managers) are urged to categorize individuals and standardize practices (Hasenfeld, 2010;Lipsky, 2010). On the other hand, the logic of individualization that is highly valued in western welfare states (and prevalent in the field) may push HSWs (and other welfare actors) to adapt to the will of the users and their unique situations. HSWs have the primary role of enabling contacts between clients and the outside world, which makes them the foremost bearers of social connectedness and relations for these clients (National Board of Health and Welfare, 2010). Research confirms that the quality of the relationship between clients and HSWs is the key to successfully providing HS (e.g., Andersson & Gustafsson, 2019;Gough & Bennsäter, 2001;Ljungberg et al., 2017).
In this article, we draw on the experiences of three occupational groups that, in their various positions, are responsible for planning, managing, and executing HS. In their voices, we can hear the various logics in play through their understandings of what HS is and the ways that the same institutional logics may potentially be interpreted and practiced differently by the representatives of those various groups. Our analytical focus is guided by the quest of identifying their institutional logics and understanding the ways in which those are enacted in practice. The complex and often contested context of providing HS for those with SMI is a rich source of material for investigating these issues.

Methods
The study is a part of a collaboration between the authors and a group consisting of former users of HS and professionals either working with HS (HSW, care managers) or with experience of people with SMI (a retired psychiatry nurse). The goal of the project is to explore the provision and organization of HS using the framework of institutional logics.
The empirical material consists of three homogenous online focus groups, suitable for capturing rich qualitative data where participants share opinions, experiences, and construct meanings about, in this case, HS (Kitzinger, 1994). Three categories of welfare workers participated: HSW, care managers, and managers for home and community-based support. In addition, one individual online interview was conducted with an HSW (the participant could not attend the meeting). These three categories of welfare workers were strategically chosen to represent the different domains that affect realization of HS at the frontline. Focus group interviews were conducted online because participants are situated in different municipalities (see Woodyatt et al., 2016) and to minimize the risks in light of the ongoing Covid-19 pandemic. The focus group interviews (audio and visual via Zoom) were recorded.
The HSW focus group comprised five participants from two different municipalities, and their levels of experience ranged from five to 20 years (the individually interviewed HSW had six years of experience). The care managers focus group comprised eight participants from seven different municipalities, and they had specialized on people with SMI from one to eight years. The managers for home and community-based support focus group comprised four participants with one to four years of experience from four different municipalities. All municipalities were in the same county in Sweden. All recorded interviews comprised of a total of 398 minutes of data.
In order to grasp the planning, management and execution of HS, interviews focused on six general themes: the meaning of HS in your context(s), how work is carried out, knowledge needed to perform the work, relevant education, examples from work considered hard or challenging, and collaboration with different stakeholders. For the focus group with managers for home and community-based support, we added two themes: competence required when recruiting and distribution of work tasks. Themes were chosen based on previous research on HS as well as discussions with the group of collaborators. Prior to the focus group interview, and as a kind of preparation and a way of triggering discussion, material based on excerpts from newspapers were sent to the participants. These newspaper excerpts contained interviews with managers and HSW and covered topics such as competence, how much HS users might need the services, and the meaning of HS. Each focus group and the individual interview started with a presentation of participants followed by a question about their thoughts and feelings about the material. The participants were also urged to talk freely, respond to each other, and exchange experiences. Typically, one or two of the participants in each focus group responded to the stimulus question by briefly commenting (such as "I found that interesting") on some part of the content in the stimulus material before moving on to talk about their own practice. That is, the participants themselves had the ability to control which paths the conversations took based on what they considered to be relevant and important. Two of the researchers conducted the interviews. Naturally, the individual interview was more of a discussion between the researchers and the respondent, whereas the researchers had a more peripheral role in the focus groups.
The recordings were transcribed verbatim and analyzed according to the following strategy. First, each researcher individually conducted an empirically-based coding and analysis of the transcripts to get familiar with the data. Terms used in this stage of analysis were thus close to the raw data. Second, after identifying overarching themes, the research team met to compare and discuss the themes each of us had found. Third, based on those discussions, a thematic matrix was constructed and supplemented with illustrative quotes from the different welfare workers. This thematic matrix was thereafter presented and discussed with the group of collaborators. This discussion highlighted the various difficulties HSW and other actors face when doing HS work in practice, for example, the need to adapt to the needs of the individual while simultaneously fulfilling duties in line with organizational imperatives. Such difficulties were understood as dichotomies of concepts and approaches creating dilemmas (cf. Lipsky, 2010) experienced and managed in the specific context of HS. Next, we expanded the analysis further by exploring and interpreting the different meanings attached to HS from the framework of institutional logics, focusing on the contradicting conditions and challenges embedded in the setting of HS and made visible by the data. From the new reading and thematization of the material, three different dichotomies of core concepts emerged in the process of analysis which makes up the findings and analysis section below.

Findings and Analysis
The analysis of focus groups revealed an apparent struggle common to all research participants in defining HS. Thus, for example, when trying to describe their own role, HSW often resorted to various metaphors, such as 'being a trigger,' or 'making the clients bloom.' While such metaphors highlight the visionary perspectives guiding individual workers, they cannot offer clear images of how to define the role and, by extension, the content of HS. In a similar vein, one of the care managers emphasized that in HS "it's about living like everyone else." Clearly, these statements can be conceived of as indicating that the normalization principle dominates the reasoning about HS. However, further analysis reveals that as a form of institutional logic, normalization can be applied and understood differently and may involve various elements.
Three separate dichotomies of core concepts were identified in analysis: process and product, independence and dependence, and flexibility and structure. These dichotomies can be understood as dilemmas inherent in the work and organizing of HS. The dichotomies also signify various types of agency and their active roles in redefining different institutional logics, pulling HS in diverse directions. These dichotomies bring forward political and organizational aspects, like the contextual settings for HS. On the one hand, HS is supposed to lead to the grand objective of individualization, however this objective is neither grounded nor situationally located in an organization. There is still a desire to fit needs into predetermined structures, however loosely constructed.

Process and Product
With the dichotomy of process and product, we illustrate the balancing act between understanding HS as a product-focused practice and as a process-oriented practice. The distinction between process and product indicates differential values and modes of working that may either embrace the notion of long-term engagement and its various aspects or focus on concrete activities and their accomplishment. We exemplify this dichotomy with two discussions that focus group members had about the practices of talking and vacuuming. First, the HSW focus group had this to say about having conversations with users: I am thinking about this thing with conversations, where we had a discussion about when and how to have conversations, because there are people that we go to that can be so caught up in a conversation and there is only conversation, and nothing gets done. And then, we had a discussion some years back, that the conversation is often a reward for a person. So, you put it at the end of… and limit it to a certain time, because otherwise you can spend an hour and there is just talk. (HSW focus group) That's the case for us as well. I mean, we don't have conversations where we sit down and talk like that, unless it results in….We always go through the schedule, what does the week look like? And you start there. Motivational talk, I mean. (HSW focus group) HS is meant to provide support in daily life and, as many research participants emphasized, the ultimate purpose is to facilitate personal growth and positive change in people's daily lives and their social worlds. Simultaneously, HS is a type of welfare practice that is constrained by the same standardization and efficiency principles as other social welfare services. Engaging in this context in HS can therefore pose challenges regarding values and priorities at work. There is a gradual shift in focus from human processes of interaction that could typically be essential for 'making the clients bloom' to concrete products that can be planned, executed, and measured. Thus, the idea of talking to clients is discussed in terms of waste and meaningless activity. The only time when talking is appreciated is when it leads to something else, for example, it motivates clients to do something. 'Just talking' is not seen as something tangible enough to be recognized as an important part of social support. On the other hand, when introducing the idea of conversation as a part of a reward system, there is a recognition that it is a meaningful activity for the clients. Yet, in this context, the notion of doing one's job seems to override the principle of attending to client's needs and supporting social life.
While talking is not perceived as productive, some typical household activities are: HSW 1: I just thought I should add that it is very important that the assignment from the case manager is very clear, very concrete, it helps extremely when we go in and do this in practice. Is it fuzzy, it's broad formulation, it's… what can partly make it difficult in practice. But also, that the client gets another apprehension about what housing support is, what we are here for. And it has become much more brief, clear and in bullet point format. Just these five recent years that I have worked. I: Can you give an example of what a good assignment might sound like? HSW1: Yes, an example can be a shift from before where it said "support in the maintenance of the home." I: That's the fuzzy version? HSW 1: Yes, that's the fuzzy version. A little clearer version could be "to motivate to vacuuming the living room." HSW 2: That's a difference. (HSW focus group) The example with vacuuming was recurrently used in various focus groups to indicate the concrete aspects of HS. In the context of no rules or guidelines to inform the practice of HS, those in charge of organizing and providing HS attempt to break the overarching goal of support into smaller activities that give specific frames for acting. In that process of translating the notion of support and with that normalization, HS is construed as including numerous activities that direct the behavior, and also allow visible outcomes of the work done. Considering the home-based context of HS, such concrete activities revolve largely around household chores. A clean apartment, washing dishes, and doing laundry thus become indicators of the effectiveness of professional intervention. Indirectly, however, such indicators reduce the potential influence of HS on the process of social integration. Too much focus on household chores that are used both as activities framing HS and providing measurable outcomes risks turning the household site into the intervention site. People with mental illness receiving HS may become more proficient in their household duties, but their social lives, which very often are affected by their underlying illness, may not be affected by this intervention at all.
The process-product dichotomy is also reflected in the research participants' talk about time: The length [of HS] varies a lot, it is not possible to decide ahead how much time is needed, there are the clients' needs and conditions that determine that. (Care managers focus group) The lack of clearness can also imply long interventions that are difficult to end. (Care managers focus group) If the clients have HS during a long time and do not move on, then home care becomes more relevant. (Managers for home and community-based support focus group) HS is provided on individual basis and conditioned upon an assessment of needs. As a part of the assessment, it is recognized that the period during which HS can be provided may vary depending on the clients' wishes and needs. However, this person-centered logic that informs practice may clash with the logics of productivity and efficiency. The logic of productivity and efficiency may turn time (or more precisely, the length of the intervention) into a criterion of success. According to such reasoning, short periods of HS are indicative of success while longer periods may suggest a client's inability to progress, indicating a failure of HS. The product-oriented frame not only presents concrete activities that are deemed appropriate, but it also provides specific time intervals that are considered reasonable. It is noteworthy that such reasoning came up in interviews with both groups of managers, but did not come up during the focus group interview with HSW.

Independence and Dependence
The second dichotomy is in the balancing act of helping clients develop an independence in their daily lives, while setting up ground rules and structure surrounding everyday life. The importance of achieving independence is emphasized in all conversations; the independence of living 'like everyone else.' Who this 'everyone else' actually is is neither detailed nor explained, but rather emphasized by various people, and it seems to be understood as something obvious and strictly positive, something to aim and strive for.
However, the practice of HS is built on structures and content that can instead emphasize dependence on others. The independence of clients was mentioned often in the interviews, though not specified at all, leaving much room for interpretation. Client independence was always set in an organizational (and hence societal) context, leaving little or no room for questioning the claim of always aiming for independence.
The dichotomy of independence and dependence illustrates how independence is talked about in positive terms as the main goal of HS. However, the dichotomy also illustrates the deeply imbedded discourse of the limitations to this so-called independence. Independence comes with expectations and limitations connected to the practical work of HS, even in terms of regulations. Discourses about and explanations of independence are accompanied with a 'but,' explaining the limitations to independence in various cases: The client should participate, of course. Maybe you can't handle everything from the beginning, but then it is our thing to find ways for them to be as independent as possible in what they want….Because participation is pretty important, and that… that we can work on this together with the client [for them] to be more and more independent simply. That's how I think about it. (HSW focus group) The quote comes from a HSW in a focus group, explaining the circumstances and daily work needed to achieve this independence for others. This comment was followed by another HSW adding support to the previous claim: I exactly agree with what you said. For me it's also about coming in when it comes to boundaries. In many of these decisions there is a very unclear limit. Where is the limit for how much we do, what we do, what we agree to do. There is always request for more in many cases. (HSW focus group) This second quote addresses the matter of boundaries and additional requests for help. The exact moment that boundaries are set can be the moment when an HSW clearly steps in and takes charge of a situation. Because there is no limit to potential further requests to be made, the HSW must be firm in their positioning and boundary setting, however challenging doing this might be. The need for HS is based on a prior assessment which is supposed to have set the boundaries beforehand, but HSW are the ones meeting clients and hence are the ones faced with more requests from clients. As pointed out by a care manager in the dichotomy of process and product, the needs assessment is challenging, because "it is not possible to decide ahead how much time is needed, there are the clients' needs and conditions that determine that" (care managers focus group. HSW are the ones that must navigate between requests and needs within the loose organizational context of HS.
Moreover, the issue of stressing things to do is emphasized. Similarly, to the dichotomy of process and product, there is an emphasis on practical matters to attend to, such as housekeeping. The role of an HSW is not merely to show up and offer whatever support is needed, but instead, things must be done and completed. Practical issues are stressed because these give a sense of accomplishment, which in turn is thought to lead to independence.
The following quote is from the same focus group and conversation as the prior quotes, and here, the withdrawal of HS is pondered: And I think that….On the other hand maybe you can think that if you need your housing support several times a week then….I am saying, that surely you need some kind of maintenance dose of your housing supporter to not fall back into something. So I think it's a good thought to not just disappear. (HSW focus group) So it is understood that mutually created dependence between HSW and client must come to an end at some point, however, the ending is not so easily completed. Everything else connected to the service and use of HS has fuzzy boundaries and limits that are difficult to determine, and the ending of HS is no exception. And, as in many cases of welfare work, clients receiving HS are not the ones fully in charge of their own situation.

Flexibility and Structure
Lastly, the dichotomy of flexibility and structure highlights the ambitions to have clients 'live like everybody else' and thus the need for professionals to be flexible and responsive in relation to a client's will, characteristics, and specific situation. However, at the same time, there are structures of practice that might complicate such ambitions. Consequently, negotiations need to hap-pen on a daily basis between both clients and organizational representatives and between different occupational groups. In the following we present two examples of this dichotomy that focus on the content of HS and control of HS intensity.
The following example demonstrates reasoning in the care manager focus group about a client who initially was considered 'hard to work with' in terms of the goal of independence. Therefore, organizational flexibility was called for, and "maybe we promised a little too much verbally, even though it is not stated in the formal decision." Now, "the user's parents are very assertive" and say that more HS-activities to be implemented: There are things like, even though the user has training once a week, is out and about with the dog every day, they [the parents] still think that my staff should take the person and go for long walks together with the user and the users pet just because the person should have someone to talk to. And this is a person who moves, is active, is at work during the day and has co-workers, has activities every week and so on. And so still you must go out….It is not even reason-able….Do you understand? Most people, they have a job, they have leisure activities, they go out and walk the dog. That's it. That's where it ends. But then they demand much more….Yes, then I have to put my foot down. Tell the staff [HSW], this is how we think. Tell the care manager….Because we are played out otherwise by the user and the parents. (Care managers focus group) Flexibility in relation to the user (and in this and other cases, the user's family) can only be accepted to a certain extent, otherwise 'we are played out,' indicating a conflict between the different parties concerning to content of, and by extension the amount of time dedicated to, the HS intervention. The rationale behind the position taken by the manager seems to be that the claims made to extend the HS stand in opposition to the logic of normalization indicated by the phrase 'most people.' If stretched too far, HS might be something that obstructs normalization and needs to stop. Even though flexibility and the logic of participation is, according to all participants, considered paramount for success, organizational boundaries need to be drawn. Since the structural dictates concerning content and intensity of HS are very loosely constructed, the manager invokes the logic of normalization as a tool for decision making.
The dichotomy of flexibility and structure also relates to relationships among organizational representatives. The structure of the purchase-provider model means that care managers assess needs and give assignments to others who then execute HS. From this follows a need to control that HS is used efficiently at the frontline and according to the assignment. The following excerpts exemplify this need for control: It is of course a challenge that it is not the housing supporters who should decide, but it is we who make the decision after investigating what the client's needs are. But if you have a good collaboration, you can give and take a lot there anyway. (Care managers focus group) Yes, it's a balance to go in and control everything, 'cause we don't see the clients as much as the housing support workers do. And they might see other things than what I do when I meet the client for an hour, an hour and a half, to make a decision. So, I think that what you are saying is very important, to have a good communication between housing supporters. In part so that it doesn't get out of hand, that they come up with many other assignments, 'cause it might not be in the assignment of housing support, but it might be some other function that should actually do these things. So it's a matter of both giving support, I think, to housing support workers, 'cause they might also find support things that I didn't catch in my needs assessment. But also to sneakily control a little so that it doesn't fly off… to have a good communication and be able to say that "this is your assignment.'' This is something else. (Care managers focus group) The excerpt above illustrates that the care managers must maintain a balance between being flexible (since their knowledge about user's characteristics and situation are somewhat limited) and their task associated with their organizational position. The fact that good collaboration and communication with other workers means that one can 'give and take a lot' when it comes to reaching decisions about the content and intensity of HS is important from the point of view of care managers, since they don't see the clients as much as HSWs do. Flexibility is thus called for. However, good communications are also necessary for HS not to 'get out of hand' because HSWs can 'come up with many other assignments.'

Discussion
In this article we have analyzed the contradicting conditions embedded in the setting of HS, handled in everyday life by people in the frontline of practice (Hasenfeld, 2010;Lipsky, 2010). The above excerpts are examples of the ways in which HSWs deal with the specific challenges within their practice, which we describe here as diverse dichotomies. The dichotomies make available a deeper insight into the everyday life of HS and the sensemaking (Thornton et al., 2012) imbedded in HS. This study has focused in particular on HS for people with SMI, who have gone from being 'patients,' to increasingly being regarded as (active) citizens (Lindqvist et al., 2012;Lindqvist & Sépulchre, 2016). Living independently, people with SMI are faced with the task of managing their social life, interacting with others, and establishing their home life in the community context.
Using institutional logics as the framework for understanding the premises of the work in HS and the special conditions from which HSWs work allowed for an analysis illustrating a diversity of dichotomies (or contradictions in practice). These perceived dichotomies reflect and imply practical dilemmas, which are not only visible when it comes to how the work is defined, but also in the way that clients are perceived within that specific context of practice. For example, in the dichotomy flexibility and structure, 'good communication' (from the care managers) means being able to see that the structures of HS drawn up in the needs-assessment are adhered to (by the HSWs), although sometimes managers 'sneakily control a little' to ensure adherence. Interestingly, in all situations mentioned in interviews, it is too much HS rather than too little HS that managers feel they need to control, indicating that a logic of efficiency is being employed by all parties.
The dichotomies also signify various types of agency and their active roles in redefining different institutional logics, which have the effect of pulling HS in diverse directions. These dichotomies bring political and organizational aspects, such as the contextual settings for HS, into the fore of the discussion. On the one hand, HS is supposed to lead to the grand objective of individualization, however this objective is not grounded nor situationally placed within an organization. There is still a desire to fit needs into predetermined structures, however loosely constructed.
The case described here, of HS for people with SMI, shows that there can be a discrepancy between the spectra of grand visions and what is described as what happens in practice. The dichotomies identified here are between value-laden concepts, heavily burdened through organizational histories of right and wrongs. These values are not easily overlooked, and they create the setting in which today's HS practice is situated. The conflicting logics and expectations regarding HS and the work performed by HSW may push the practice in different directions, especially given the policy vacuum surrounding HS in Sweden.