Incorporation and Individualization of Collective Voices: Public Service User Involvement and the User Movement’s Mobilization for Change

‘Service user involvement’ is a widespread and well-known phenomenon within welfare policy and practice in Western countries and is usually perceived as a way of improving welfare services to better aid service users in managing their predicaments. However, the presented ethnographical study of service user involvement within a Swedish psychiatry organization shows that user involvement initiatives might also result in unintended and undesired effects on the collective user movement (i.e. the service user organizations involved in the activities). The analysis suggests that initiatives on user involvement might affect both the constitution of the user movement as well as the way the movement operates. Theoretically co-optation theory informs the analysis.


Introduction
The notion of citizen participation is prominent in contemporary political discourse (see Barnes et al. 2007;Clarke 2013), and the concept of 'service user involvement' is widespread within welfare policy and practice in Western countries (see Beresford and Carr 2012;Rantamäki 2017). 1 The core idea is that the users of welfare arrangements should be able to comment and affect policy and service provision, supposedly benefitting service users and the welfare administration alike. Before (and besides) governmental efforts to create user involvement in the welfare system, service users have a history of organizing themselves collectively and independently of the state (see Barnes 1999;Crossley 1999). Such acts of social mobilization-where individuals or groups organize in civil society with the aim of changing politics or social conditions that they perceive as incorrect, unjust or discriminating-are often understood in terms of social movements (see Melucci 1989;Scott 1990;Tilly 1999;Della Porta and Diani 2006). Hence, the 'service user movement' (see Crossley 1999;Näslund et al. 2017) can be understood as a social movement where users of various public services organize (predominantly through 'user organizations') with the aim to improve the situation for its target group. Alm Andreassen (2008) highlights that the practice of user involvement stages a specific kind of relationship between the public administration and the user movement-a relationship that is the centre of attention for this article. Davies et al. (2014: 120) emphasize that the user movement can work 'alongside, against or within government-sanctioned service user forums'. But over the last decades, the alignment between the state and the voluntary sector has grown stronger, increasing the emphasis on collaboration between the parties (Salamon 1994), a development also evident in the Swedish context (Johansson 2011;Lundberg 2012). Within Swedish public policy, user involvement is articulated as the government working in harmony with the user movement, collectively pursuing the same goals (Eriksson 2015;Hultqvist and Salonen 2016). In such rhetoric, public initiatives on service user involvement are associated with the same mindset as social movement mobilization. They are depicted as conjoining ways of reaching the shared goal of better services and social conditions for vulnerable groups in society, and governmental initiatives on service user involvement are understood as facilitating user movement mobilization and supporting the user movement's cause. Such consensus rhetoric risks concealing some of the fundamental differences in positions and interests between the public administration and the user movement (see Forbes and Sashidharan 1997), as well as disregarding the possibility that service user involvement could have harmful effects on the user movement. Service user involvement might indeed contribute to local welfare service development and improve the situation for local service users. At the same time, public user involvement initiatives could potentially have unintended, undesired, and undetected effects on the user movement.

Purpose and Questions
An almost self-evident theme when researching user involvement is the service users' impact (or absence of impact) on the public administration (see Hodge 2005: 164). This article furthers the understanding of user involvement by demonstrating that the public administration at the same time also might impact the user movement.
The ambition to work from a user perspective is a core value in professional health care and social work (British Association of Social Workers 2012), and my intention is not to criticize that ambition. Rather, the purpose of the article is to critically examine how initiatives to create user involvement might unintendedly affect the autonomous user movement. This is demonstrated through an ethnographical case study of the practice of service user involvement within a large public psychiatry organization in Sweden. The question guiding the analysis is 'What potentially harmful effects could public service user involvement initiatives have on social movement mobilization among service users?' 2 Drawing on the empirical study, the article presents two potentially harmful ways that involvement initiatives can influence the user movement: by affecting the constitution of the user movement (i.e. service users' ways of mobilizing within the movement), as well as the way the movement operates (i.e. how the user movement formulates and pursues its goals).

The User Movement and Service User Involvement in Sweden
At the centre of attention in this article is (a local part of) the Swedish service user movement, advocating the perspective of service users in the field of psychiatric care, and I understand this user movement as a social movement. Definitions of social movements differ but usually contain a few overarching elements, the most central one being the collective action of individuals mobilizing in the sphere of civil society (i.e. autonomous from public or corporate initiatives). Furthermore, social movements are characterized by shared collective identities or viewpoints, solidarity within the group and collective action to pursue a common goal-typically some kind of political influence, policy change or social change (see Melucci 1989;Scott 1990: 6;Della Porta and Diani 2006: 20). Conflict with the political establishment, power structures, institutions or institutionalized norms in society is another key feature of social movements (Della Porta and Diani 2006: 21;Tilly 1999: 257). A social movement must also show some degree of internal structure and organization. Nonetheless, a social movement can never consist of one single (formal) organization. A social movement is always a network of individuals and/or organizations interacting with other actors (Della Porta and Diani 2006: 21).
In the Swedish welfare setting, the disability movement, formed in the 1860s, is an early example of a social movement (Persson Bergvall and Sjöberg 2012). About a century later-with the labour movement and civil rights movements as role models, and the anti-psychiatry movement as inspiration-social mobilization among users of welfare services grew in the 1960s and 1970s, laying the foundation for the modern Swedish service user movement (Sunesson 1989;Johansson 2011;Markström and Karlsson 2013). 3 With support from radical professionals, psychiatry patients, prisoners and service users within addiction treatment formed their own organizations trying to influence the welfare practice they perceived as high-handed and paternalistic. They challenged the state by demanding democratization and the sharing of power, and especially during the 1970s the user movement was a powerful actor in the socio-political arena in Sweden (Modig 2006). The contemporary Swedish user movement is perhaps not as influential, but has been extended by many new organizations gathering together individuals using a variety of welfare services. While parts of the movement still work primarily through advocating political and social change, other parts put more effort into supporting their target group or empowering collective identities (Johansson 2011). In the year 2007, twelve of the largest user organizations within the field of psychiatry formed a national network (NSPH) to promote their agenda. NSPH has regional divisions in all parts of Sweden and was originally initiated by the state as part of a political strategy to strengthen user perspectives within the field of psychiatry (SOU 2006:100, p. 531).
Even if advocacy and voice (rather than service provision) have dominated the activities of the Swedish voluntary movements, Kuhnle and Selle (1992) have shown that the relationship with the state has been characterized by agreements and reciprocity, rather than outright conflict (see also Lundström and Wijkström 1997). Thus, the relationship has been significantly substantiated by the strong Swedish tradition of corporatism (see Ö berg et al. 2011). Inspired by the UK 'compacts' (see Lewis 1999), in 2008 the relationship between the Swedish voluntary movements in the social field (Including the user movement figuring in this study) and the state was formalized through a collaboration document called 'the agreement' (Ö verenskommelsens 2008). This agreement clearly declares the importance of the civil society taking an autonomous, 'critically scrutinizing role' in relation to the state (ibid., pp. 5, 11, 22). However, as Johansson (2011) notes, the agreement inevitably affects the user movement's autonomy by connecting it more closely to the state. The document emphasizes that the relationship between the parties should 'safeguard the perspectives of both [my italics] parties', and that the purpose is to create 'mutual understanding and trust' (Ö verenskommelsens 2008: 22).

Service User Involvement Becomes an Administrative Concept
The early user movement's demand for influence gave rise to the first notions of user involvement within Swedish welfare, articulated in terms of democratization, policy transformation, power sharing and service users' collective rights. Croft and Beresford (1992) refer to this as 'the democratic approach' to user involvement. While the idea of user involvement was initially formulated from outside the public administration by an independent grass-roots movement, those ideas have gradually been adopted by the state. Today user involvement is largely a political and organizational concept managed from above and within the Swedish welfare administration. In political discourses infused by the logics of new public management, user involvement is formulated as a matter of measuring and improving efficiency and service quality (Barnes 1999;Carr 2007), service users are reimagined as 'customers', and individual freedom of choice becomes a core intention (Davies et al. 2014). Croft and Beresford (1992) describe this formulation of user involvement as 'the managerialist approach'. Compared to other Western states, the Swedish government was quite late to incorporate the concept of service user involvement in its welfare politics. Lindqvist (2007: 14) argues that this was because the Swedish welfare state was built on a strong common trust in 'the good state'. In a context where the state was already understood as working in the best interests of the citizens, organized service user involvement was perhaps not as necessary as in countries where more suspicion and distrust characterized the common view of the state. However, the first Swedish government report on service user involvement appeared in 1991 (Civildepartementet 1991), and for the last 20 years official government reports in the welfare field steadily highlight the importance of user involvement.
Although Swedish policy on user involvement draws on both the democratic and a managerialist approach, the managerialist approach dominates the formulation (see Eriksson 2018). Thus, when transformed into a public concept, service user involvement becomes something distinctively different from service users mobilizing themselves to make a social impact. Rather, it becomes a case of service users being activated in welfare organizations' internal evaluation, development, and quality management. Much of the tension between the psychiatry organization and the user movement described in the analysis can be understood as a tension between the democratic and managerialist approach to user involvement, where the actors' aims, ambitions and purposes conflict. However, these potential conflicts are never acknowledged in the national policy (see ibid.). Instead, all actors' interests are implied to be realizable at the same time, through collaboration. This formulation of user involvement contributes to confounding the user movement with the public administration by implying that they are partners that share the same goals and objectives, while the conflictual position in relation to the state-crucial to many definitions of social movements-is left out of the equation.

Co-optation Theory
The theoretical framework of co-optation has been widely used to understand the dynamics between a larger organization (corporations or authorities) and the grass-roots movement (NGOs or social movements). Through a study on the Tennessee Valle Authority and its relations to coalitions of local farmers, Selznick (1949) established the concept of co-optation in institutional theory. Selznick defines co-optation as 'the process of absorbing new elements into the leadership or policy-determining structure of an organization as a means of averting threats to its stability or existence' (p. 259). Through processes of co-optation, organizations can avoid overt conflict by incorporating external, less powerful actors in 'collaboration'. Baur and Schmitz (2011: 12) points out that co-optation is not about an organization reacting to external pressure to change, but rather about taking control over the entire change process in a way that makes it comply with organizational interests. The transformation of service user involvement into a political and organizational concept corresponds to this line of thought; today the public administration runs and controls the user involvement initiatives, inviting the user organizations to participate in terms dictated by the administration.
A co-opting relationship must be understood in terms of power. Following Foucault's reflexive understanding, power is exercised through relations and all actors exercise power (Foucault 1975). Yet, the positions in the societal web of relations are unequal and some actors are-by the strength of discursive power or by representing dominating institutional logics (Douglas 1986)-in more advantageous strategic positions to exercise power than others (Foucault 1975). In the case of co-optation, the larger organization incorporating external actors usually represent that advantageous position, exercising its power by influencing external actors to align with the dominating logic of the organization (Coy and Hedeen 2005). Nevertheless, several scholars have opposed a rigid understanding of co-optation as a social movement being either co-opted and thereby neutralized or not co-opted and thereby freely able to express criticism and create change. A more nuanced image is required, where it is recognized that the incorporated party indeed might be able to influence from within the co-opt relationship (see Burke 1968;Austin 1972). As Sunesson (1989) notes, the key issue for the incorporated movements is how far they can manage to maintain selfdetermination and an independent agenda within the coopting relationship.
Like co-optation theory, theorization of corporatist exchange (see Ö berg et al. 2011: 368) highlights that interest groups adapt to state regulation and moderate their critique in exchange for the possibility to influence policy. However, theories concerning corporatism primarily account for the structures and prerequisites of the social movements-state relationship on political policy level. Thus, for this article, the main advantages of co-optation theory are that it focuses on the processes of incorporation, that it accentuates the inquiry of power-asymmetries and that it is suitable for organizational-level analysis.

User Involvement as a Co-opting Relationship
Even though the relationship between the Swedish state and the service user movement has been conceptualized in terms of co-optation (e.g. Sunesson 1989;Meeuwisse and Sunesson 1998), the phenomenon of user involvement has not. Nonetheless, the practice of user involvement has fundamental features that makes it suitable for analysis in terms of co-optation (see Pilgrim 2005;Martin 2011). Cooptation deals with situations in which external grass-root actors are involved in the policy-determining structures of a larger and more powerful organization with which a potential conflict of interests exists, which is basically what user involvement is all about. The user movement enters this relationship hoping to achieve changes, at the same time adapting to the general premises for the activities set by the public organizations. Meeuwisse and Sunesson (1998) recognize that: … co-optation often requires that the [user] organizations adapt to demands and expectations [from the state]. For example, the user organizations are treated as a group of experts on their own situation, rather than a counterweight or counterpower (ibid., p. 177, my translation).
I have shown elsewhere (Eriksson 2015) that the practice of user involvement creates a co-opting relationship between public organizations and the user movement based on four distinctive features: (1) a strengthened bond between the parties, (2) the public organization's possibility to frame and (3) control the joint activities, and (4) the possibility for the user movement to influence from within the public administration-to exercise a 'sanctioned resistance'. The aim of this article is not to demonstrate once again that the relationship between the public administration and the user movement is a relationship of cooptation-this is, rather, the starting point of the analysis. Instead, this article uses co-optation theory to increase the understanding of how and why the 'collaboration' between the actors can have certain (harmful) effects on the user movement. Using co-optation theory to analyse the microlevel interactions between employees and user movement representatives-played out through the practice of user involvement-creates an opportunity to understand the processes of incorporation that affect the user movement.

Method and Empirical Material
The analysis draws on a broad ethnographic case study on service user involvement within a large public psychiatry organization. The organization is one of Sweden's largest psychiatric administrations. It has approximately 3000 employees, with facilities throughout the county providing care across the entire spectrum of mental health to both adults and minors. In line with national political policy, the organization has a clearly stated ambition to 'implement service user involvement at all levels in the organization'. Voluntas (2018) 29:832-843 835 The fieldwork lasted for approximately one and a half years as part of my dissertation project (Eriksson 2015), and the study aimed to investigate user involvement at organizational level. Hence, the analysis is delimited to those kinds of user involvement initiatives that invite service user representatives to influence organizational development, ruling out influence on political policy (structural level) or individual treatment (individual level). The data analysed consists of field notes from participant observations, where I followed activities defined by the organization as working with user involvement. Common to all activities observed was that the user movement, through representatives of local user organizations, was invited to participate. In sum, the material contains approximately 600 A4 pages of field notes from observations on 37 separate occasions. 4 The material was coded using the Atlas.ti 7 software, and the analytical method resembles a thematic analysis (Willig 2014) conducted digitally instead of using printed material. Following the ethnographic approach, the analysis was guided by what took place in the field (ibid., p. 128), and through the analysis it became evident that the involvement activities not only enabled the user movement to influence the psychiatry organization, but also enabled the organization to influence the user movement.
To ensure anonymity, all personal names in the presentation are fictitious and names of places are not specified. Since user involvement has not been investigated on an individual level, no part of the study has run the risk of interfering with the practice of psychiatric care to individual patients. All observations were performed openly, meaning that I at every separate occasion of observation presented myself and the research project, asking those present their informed consent to partake in the study. The study was carried out as an autonomous research project, but with the approval of the administration of the psychiatry organization.
The analysis presented is a single case study, which means that it gives a thorough examination of one instance of user involvement practice (cf. Flyvbjerg 2004). Hence, the conclusions concerning how this particular example of user involvement affects this local part of the user movement cannot be directly generalized to all instances in which user involvement is implemented and also not to effects on the Swedish user movement as a whole. The study gives in-depth insight into how the relationship between a public administration and the user movement can unfold, and how the practice of user involvement can affect the user movement. Moreover, the study is situated in a Swedish welfare setting and the relationships between the state and civil society movements might differ in other national settings. The analysis shows that public initiatives in user involvement might (in some cases, in some contexts and in some ways) have disadvantageous effects on the user movement-something worth bearing in mind when working with user involvement in other contexts (cf. Alasuutari 1995).

The User Involvement Relationship Scrutinized
When organizational-level user involvement is implemented in local welfare organizations, the local user movement is typically invited to participate. In the study, about ten different local service user organizations were commonly involved by providing representatives who participated in different kinds of user involvement activities within the psychiatry organization. This included participating in working groups developing organizational policy, attending user councils and other dialogue forums, being involved as educators in staff training on user perspectives, managing support groups, and acting as 'user perspective consultants' or 'supervisors' to staff groups. The user organizations involved were basically the local divisions of the same organizations that gathered in the national network of user organizations (NSPH, described above). This local user movement generally understood the user involvement activities as an opportunity to pursue the movements cause and promote service users' issues. Common to the local user organizations criticism of the psychiatry organization included accessibility problems and under staffing; lack of psychotherapy and other longterm conversation-based treatment; overuse of medication; too vivid and too instrumental use of the diagnostic system; lack of holistic and recovery-based perspectives on psychiatric care (see Borg and Kristiansen 2004); as well as problems with staff members degrading views and treatment of individuals suffering mental health problems. Two or three of the user organizations had more fundamentally alternative perspectives on mental health and treatment, conflicting with the clinical, biomedical discourse prevailing within the psychiatry organization.
However, when formulated rhetorically within the psychiatry organization, it was repeatedly stressed that user involvement is about cooperation, partnership, and mutual understanding between the psychiatry organization and the user movement. A typical example of this rhetoric is captured in the following field note extract, where the head of department within the psychiatry organization opens an annual meeting between clinic managers and user organization representatives: Thomas (head of department) welcomes everybody to the meeting. He says that the day will be devoted to discussing questions concerning the organization and its services. He believes that there are many 'shared interests' that they can all agree on.
-I see this day as an opportunity to muster a 'joint force' to drive development forward through constructive dialogue […] So, I'm looking forward to the discussions and I hope that we can put forward many concrete proposals. We can then use these and continue working together to make things better. Thomas uses the term 'we' on many occasions in his welcome. He stresses the word and looks around the room, nodding towards representatives from the psychiatry organization and the user organizations alike, opening his arms in a gesture to show that 'we' means professional psychiatry and the user movement together. (Field note extract, dialogue meeting) The widespread consensus rhetoric amplified the understanding that the psychiatry organization and the user movement shared the same interests and strived for the same goals-a conception of user involvement that contributes to obscure the boundaries between the public administration and the social movement, linking them together through the construction of joint efforts, shared objectives, and mutual gain. As the analysis will show, this dominating consensus perspective affected the user movement's potential as a critical voice through discouraging extensive criticism that would pose a threat to the 'mutual understanding'. Moreover, within such a strong consensusdiscourse it might indeed be difficult to perceive that some aspects of user involvement might actually disadvantage the user movement.

Affecting the Constitution of the User Movement
In the administrative rhetoric within the psychiatry organization, it was argued that user involvement strengthened the user movement, following the logic that user involvement initiatives had the potential to mobilize new participants to the user organizations. Some of the activities defined as user involvement also aimed specifically at supporting user movement recruitment: such as inviting user movement representatives to wards to present their organizations or making rooms available within the psychiatric premises where service users could meet and organize.
But if the prevailing picture within the psychiatry organization was that user involvement initiatives supported the user movement, members of the user movement had a more nuanced understanding. At a meeting with the user council-the most formalized forum where highranking user movement representatives met with the administrative management of the psychiatry organization-some user organizations raised the concern that the extensive user involvement activities initiated by the psychiatry organization could affect the user movement negatively. Below is a transcribed extract from a discussion on a joint education programme called the 'Increased Dialogue Initiative' (IDI). IDI was constructed as a large-scale in-house training programme led by the psychiatry organization in collaboration with the user organizations, where representatives from the user movement would convey a patient perspective and train employees in patient-centred care. 5 The initiative was intended to reach all clinical staff within the psychiatry organization. The transcript starts with Magnus (a user organization representative) commenting whether the user movement should participate in the joint project proposed by the psychiatry organization: Magnus: What I'm a bit afraid of is that, more generally, we as [user] organizations must think about how to use our extremely limited resources. I'm not thinking about economic resources, but human resources. Several others: *affirmative sounds* Magnus: Because this is, actually, a bigger problem. Those few individuals can't sit on an endless number of boards and committees and participate in all sorts of other projects [within the psychiatry organization]. Ultimately, we will lose our own [members]. Someone: Exactly. Magnus: So, the question is whether we really have the time and resources [to participate in the education program], I don't really think we have. /…/ I don't think we should work with that. Kasper (psychiatry organization representative): Yes, it's possible that this could take energy from more independent, more critical [work]… That could happen. Magnus: Yes, from our own projects, our own goals, own… Ö rjan (user organization representative): From certain things that we want to carry through on our own. (Transcription, user council) Scott (1990: 6) stresses that the key resource of any social movement is the mobilization of individuals. The central concern raised by Magnus was that public user involvement initiatives might affect that fundamental resourcethereby reducing the movements' capacity to run their own 5 See Eriksson and Jacobsson (2016) for a detailed account of the training program. Voluntas (2018) 29:832-843 837 autonomous agenda. The empirical material indicates that Magnus's worries were, at least partly, justified. As the user organizations later decided to take part in the IDI, some of their members were recruited as tutors in the programme, which claimed a significant amount of their time. In similar ways, several other user involvement activities-such as partaking in organizing conferences, conducting evaluations, and participating in various working groups and committees-connected user movement representatives to processes of service development within the psychiatry organization, leaving them less time to engage in the user movements autonomous projects. Hence, many of the user involvement activities affected the constitution of the social movement by tying its members closer to the psychiatry organization. This process had both physical/social dimensions [i.e. user representatives working more with(in) the psychiatry organization] and ideational dimensions (i.e. user representatives connecting to the conceptual framework of the organization). The IDI programme can continue to serve as an example. Despite the doubts raised by some representatives, the user organizations decided to join in a collaboration with the psychiatry organization, managing the programme together. Similar staff training programmes had earlier been managed by the user movement alone, which is a critical topic discussed at a user council meeting: Viktor (user organization representative): When we [the user movement in their own training programs] are out talking, we give our picture of reality. But how is our view going to be expressed in this kind of joint project? That is an interesting question. (The room falls silent.) Kasper (psychiatry organization representative): Well… Viktor: That's the question.
Kasper: Well, I should clarify. The aim of this project is not to give the user organizations' interpretation of the problems within society or Swedish psychiatry… Viktor: No, exactly. Kasper: …it is just people expressing their own experiences [of psychiatric ill health], or their experiences of being a relative of persons suffering from, for example, schizophrenia. The idea is to give increased knowledge, tell about their lives, create contact, talk to people. And anything can come up in these discussions, of course. But it is not supposed to be a mouthpiece for the user movement.
[…] just to be very clear. Viktor: Yes. And it's not a mouthpiece for the psychiatry organization?
Kasper: No. What it is… it's just the experiences of individuals. (Transcription, user council) The representative from the psychiatry organization clarified that the project was not supposed to give the perspective of the user movement. Yet the psychiatry organization wanted the user movement to partake in running, and particularly staffing, the project-which they did. This demonstrates how user involvement initiatives, through the rhetoric of collaboration, can result in members of the user movement mobilizing less within the movement's autonomous initiatives (the prior education programmes, run by the user movement alone), and instead working more closely with the public administration following its institutional logic. The situation can be interpreted as a trade-off, because the user movement could hardly have managed such a vast training programme on its own. Participation in the IDI made it possible for the user movement to reach a large numbers of welfare employees, and in exchange the user movement agreed to downplay their political and ideological agenda. Likewise, many other user involvement activities implemented requested service users' individual experiences of psychiatry and psychiatric ill health, rather than the voices of user movement representatives. Hence, even if the individuals involved were engaged through the user movement, in many instances they were not expected to act primarily as movement representatives. This can be understood as a neoliberal, individualistic notion of service user involvement infringing on the collective perspective of the user movement.
The psychiatry organization's user involvement initiatives also affected the constitution of the user movement by formally recruiting leading user representatives from the user organizations, employing them within the psychiatry organization to implement user involvement. The fieldwork showed several examples of this. Such recruitment could be interpreted as the organization taking user involvement seriously, and the appointments were usually a great step for the recruited individuals personally. At the same time, this procedure might drain the user movement of important cohesive forces and key representatives, now instead actually employed by the state administration (cf. Selznick 1949;Coy and Hedeen 2005).
Moreover, some user involvement initiatives had the potential to prevent potential members from ever mobilizing collectively within the user movement. Instead, their will to participate could be absorbed by the psychiatry organization directly. As noted, in many involvement activities, service users' personal experiences were requested rather than the collective voice of the user organizations. In such activities, the welfare organization could very well bypass the user movement altogether, single-handedly selecting and asking service users considered as suitable to participate. As a representative from the psychiatry organization put it, the organization had its own 'pool of representatives' to contact when needing user representatives. 6 In the competition over user representatives' time and commitment, the psychiatry organization had the advantage of being able to remunerate the service users for their participation, while the user movement usually expected their members to work on a voluntary basis. This could tempt present members of the user organizations to engage in initiatives led by the public administration rather than in the movement's autonomous operations. Moreover, it could tempt potential members of the user organizations to work directly with the psychiatry organization, preventing them from joining the user movement. Alexandra-a user representative not affiliated to the user movement but often invited by the psychiatry organization to talk about her experiences of psychiatric care-spoke to me before one of her talks at a training session on 'Case management'. She told me that she was a bit stressed: She doesn't know if she will be able to keep her job as a pre-school teacher. It's tough, she says, because she really wants to stay. But if she loses her job, she's been thinking about starting her own business, getting contracted by psychiatry organizations to give this kind of lectures on user perspectives, and get involved in the user involvement. She has heard from others that you can make quite good money giving lectures and writing books about your experiences. (Field note extract, case management education) The extract, yet again, reveals individualizing (as well as commodifying) tendencies within user involvement. Instead of joining the user movement, individuals can 'sell' their personal experiences and expertise directly to the public organization. These service users, engaged in user involvement without being connected to the user movement, lack the support and autonomous ideological foundation that a social movement provides, and a strong tendency in the empirical material was that service users not tied to the user movement were less critical when participating in user involvement activities, working more in line with the rationales of the psychiatry organization.
Given that there are a limited number of individuals to engage, with a limited amount of time to invest, an extensive increase of user involvement initiatives run the risk of creating a situation where those individuals work more with the psychiatry organization's initiatives, and less with the user movement's autonomous initiatives. Through the co-opting relationship, power works to tie the participating user movement representatives more closely to the welfare administration, incorporating them in the dominating institutional practice and logic. Simultaneously, they are to varying degrees decoupled from the user movement. Hence, public user involvement runs the risk of fragmenting the service user collective-either by detaching user movement members from the movement's autonomous ideological framework and independent projects or by preventing some service users from getting involved with the user movement at all.

Affecting and Directing the Voice of the User Movement
The ties between the public psychiatry organization and the user movement discussed above have a physical and social character, where user representatives (and indeed in some sense the whole user movement) get more closely connected to the public administration. But as touched upon, the tightening bonds also had ideational dimensions. The increased contact between the psychiatry organization and the user movement enabled the welfare organization to influence the user representatives, making them accept, adapt to, or even take over, the dominating organizational logic. Such conceptual or ideological bonds affect what questions and demands user representatives raise, as well as how they raise them-disciplining them to act within the boundaries of the 'acceptable' (McKay and Garratt 2013)-thereby impacting how the user movement operated and performed.
A significant characteristic of the observed interactions was that the user representatives engaged in user involvement were socialized to act in a specific way: not to be too critical or far-reaching in their demands. This socialization (or disciplining) was achieved through conveying the psychiatry organization's depiction of user involvement, either through subtle comments or by directly telling or training the users how to act. For example, at a user council meeting, Thomas (a high-level manager in the psychiatry organization) gave his view on the development of the relationship between the psychiatry organization and the user movement: Thomas: When I first took up my appointment, I remember a meeting with the user organizations. It felt like being back in the 1970s. It was all arguments and opposition, accusations flying through the room. And I thought, where the heck am I? The managers and doctors became defensive, and the user organizations where just mad, criticizing the psychiatry organization. But since then, I feel we've found ways to work constructively with questions of user involvement. And I want it to continue that way, in a constructive manner, where we can develop psychiatry together, accomplish things together, because we have a lot of interests in common. (Transcript, user council) The transcript clarifies the psychiatry organization's expectations on the user movements when they engage in user involvement. They are not supposed to be as critical or 'radical' as they were back then. Instead, they are expected to be 'constructive' and focus on those issues where it is possible to find common ground. The psychiatry organization also held a recurrent (however not compulsory) 2-day 'communications training' for all new user representatives. The training focused on one specific form of user involvement: individual service users 'telling their story' to change negative attitudes towards service users and motivate employees to take on a user perspective (see Eriksson 2013). A significant element of the training was to encourage the user representatives not to be too critical. The education leader concluded the training programme with the following words: -If there is one thing that's important, no matter what situation you are in out there talking to people, it's this: 'Be a good example!' and 'Focus on what has helped you!' (Field note, communicators training) The structure and content of this communications training both directed the user representatives to carry out user involvement in a specific way-through attitude discussions with clinical staff rather than through policy discussions with organization management-and affected how they were to present their message-through individual narratives told in a pleasant manner.
Apart from how to act, the psychiatry organization could also direct what topics the user movement pursued. To exemplify, at another user council meeting, a critical external audit of the psychiatry organization-which highlighted four areas of improvement-was to be discussed. The manager leading the meeting (Thomas) started the discussion with a 15-min talk, giving his view of the report. While the user movement usually criticized the psychiatry organization on all four topics of the report, Thomas only acknowledged one of these ('documentation and care planning'). Concerning the other three, he gave vivid reasons to explain why the audit was mistaken. After the introductory speech by Thomas, the entire 30-min discussion that followed centred on the possibilities of improving the matter of documentation and care planning, while none of the other three topics was mentioned. Finalizing the discussion, Thomas urged the user representatives to continue highlighting that specific part of the audit: Thomas: It would be perfect for me [sic!] if you wanted to reconnect to this audit report [in future discussions], and if you do, focus on this topic, the one concerning documentation. Because on all other three topics we will defend ourselves. (Transcript, user council) This example shows how the top-manager directs the discussions towards certain 'allowed' issues-the 'common ground'-also explicitly establishing the kind of criticism that will be tolerated by the psychiatry organization.
Concerning what topics to pursue, another distinct expectation of the user representatives was to make 'realistic suggestions'; meaning that changes proposed should not be too profound. If deemed 'unrealistic' by the employees, the suggestions were usually dismissed as not possible to realize. At a conference gathering user movement representatives and psychiatry organization managers, the participants were divided into smaller, mixed, groups to develop ideas for service development. The moderator described the exercise like this: -View your group as a team! It's about working together to win the prize! When you work, try to use the clinic managers as a resource, because they know what's possible. What resources are available? What is realistic? If you're from the user organizations, please give your ideas, but remember that the suggestions should be viable, and this is where the managers can contribute. (Field note, user conference) The extract clarifies the rhetoric concerning realistic suggestions as well as the fact that it was up to the welfare administration to judge whether a question was realistic and adequate for further discussion (cf. Hodge 2005). Many user representatives adopted this rhetoric as well. At another meeting, a group of user movement representatives were gathered to comment on the plans for a new psychiatric hospital. During the meeting, the participants were divided into smaller groups and asked to decide on the three most important issues to change with the new hospital. In the group in which I was sitting, the discussions started like this: Fredrik (user organization representative) says that he is used to these kinds of exercises from other user involvement activities, and the key is to present realistic suggestions. What can be achieved in practice? If we aim too high, it's likely that nothing will get done, he says. The others in the group nod in assent. (Field note, meeting concerning a new psychiatric hospital) Here, the user representatives deviated from the task to decide on the most important issues to change, instead restricting themselves to issues believed 'realistic' to change. This can be interpreted as a conscious strategy to enable at least some influence. At the same time, it is noteworthy that the restriction to considering only 'realistic suggestions' is something that Fredrik had learned through his prior work with user involvement and later used to conduct self-discipline within the user group. In this way, the institutionally established boundaries that determined the range of user influence acceptable within the organization affected the user representatives, causing them to lower their expectations of what changes were possible. In the words of Coy and Hedeen (2005), the social movement's '''dreams of justice, dreams of peace'' have been considerably scaled down […] only to be replaced by more modest goals and measures' (ibid., p. 418f). If such ideological influences on the user movement representatives are strong enough, they might even cease to comprehend visions alternative to the ones formulated by psychiatry.

Concluding Discussion
The main arguments of this article have been that the practice of public service user involvement risks having a harmful impact on service users' capacity to mobilize, organize and act autonomously in the civil society, and that a positive consensus-discourse surrounding service user involvement serves to obscure this fact and to enable such a harmful impact.
The psychiatry organization's extensive ambition to create user involvement demanded and attracted the attention of many user movement representatives, and the case study shows that this can affect the constitution of the user movement by tying user representatives closer to the welfare organization. The service users were to varying extents disconnected from (or never gained access to) the social movement's collective action frames (Benford and Snow 2000) or collective identities (Melucci 1989), which threatened to fragmentize and weaken the user movement's potential for cohesive autonomous action. This development was interlinked with the tendency of the psychiatry organization to individualize organizational-level user involvement, where several activities requested user representatives' personal experiences, while downplaying the role of the user movement as a collective voice. Just as Meeuwisse and Sunesson (1998) note, the user organizations were often treated as a source of (individual) expertise rather than a socio-political counterpart, and being entangled in the co-opting relationship made it difficult for the user organizations to renounce this individualization, while they at the same time became directly exposed to it. This tendency to individualize collective action and organizational-level user involvement can be understood against the background of broader liberalizing and individualizing trends that permeate contemporary society (see Bergh and Erlingsson 2008;Bauman 2001).
The analysis also shows that public service user involvement can affect the user movement by directing the movement's attention towards certain issues (i.e. towards subject matters accepted by the welfare organization), and by influencing the movement to act and operate in certain ways (i.e. in cooperation with the authority in harmony with the organizations institutional logic). The service user representatives were socialized to act in a 'reasonable' and 'active but not activist' manner (Clarke 2013: 214). Moreover, as Marian Barnes et al. (2007: 94) concludes, the user representatives were also expected to be 'realistic' and show understanding towards the limited capacity of the psychiatry organization to change. From a critical co-optation perspective, this form of impact of public administration on social movements-governing and disciplining their autonomous, critical voices-might be understood as co-optation at work. Following Stickley (2006), what happens can be understood as the dominating psychiatric institution exercising power towards the user movement, influencing the social movement to align itself with the dominating regime. If conflict and opposition from a position outside of the institutionalized administrative system are considered as defining features of social movements, the user movement's character as such might even be questioned. Some suggest that strong affiliation with (and incorporation into) the state results in a social movement ceasing to be a social movement has instead become an institutionalized part of the political order (cf. Touraine 1982), or at least something different than it was before (such as an 'interest group', Costain 1981). 7 As the case study indicates, some aspects of public user involvement initiatives can affect the ability of the user group to organize and act autonomously within civil society (cf. Näslund et al. 2017). From a social movement theory perspective, such mobilization is vital to a democratic society, because it allows critical scrutiny of governments and power elites (Tilly 1999: 257), enables the inclusion of excluded groups (Scott 1990: 150), promotes (or resists) social change (Della Porta and Diani 2006: 21) and endorses renewal of the political discourse (Eyerman and Jamison 1991). For the user movement, autonomous mobilization is vital to enable the movement to influence social and political issues that cannot be changed by addressing the welfare administration on an organizational level. This is because, whereas social movements are free to pursue any question they perceive necessary-usually aiming at political impact and more fundamental reform of policy, practice, or discourseservice user involvement is inevitably constrained by the public's institutional logic, and typically formulated in terms of in-organizational quality adjustments, rarely enabling (or even actively avoiding) in-depth organizational rearrangements (Forbes and Sashidharan 1997: 485;Hodge 2005;Eriksson 2015). When user involvement in political and organizational discourse is formulated as an issue of in-organizational quality, it steers the user movement's attention away from the political arena, directing its efforts to influence the organizational level. This could be interpreted as a sign of declining or changing corporatist structures (cf. Ö berg et al. 2011), and a 'de-politicization' of questions that are, in fact, inherently political (see Clarke 2013: 216). If the user involvement initiatives at the same time weaken the user movement's potential to mobilize for social and political change, the prospects of more exhaustive improvements for marginalized and vulnerable groups in society decrease.