Four features of cooptation – User involvement as sanctioned resistance

The article draws on a three-year ethnographical study investigating how “service user involvement” was constructed (i.e. understood, implemented, and performed) within two large Swedish welfare organizations – a county-based psychiatric organization and a municipal social service administration (see Eriksson 2015). When analyzing the interactions between the user movement and the welfare organizations, a relationship much like cooptation (Selznick, 1949) was revealed. The article outlines four characteristic features of this coopting relationship: (1) The bonding between the parties, incorporating the user representatives in the organizations and their institutional logic; (2) The organizational framing of the user involvement activities; setting the initial rule for how to act/speak, where to act/speak, when to act/speak as well as what to speak about; (3) The organizational control exercised as the activities took place, directing the discussions and interaction to align with the interests of the welfare organizations; and (4) The resistance exercised by user representatives, enabling them to influence the organizations and contribute to change. Together, these four features disclose service user involvement as a “sanctioned resistance”: At the same time as the institutionalized service user involvement controls and constrains the way service user representatives act and pursue their goals, it gives them a possibility to challenge the welfare organizations from within. However, the influence that is permitted can be understood as adjustments within the prevailing institutional logic, rather than changes that transformed the organizations in more profound ways.


INTRODUCTION
Service user involvement is a popular concept within contemporary western welfare administration, and takes many shapes and forms (Eriksson, 2015;Alm Andreassen, 2018).Its roots can be traced to the democratic aspirations of the civil rights moment and ÅRGANG 3, NR.1-2018, S. 7-17 ISSN ONLINE: 2464-4161 the demand for influence put forward by the user movement in the 1960s (Sunesson, 1989;Crossley, 1999).Incorporated into current political discourse, the idea of user involvement is also infused by the new managerialism of public services, which have been implemented to modernize, evaluate, and make welfare administration more efficient (Croft & Beresford, 1992;Martin, 2011).The political rhetoric of user involvement is optimistic, promising improved services, conditions, and positions for service users.The practice of user involvement may indeed render beneficial outcomes, but such an alluring concept must be critically investigated.There is an inherent contradiction in political policy, where there is an aspiration to increase user influence and freedom of choice, at the same time as social assistance is increasingly conditioned and the social sector is expected to cut costs (Davies et al., 2014;Andersson et al., 2015).Several researchers have noted that user involvement serves to legitimize the welfare administration, while leaving limited room for service users to exert true influence (see e.g.Forbes & Sashidharan, 1997;Böcker Jakobsen & Mik-Meyer, 2001).
User involvement has been theorized in terms of, for example, governance and disciplining (McKay & Garratt, 2013), power relations (Hodge, 2005), professionalization (El Enany et al., 2013) and participatory spaces (Näslund et al., 2017).The aim of this article is to highlight how the concept of cooptation can be used to understand the relationship between the public administration and the collective of service users that the practice of user involvement produces.The article also increases our understanding of the theoretical concept of cooptation.
The analysis draws on an in-depth ethnographical study of how organizational-level user involvement was realized within two large Swedish welfare organizations.The practices were predominated by activities characterized by "logics of democracy" (Alm Andreassen, 2018), where user movement representatives participated in organizational policy and development processes.The activities can be understood as "interest politics" or "collective-voice activities", thus contrasting expressions of user involvement taking the shape of collaborative service provision, so-called "co-production" (ibid.).However, even though the activities were based on a fundamentally democratic logic, the activities were permeated in practice by managerialist/consumerist aspirations, and by the uprising discourse that underpin the logic of co-production that stress the importance of mutuality and shared responsibility (see Askheim et al., 2017).

COOPTATION THEORY
Philip Selznick (1949) established the concept of (formal) cooptation in institutional and organizational theory.Selznick investigated the relationship between the Tennessee Valley Authority − a federal organization governing agriculture in the Tennessee Valley − and the local farmer community.He defined cooptation 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 cooptation, a larger and more powerful organization can moderate external criticism and take control over processes of change by inviting collaboration with external, less powerful parties with whom a potential conflict of interests exists.Selznick showed how the incorporated actors began to adapt to, or even internalize, the visions and logics of the organization (see also Baur & Schmitz, 2011), without having a substantial impact.Several scholars have subsequently argued that the external party may be able to exert influence from within the coopting relationship (Burke, 1969;Sunesson, 1989).Hence, cooptation should not be understood as a static situation of neutralization -rather, the strength of the coopting mechanisms depends on the nature of the relationship between the parties.Nonetheless, cooptation theory is typically used to understand compliance and institutional stability (see, for example, Pilgrim, 2005;Martin, 2011;Baur & Schmitz, 2011), and the concept has been widely applied to conceptualize the relationship between the state and the voluntary sector.Coy and Hedeen (2005) analyzed cooptation as a process of homogenization, during which voluntary sector agendas are gradually taken over and transformed by the state in order to conform with government policy.As Coy and Hedeen, many scholars (including the author) understand cooptation primarily as an institutional process.Najam (2000), in contrast, understands cooptation as deliberate action and sees it as one of four possible relationships between the state and the voluntary sector -the others being cooperation, confrontation, and complementarity.Najam defines cooptation as a relationship in which a more powerful party "attempts to change the preferences that others have about particular ends and means" (p.388); cooperation is a situation in which the parties have a shared view of ends and means and act together to realize these; confrontation occurs when the voluntary sector falls into dispute with the state, and works autonomously to contradict or alter state policy or practice; and complementary describes the situation in which the voluntary sector works with its own strategies parallel to -but not in opposition to or in collaboration with -the state.
User movement cooptation usually requires user organizations to adapt to demands and expectations from the state (Meeuwisse &Sunesson, 1998), andMartin (2011) showed that the autonomy of the voluntary sector could be compromised by closer relationships with the welfare administration.Related to this, Sunesson (1989, p. 410;my translation) argues that: Cooptation should not be understood as a failure.Rather, what is important is the extent of self-determination that the user movement is able to maintain within the boundaries of a necessary cooptation.
Hence, autonomy is a key issue that we must consider in any analysis of user movement cooptation.If the user representatives can sustain an autonomous agenda, they may indeed be able to influence from within the coopting relationship.This article increases the understanding of how the processes of cooptation work: it explains why cooptation is likely to reduce conflict and engender compliance, and discusses opportunities for influence within a coopting relationship.

THE STUDY
This article draws on a constructionist ethnographic study that aimed to investigate how service user involvement was constructed (i.e.understood, implemented, and performed) within two large Swedish welfare organizations -a county-based psychiatry organization, and a municipal social service administration (see Eriksson, 2015).Over a two-and-a-half-year This article is downloaded from www.idunn.no.© 2018 Author(s).This is an open access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).period, participant observations were conducted within the organizations, following activities that had been defined by the organizations as working with organizational-level user involvement.The study was initiated and completed by the author with the approval of the organizations.All observations were performed openly, informing the individuals who were present about the study, and asking their consent to participate.The empirical material consists of approximately 1000 A4 pages of field notes from 208 hours of observation on 65 occasions.The study was guided by what took place in the field, and when the fieldwork was finalized a structured analysis of the data was conducted.Recurrent themes were identified and linked.Implementation of user involvement somewhat differed between the two organizations (see Eriksson, 2015, p. 278), but four comprehensive, reoccurring themes stood out and characterized user involvement within both organizations: (1) the bonding between the parties, and the processes of (2) framing, (3) control and (4) resistance.I interpret these themes as vital elements of a coopting relationship.Thus, the analysis led me to adopt the theoretical framework of cooptation.The article emphasizes the theoretical argument and analyzes user involvement using cooptation theory.Examples will be briefly described, but the scope of the article does not permit the empirical material to be presented at length.The full empirical grounds for the argument, however, have been presented elsewhere (Eriksson, 2015, see also Eriksson, 2018).

USER INVOLVEMENT AS A COOPTING RELATIONSHIP
The user movement can work "alongside, against or within government-sanctioned service-user forums" (Davies et al., 2014, p. 120), and when the user representatives agree to join institutionalized forums, they enter an "invited space of participation" governed by the welfare administration (Näslund et al., 2017).Such user-involvement activities create a specific kind of relationship between the user movement and the public administration − a relationship that it is possible to understand in terms of cooptation (Forbes & Sashidharan, 1997;Pilgrim, 2005).Indeed, the core idea of user involvement is to invite external actors (with whom potential conflicts of interests exist) to participate in organizational policy processes.In this case, the external actors that are incorporated is the collective of service users, typically represented by user organizations and their representatives.These actors regard the activities as an opportunity to influence and change the policy and practice of the welfare organizations such that they align better with service users' interests.The work presented here revealed that the practice of user involvement had four distinctive features that mirror four aspects of a coopting relationship.

Bonding
The first feature of user involvement that was identified was that it created connections between the user movement and the public administration -a process I call "bonding".The bonding was formal because agreements were made to conduct activities together, and it was physical-social because user representatives spent time with(in) the organizations, getting to know the organizations and their employees.These aspects of the bonding were the basis of the relationship, and they are inevitable if any activities are to be conducted.The This article is downloaded from www.idunn.no.© 2018 Author(s).This is an open access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).bonding, however, also showed significant ideational and emotional dimensions, in which user representatives started to adopt the dominating institutional logic of the organization and began to feel solidarity with it.
When user involvement was implemented, it was embedded in a strong consensus rhetoric -resembling the harmonious discourse that underpins the co-production perspective (see Askheim et al., 2017, p. 586, p. 599).User involvement was conceptualized in terms of "joint efforts", "collaboration", and "mutual understanding"; creating a sense of a common "we".This consensus rhetoric reinforced the ideational and emotional bonding, and encouraged user representatives to accept or adopt the standpoints of the welfare organizations (see further on these processes in Eriksson, 2018).For many (but not all) user representatives, the more time they spent with(in) the welfare organizations (i.e. the stronger the physical-social bonding), the stronger the ideational and emotional bonding tended to become.This was especially true for those user representatives not closely connected to the user organizations.The following extract is an example of ideational bonding.During a group discussion, David (a long-time user representative within the municipality) proposed how the user movement could help the social services: One thing the user organizations can do, David says, is to create realistic expectations.Many users within drug treatment have a serious misconception.They think they can just walk into the social office and ask for any treatment they want.This guy might have heard of a treatment that he really wants, and it might be the cheaper alternative as well, but you [the social services] can only offer the treatments you have contracted.Here, we can aid you by lowering expectations, telling this guy that it is not likely he will get the treatment he wants.

(Field note, open-space conference)
The extract shows how David has adopted the perspective of the municipality.Rather than speaking on behalf of the service users, arguing their right to obtain proper treatment of their preference, David acknowledges the need (of the municipality) to ensure that service users comply with what the social services can offer.However, while formal and physicalsocial bonding is inevitable, the strength of the ideational and emotional bonding can differ.Some user representatives maintained a strict alternative ideological agenda and avoided emotional attachment, even when they spent significant time working with user involvement.
In terms of cooptation, bonding can be understood as the process whereby the user representatives are incorporated into the welfare organizations.Through formal bonding, the user representatives acknowledge the right of the welfare organization to govern the activities, whereas the ideational and emotional bonding can be understood as processes in which power works to persuade the external actors to embrace the organizational logic.Bonding explains many of the conflict-reducing tendencies of a coopting relationship: the stronger the bonding, the more likely it is that user representatives will act in accordance with the organizational logic.Through bonding, the autonomy of the user representatives is reduced, and if the ideational bonding becomes sufficiently strong, they may even accept the institutional goals as their own (Coy & Hedeen, 2005, p. 418).

Framing user involvement
When user representatives agreed to participate in the involvement initiatives, they acknowledged the welfare organizations' formal authority over the activities (cf.Sunesson, 1989;Pilgrim, 2005).This enabled the organizations to set the overarching premises.Such "framing" of the activities regulated how user representatives were supposed to act/speak, where to act/speak, when to act/speak, and what to speak about.Concerning how to act, user representatives were expected not to be too critical or "radical" in their argumentation, as this could jeopardize the "collaboration" and "mutual understanding" of user involvement.It was made clear in training sessions on how to participate that user representatives were expected to moderate their voice, and by employees subtly or explicitly telling user representatives how to behave "reasonably".Concerning where and when to act, user representatives were invited to act in numerous spaces, but never in those spaces in which formal organizational decisions were made.Concerning what to speak about, every activity had a defined objective, which in turn defined what it was appropriate to discuss.For example, at a meeting on how to design a user survey within the municipality, the user representatives repeatedly raised issues that concerned how the social services functioned.Instead of taking this opportunity to hear the users' opinions of the services (which was exactly what the survey was supposed to investigate), the organization rejected these matters as "not the issue", as they did not fit the framing of the meeting.In general, the activities were also framed to concern topics deemed to be "realistic" -i.e.issues that the organizations assessed to be possible and/or appropriate to address.This tended to discourage the user representatives from demanding profound changes.
Together, the framing constituted a norm system for user involvement: it disciplined the user representatives to behave "acceptably" (McKay & Garratt, 2013) and delimited the invited space for participation (cf.Näslund et al., 2017).The user representatives accepted this norm system because it allowed them access to the organizations, even though it enabled the organizations to take command of the change process (cf.Baur & Schmitz, 2011).The framing was sometimes contested, but if the user representatives had rejected it entirely, the relationship would probably have ceased to exist (Najam, 2000).Hence, compliance with the framing was a vital part of the coopting relationship because it maintained its stability.

Controlling the activities
Framing the activities is a way to control user involvement in advance by setting the initial principles, but employees were also able to control the activities as they took place.As representatives of the organizations, the employees possessed a position of formal power.The employees appointed to implement user involvement, in particular, had far-reaching possibilities to control the activities.As formal administrators they were expected and obliged to govern the involvement initiatives in ways that benefitted the organizations.They exercised an administrative control by planning and managing the involvement activities, by controlling the agendas and meeting minutes, sometimes by presenting complete suggestions for how to move forward on an issue, and by choosing which user representatives should participate.For instance, administrative control was exercised when Johan -an This article is downloaded from www.idunn.no.© 2018 Author(s).This is an open access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).employee of the psychiatry organization responsible for the implementation of user involvement -raised at a user council meeting the issue of choosing user representatives to participate during a series of conferences on user perspectives: Johan (employee): In this case, there are certain individuals whom I would prefer, certain of your [the user organizations'] members.So, I will handpick them, the ones I think have a lot to contribute.
Anna (user organization representative): You do that.

Nils (user organization representative):
Yes, it is probably the best way to do it in this case.
Johan: Yeah, it doesn't work if you [the user organizations] appoint everybody who participates.

Anna: No.
Johan: That doesn't work for me.

(Transcript, user council meeting)
The user representatives rarely questioned the administrative control exercised by employees, since they considered it to be part of the employees' duties, rather than an expression of power.Even when employees attempted to share administrative control, it proved difficult, as the user representatives typically expected employees to lead the activities.In addition to the administrative control, employees also controlled the events by guiding the discussions in directions that suited the organizations.The user representatives also tried to direct the discussions to suit their ambitions, but the employees − who had the advantage of authority within the organizations − were usually more successful.
The control exercised by employees should not be interpreted as a calculated way to avoid user influence.Nevertheless, given the institutional setting, control was typically exercised in ways that made user involvement avoid issues that deviated too extensively from the interests of the organizations.And such exercise of control inevitably encouraged certain user voices and restricted others.Baur and Schmitz (2011) describe cooptation as a way for an organization to ensure that processes of change align with the ambitions of the organization, and this is typically what the control exercised by employees tended to do.

Resistance
A final distinctive feature of the coopting relationship was the opportunity of the user representatives to exercise "resistance".Here, the definition of "resistance" proposed by Trethewey (1997, p. 288) is used: "any behavior or discourse … that countered or disrupted the dominant bureaucratic discourse".As noted by Burke (1968), the coopting relationship may require concessions from the incorporating organization, allowing the user organizations the opportunity to influence the organizations from within.
Not all instances of user involvement were characterized by resistance.In some cases, the user representatives conducted activities in full agreement with the organization and in accordance with its logic.When interests differed, however, resistance was demonstrated, and its most common expression was argument or debate between the two sides.Even if certain processes worked to delimit and regulate the user influence, the invitation to par-This article is downloaded from www.idunn.no.© 2018 Author(s).This is an open access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).ticipate enabled the user representatives to oppose the organizations.Resistance typically took the form of argumentation during discussions, but alternative strategies, such as transgressing the framing or disobeying an agreement about how to participate, were used in some cases.An example of the latter arose when three user organizations that were partaking in user involvement within the psychiatry organization published a six-page press release that severely criticized the psychiatry organization for not considering users' perspectives.This caused the psychiatry organization to organize a crisis meeting with the user organizations, and tried to convince them that the organization indeed took their perspective seriously.At the meeting, however, the psychiatry organization employees also reproached the user representatives for their action: "You are supposed to keep an eye on us, " Göran [head of the psychiatry organization] says, "it's your function.But this is just six pages of negativity.You haven't taken any of the positive things we try to do into consideration.None.And then, at the end, you write that you still want to keep your chairs on our user council.I don't understand how to interpret that.Of course, we want you to continue on the council, but then you must be able to cooperate.We must be able to work together; you can't just go out throwing allegations about like this."

(Field note, crisis meeting)
It is clear that Göran perceives the press release as a violation of the agreement about how to participate.If the user organizations want to continue to participate in the user-involvement activities, they must refrain from such confrontational behavior as expressing public criticism.The user organizations eventually agreed to continue the collaboration.The psychiatry organization, in return, agreed to open new forums for user representatives to express their opinions.Hence, the example shows how confrontational forms of resistance -challenging the framing of user involvement and the very existence of the relationshipcan achieve substantial influence.
The opportunity to exercise resistance is vital to the coopting relationship.If no such potential existed, the user representatives would probably not participate.Since user involvement served an important function in this case to legitimize the welfare organizations (cf.Martin, 2011), and enabled them to control the situation, the organizations were eager to sustain the activities.Indeed, they granted an amount of user influence to ensure that the user movement were prepared to continue the relationship, and not regress to autonomous actions.

SANCTIONED RESISTANCE
A certain practice can simultaneously contain elements of reproduction and contestation of the current order (Trethewey, 1997), and the involvement activities in the study had such a paradoxical nature: Through bonding, influence and resistance were encouraged and requested at the same time as they were governed and suppressed through processes of framing and control.The analysis showed that user involvement can be understood as "sanctioned resistance", where "sanctioned" here is used in both of its meanings.The user representatives were invited and expected to exercise resistance -i.e. the resistance was This article is downloaded from www.idunn.no.© 2018 Author(s).This is an open access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).sanctioned (as in approved) by the welfare administration.This resistance, however, must lie within certain limits and fulfill certain conditions set by the welfare organizations -i.e. the resistance was sanctioned (as in disciplined).The user-involvement activities governed the processes of change, and transformed potential autonomous external resistance into a sanctioned, interinstitutional resistance.Najam (2000) understands cooptation as an inherently unstable relationship, which soon becomes either overt confrontation or consensus-based cooperation.However, the work presented here shows that cooptation can be an enduring aspect of the relationship between public administration and user movements -a fact that may be particularly true in Sweden, where the affiliation between the state and the voluntary sector has traditionally been characterized by closeness and reciprocity (Johansson, 2003).When analyzing user involvement, we should not consider cooperation, confrontation and cooptation necessarily to be distinctively different kinds of relationship.Rather, the relationship can be understood as simultaneously containing aspects of cooperation, confrontation, and cooptation, where the coopting aspects work to avoid uncontrolled confrontation by maintaining a relationship that is characterized by cooperation on terms dictated by the administration.
When able to maintain an independent agenda, user representatives have the opportunity to convince employees to adopt their viewpoint, and in this way challenge the dominating organizational logic from within the coopting relationship.Hence, user involvement can contribute to changes other than those initially fully sanctioned (allowed) by the organization.However, the further the user representatives' agenda is from the dominating institutional logic, and the stronger the intra-organizational forces that safeguard the current order, the more difficult it is for the user representatives to have an impact.In this study, the involvement activities sometimes engendered changes that improved service delivery and treatment.Yet, the influence was generally far from having a decisive impact on the welfare organizations or radically improving the situation of the service users.Political reform is necessary to achieve such an impact, and user organizations that aim to achieve fundamental changes should probably consider other forums than invited organizational spaces.The analysis further suggests that bodies in public administration that wish to create user influence should be critically aware of how they frame and control the involvement activities, consider the possibility to reduce any restricting aspects of the practices, and try to make it possible for the user representatives to maintain their autonomous position and agenda.

Compliance with Ethical Standards
The author declares that there is no conflict of interest.
This article is downloaded from www.idunn.no.© 2018 Author(s).This is an open access article distributed under the terms of the Creative Commons CC-BY-NC 4.0 License (http://creativecommons.org/licenses/by-nc/4.0/).