Employees’ resistance to users’ ideas in public service innovation

Correspondence Line Margrethe Jenhaug, Department of Social Work and Guidance, Inland Norway University of Applied Sciences, 2624 Lillehammer, Norway. Email: line.jenhaug@inn.no Abstract This article contributes to the existing criticism of the positive emphasis on user participation in service innovation as co-creation by examining employee resistance to user-driven innovation. The empirical base comprises interviews, document studies, and observations from a project that focused on implementing a user-initiated idea in public care services in Norway. To discuss employee resistance to innovative user ideas, a power perspective is included by drawing on the Foucault-based theory of identity regulation and discourse. Employees resist the required identity regulation by distorting the initial innovative idea to align with their problem representations, which is facilitated by entangled discourses. The power relations embedded in the different parties’ subject positions emphasize how governing the user side is incompatible with being governed by the users. The article contributes to our knowledge of service innovation and the co-creation of value by demonstrating discursive mechanisms for twisting value propositions.


INTRODUCTION
In Western welfare states, user participation in service innovation is being promoted increasingly and conducted in both private and public sectors (Hasu, Saari, & Mattelmäki, 2011;Jaeger, 2011;Osborne, 1998;Osborne & Brown, 2011). This is a result of the realization that positive outcomes from innovation in services require user involvement to ensure user values. Inviting users into the innovation process is frequently referred to as user-driven innovation and is embraced for the purposes of ensuring services that are higher in quality, less costly, and more effective (Sørensen, 2012;Torfing, Sorensen, & Roiseland, 2019;Trischler & Scott, 2016;Von Hippel, 2005). Policy documents in Western welfare states also increasingly refer to users as co-producers necessary for the sustainability of services and the public sector (Jaeger, 2011;Jenhaug, 2018). Currently, research is focusing on drivers for and barriers to including users in public innovation at a system level and at an individual level as resource integration and value propositions from different stakeholders. User-driven innovation presupposes the acknowledgement of users as 'experts-byexperience', as Meriluoto (2018) called them. Public innovation with user input is placed into a cocreation paradigm, where innovation represents the highest level of co-creation as it includes and activates users (Torfing et al., 2019). However, a growing body of literature examines the phenomenon of co-creation in a power perspective as a neoliberal governmentality discourse in marketing (Beckett, 2012;Beckett & Nayak, 2008;Cova & Cova, 2012;Zwick, Bonsu, & Darmody, 2008), which has yet to mature within public sector research and is not included in the service innovation literature. This article considers user-driven service innovation from the perspective of this critical research on co-creation in the private sector.
A few contributions have provided insights into the difficulties related to implementing usergenerated innovative ideas (Magnusson, Matthing, & Kristensson, 2003). Sundbo (2011) found that employees tend to 'other' users who suggest service changes, thus distancing themselves from the users and demonstrating an unwillingness to adapt to users' wishes. As user-driven innovation demands employee involvement and changes in practices, resistance to users' ideas can express reluctance to modify professional roles or professional identities as decision makers and sole experts. Foucault (1982) theoretically linked resistance connected to identity to power relations, constituting a gap in the field of critical studies of user involvement in service innovation. This article challenges the harmonizing view of the user's role in user-driven innovation by exploring how employees resist user-driven innovation from a power perspective.
The analysis is based on an in-depth study of the implementation process of a user-driven and innovative idea in care services. In this study, family carers are represented as users by virtue of their role in relation to end users. Family members are increasingly seen as important co-producers of public welfare services (Pestoff, 2012;Realpe & Wallace, 2010), and how such services are organized and delivered will have important consequences for the care the family members provide (Jenhaug, 2018). The aim of the idea studied was to give family carers greater influence in their collaborations with care services on a daily basis.
In the next section, I will first further outline the theory of user-driven innovation and service innovation and frame the idea of giving family carers more influence as a potential innovation. In order to understand employees' resistance to user-driven ideas, power/knowledge and the associated concepts of resistance and identity regulation are introduced. Second, the empirical study will be explained in greater detail, followed by a description of the methodological and analytical approach. The findings and analysis will then be presented, followed by a discussion that aims to offer a practical and theoretical contribution regarding resistance to user-driven service innovations. A short conclusion summarizes the content and suggests further research.

USER-DRIVEN SERVICE INNOVATION
Research has concluded that users' input is necessary during service innovation processes (Henkel & Von Hippel, 2004;Karlsson, Skålén, & Sundström, 2014;Von Hippel, 2005). Users have even been described as the 'driving force' of innovation (Sundbo & Toivonen, 2011, p. 4). To contribute to innovative results, Agger and Lund (2011) highlighted the importance of giving users a role as co-producers and public innovators, which means giving them the responsibility, in addition to the right, to innovate in the public sector.
Public sector innovation research is inspired by the marketing tradition in service innovation, which has introduced value co-creation as a distinct concept in the innovation process (Osborne, 2018). Cocreation is defined as 'joint collaborative activities by parties involved in direct action, aiming to contribute to the value that emerges for one or both parties' (Grönroos, 2012(Grönroos, , p. 1520. Skålén (2018, p. 93) defined service innovation as 'the development of new or existing value propositions that have a positive impact on the customer's value creation'. Whether the effort involved in change is considered an innovation or not is, therefore, dependent on the user's or the customer's considerations of the outcome of the innovative effort. Giving family carers more influence with regard to daily service production for end users can be considered a new value proposition that, presumably, could have a positive impact on family carers' value creation. This gives the idea that is studied innovative potential.
Conceptualizing a service logic, users' experiences of a service are considered important to a company and, in order to secure value for the users, they must be included as partners during service development (Grönroos, 2008;Grönroos & Voima, 2013;Vargo & Lusch, 2004, 2008. In public sector research, the user's conceptualization of value is also emphasized as the main element of innovation (Osborne, Radnor, & Nasi, 2013;Skålén, Karlsson, Engen, & Magnusson, 2018), arguing that public sector sustainability is dependent on the value experienced by service users (Alford, 2016;Osborne, Radnor, Kinder, & Vidal, 2015). To enable users, citizens, or third parties to be co-creators of value, services have to recognize what individuals value and provide it. Skålén et al. (2018, p. 704) emphasized problem-finding and problem-solving as the most important factors for co-creating public service innovation: By being involved in problem-finding and problem-solving practices, intra-organizational actors, such as frontline employees, and extra-organizational actors, such as users, may detect problems with existing value propositions and offer solutions resulting in ideas that may develop existing value propositions or create new ones.
To understand employee resistance, this article explores the mechanisms of implementing value propositions based on problem-finding and problem-solving in a power perspective. Sørensen and Torfing (2011b) emphasized the need for different perspectives among the parties involved in order to ensure the quality of innovative ideas. Furthermore, they stressed framing innovation across organizations as being especially difficult because the process requires that roles and ways of thinking have to change (Sørensen & Torfing, 2011a). If users are given more influence in practice, then professionals may be required to downplay their identities as experts, which has been found to be difficult. The study by Sundbo (2011) demonstrated the hindrances of user-driven innovation related to employees' roles as professionals at a café. Taking on a professional role resulted in the employees emphasizing the differences between themselves and the users, as well as the perspective that they had knowledge and expertise that users did not have. This led to their reluctance to ask for or act upon innovative ideas that originated with users. This indicates that position and identity as a professional play a significant role with regard to resisting user-initiated ideas, and that both pave the way for a power perspective in the study of user-driven service innovation. The next section outlines a Foucauldian-informed view of how to study resistance to user-driven service innovation.

POWER, IDENTIT Y, AND RESISTANCE DURING SERVICE INNOVATION
When services promote users as innovators and include them in innovation processes, employees' identities as experts need to be regulated towards seeing these users' expertise as equally important, or as what Hasu et al. (2011, p. 252) described as 'a new type of expertise and expert identity' for professionals. Foucault (1982) related identity regulation and resistance to power constructions based on knowledge and conveyed through discourse. Based on Foucault (1991), discourse can be defined as the difference between what you 'could' say during a specific period and what actually becomes institutionalized as the truth and spoken by a social group. The power of discourse is dependent on the knowledge it conveys, with power and knowledge being interdependent. Discourses affect people's identities; as Foucault (1991, p. 58) put it: 'Discourse is not a place into which the subjectivity irrupts; it is a space of differentiated subject-positions and subject-functions'. In other words, discourse affects subject positions rather than the opposite. Some individuals are given subject positions as 'users' and others as 'professionals', which also -to a certain degree -determines the behaviours of both groups. Some discourses are accorded a higher status than others, whereas some positions from which discourse is conveyed are, according to Bacchi (2009), more influential than others, for example the media, politics, and science. Consistent with a social constructionist epistemology, a person's identity will always be regulated by discourse reflecting certain beliefs and truths, depending on access to knowledge and the kind of knowledge that is accessible. In this way, discourses, or regimes of knowledge, embed power as they produce identities, and 'Discourses exert power because they transport knowledge on which collective and individual consciousness feeds' (Jäger & Maier, 2009, p. 39). Through policy, science, and the media, for instance, a society will convey a discourse that is believed to serve the common good. Identity regulation is conducted through techniques of governmentality, which are concretized as societal evaluations, punishments, and rewards, called disciplinary power, and through the inner yearning to satisfy societal expectations, for example by means of scrutinizing oneself, confessions, and receiving guidance, called technologies of the self (Foucault, 1988). Skålén (2018) concretized the fact that 'power manifests itself through identity construction'. One governmentality technique described by Foucault (1982, pp. 177-178) is 'dividing practices', that is objectivizing subjects into categories more or less attractive. McKay and Garratt (2013) identified 'dividing practices' regarding the responsibilization of parents in schools. They found that schools divided parents into the categories of good and bad, with 'bad parents' being recognized as opposing the established truths to which the schools related, and 'good parents' accepting the problems represented by the schools. The schools aimed for parents to participate the 'right way'. Similarly, the co-creation discourse is, according to critical marketing theory, a governmentality technique aiming for customers' resource integration by constructing an attractive, active customer identity and helping them to make choices that are beneficial to the firm (Beckett, 2012;Beckett & Nayak, 2008;Cova & Cova, 2012;Zwick et al., 2008). Considering user-driven innovation as value co-creation and as the implementation of value propositions based in problem-finding and problem-solving , these findings of co-creation as a governmentality technique form a background for discussing employees' resistance of identity regulation. Järvinen (2004) and Villadsen (2003), in agreement with Foucault, described employees helping users make good decisions as 'power mechanisms'. The profession of care worker is constituted through the user's need for guidance, whereas knowledge of users' weaknesses, needs, and wishes makes employees capable of leading them towards the welfare state's ideals. As professionals are given the role of guiding users into responsible subjects, they are, through discourse, placed in a superior subject position compared to service users. Nevertheless, during user-driven service innovation processes, employees have to acknowledge users as 'experts-by-experience', which challenges employees' subject position as supervisors.
According to Foucault (1976Foucault ( , 1980, identity regulation and resistance against it are embedded in the nature of power relations. Instances of resistance 'are spread over time and space at varying densities, at times mobilizing groups or individuals in a definitive way, inflaming certain points of the body, certain moments in life, certain types of behavior' (Foucault, 1976, p. 96), and resistance is described as potentially utilizing the same techniques used for regulating identities. Covaleski, Dirsmith, Heian, and Samuel (1998) found that, in organizations, resistance to governmentality techniques of identity regulation from organizational leaders is rooted in subject positions, such as those of autonomous professionals, and that this resistance is an exercise of professional power. Institutionalized subject positions may, therefore, form the basis of resistance to new subject positions. Such resistance is likely to emerge in cases where the new subject position counters the institutionalized one.
Hence, putting individuals or groups of people in a structural position as innovators does not automatically mean that they will be in a position to make changes because the structural position does not necessarily reflect the parties' subject positions; as Foucault stated, 'Relations of power are not in superstructural position[s]' (Foucault, 1976, p. 94).
Tensions between parties' identities as experts, in relation to power positions, are relevant when examining how employees might resist user-driven innovations.

Empirical site
A centre for family carers (Pårørendesenteret i Oslo; here translated as the Family Care Centre in Oslo (FIO)) has developed, in collaboration with the health department of Oslo, an innovative idea and helped implement it in diverse municipal care services. The FIO is led and organized by family carers, and both the FIO and the municipal health department representative will, henceforth, be referred to as 'the initiators'.
The aim of the idea was to improve the collaboration between family carers and municipal service employees on a daily basis by seeing the two as equal partners -as well as experts -with regard to the needs of end users.
The municipal service leaders were introduced to the idea of 'enhanced collaboration with family carers' as 'taking the family carers more seriously'. The service leaders agreed to implement the initiators' idea, and employees were introduced to it by the initiators at workshops that included teaching, role-playing, a 'toolbox' of forms presented and given to employees, and discussions about specific cases concerning collaboration. The aim was to regulate the employees' practices and attitudes towards taking family carers' perspectives. In addition, a family carer representative was present and discussed her experiences of having a child with a mental disability. For role playing, the employees played the parts of family carer and an employee in pairs, practicing dilemmas they might encounter based on notes describing both positive and negative experiences. In advance, the employees had submitted cases describing cooperation with family carers that were to be discussed with the initiators at the workshops. The toolbox for the employees to use was developed by the FIO centre and included forms to complete with the family carers, including a consent form, a form regarding what to talk about at the first meeting (the family carers' duties and needs and their recommendations for the end user), a form about what to do in emergency situations, and one for mapping the end user's social network. Additionally, the toolbox had an introduction and a guide for dialogue with family carers. The health department representative conducted the instruction using a PowerPoint presentation and focused on the same topics as those addressed by the toolbox: preconditions for cooperation, clarification of roles, dialogue, network, and consent. Together, the instructional component and the toolbox were developed to inform employees about ways to include family carers, and the role playing and discussions aimed to help the employees identify ways to include family carers in practice.
The implementation process is, technically, putting the family carer representatives in positions as experts and innovators. The question to be considered is whether these positions give the initiators influence in practice. This article will discuss whether the implementation process was a success compared to the initial idea and aims to determine what factors influenced the outcome by interviewing the employees following the workshops.

Methodology
The empirical study took place between 2016 and 2018, and a mixed-methods approach was used that included individual interviews, focus group interviews, observations, and document analyses with regard to the implementation of the innovative idea (see Table 1).
To identify the discourses determining the current identities, I found it important to ask questions about the participants' perceptions in the form of semi-structured interviews. The initiators were asked what they would like to change, how they had collaborated to create the concept, and how their experiences of the workshops had been thus far. Questions to the employees were, for example about experiences during their day-to-day collaborations with family carers, what they would like to change, whether the workshops had influenced their thoughts, and what they remembered from them. Each individual and focus group interview lasted about an hour, after which the interviews were transcribed. All the interviewees were informed about the project and signed a participation consent form; the employees, the leader of the FIO, and the health department leader were informed of the potential for them to be recognized because of their positions. The workshop observations lasted approximately 3 hr; I noted what happened, transcribed some parts of the discussion, and sketched out several analytical thoughts. Those taking part in the workshops agreed to be observed.

Data analysis
To understand resistance as being related to identity and dependent on discourse, for this article, critical discourse analysis (CDA) is used, based on Foucault's discourse theory. This CDA is aimed not only at analysing what is happening but also how and why it is happening. Inspired by Foucault, discourse analysis was conceptualized by Bacchi (2009) as the 'what's the problem represented to be' approach. Inspired by this approach, I analysed how solutions are grounded in certain problem representations through the discourses present. Although Skålén et al. (2018) emphasized service innovation in terms of problem-finding and problem-solving, Bacchi insisted that parties often have different problem representations, with the solution being based on power positions among the stakeholders. Thus, the mission of the analysis is to shed light on the problem from different perspectives, as I have done in this article. As described by Alvesson and Kärreman (2007) and Samuels (2000), the analytical approach is inspired by abduction, being a combination of deduction and induction. I had a pre-understanding of user-driven innovation and employee resistance, as well as an understanding of the lack of a power perspective, and this basis in theory, combined with the inductive elements of scrutinizing the empirical material, was aimed at developing new knowledge of employee resistance to innovative ideas developed by the user side.
All data were categorized and analysed using NVivo 11. To examine employee resistance, I first coded the utterances, observations, and documents into 18 categories, for example 'good cooperation', 'bad cooperation', 'performance of care work', 'the municipalities' responsibility', 'information from family carers', and 'information to family carers'. In scrutinizing identity regulation and resistance among employees, it is critical to identify their conceptions regarding the family carers before and after implementation of the idea, compared to the family carers' conceptions of themselves. A close reading of the coding and analysing the empirical material as a whole, in line with Carter and Little (2007), revealed that the different nodes could be placed into three discourses with regard to family carers: 'family carers as experts', 'family carers as difficult', and 'family carers as responsible'. Among the participants, these three discourses appeared to various extents on different occasions, as shown in Table 2. The results of the analysis will be outlined in the Findings section; all quotations in that section were translated into English by the author.

FINDINGS
This section begins with an analysis of discourses that represent the initiators and the employees 'before and during' the workshops (where implementation of the innovative idea of family carer influence was attempted) and is structured according to the three discourses that appeared. Then, these findings are compared with the discourses among the employees 'after' the workshops as a basis for discussing identity regulation and employee resistance to the user-driven and innovative idea. These findings will be discussed in the next section in relation to earlier critical contributions of co-creation from a user perspective.

Family carers as experts
The interviews with the two initiators and the observation of one meeting between them made it clear that the FIO and the health department representative agreed on the aim of the workshops, that is, to implement the idea of enhanced collaboration by encouraging employees to give family carers more influence. This safeguarding of family carers' perspectives and values enables the idea to be called 'user driven'. Further examination of the process of developing the idea is not included here as the issue is the implementation of the idea into the services.
The idea concerning enhanced collaboration by giving family carers more influence is expressed in terms of being taken seriously as an equal partner or being regarded as experts. The family carers' problem is that they are often ignored as sources of knowledge about the persons in need of care. The innovative concept of giving users more influence is aimed at addressing this problem. Improved cooperation will, therefore, take into consideration family carers as experts in order to recognize and integrate their knowledge and, thereby, provide beneficial services for users. Accomplishing this calls for changing the employee/professional's role to one as a facilitator of services. According to one staff member at the FIO, 'The thing is that health staff have new roles, but they don't understand this. They're not authorities anymore, but cooperators, facilitators, secretaries'. The emphasis on employees having less authority requires identity regulation among these employees and aligns with the goal of giving family carers more authority as experts, which is what the initiators aimed to convey to the employees as the core of the innovative idea during the workshops.
At the workshops with the service employees, the discourse about family carers as experts is expressed by the initiators. Sometimes this is vague, as in the following comments: 'Why not do as the family carers sometimes want you to do?' 'We need each other, employees and family carers'. Other times, it is expressed more explicitly, as in this comment by the FIO leader: '[The employees] have to understand that the family carers' knowledge of the user is important for the job they're doing…'. One of the forms included in the employees' toolbox instructs them to ask about the family carers' wishes and how the family carers would prefer to collaborate, thus demonstrating a 'family carers as experts' discourse.
The initiators see the problem in terms of the family carers having too little influence during service provision. If the employees agree and respond to the solution to the problem in terms of seeing family carers more as experts, then they will be demonstrating their willingness to regulate their identities.

Family carers as difficult
Even before the workshops, the employees revealed their perception of the problem as troublesome collaborations with family carers, with their aim being to acquire help addressing and preventing these conflicts analysed from the cases they sent to the initiators in advance: … The new primary contact and some other employees are described as incompetent [by the family carer], and some employees are receiving harsh feedback from this family carer. There have been several meetings between leaders of the unit and family carers to try to improve cooperation, without success…. How can we prevent these situations, and how should we handle things when such situations occur?
From a distribution office, the following case was submitted:

… This family carer clearly says that the family is dissatisfied with the service in regard to follow-up and activation. The user is closely followed by specialists regarding different diagnoses. The family carer regularly asks for access to and printouts of the journal, and we and the medical supervisor spend a lot of time going through questions dealing with interventions and reports. The family carer is perceived to be critical of the staff and of the doctor's conception and follow-up of the user's health status.
It seems as though the family carers are considered problematic when they take on the roles of 'experts by experience' and supervising the professionals. Conflicts may escalate when family carers are not treated as equals and when services are not provided according to the family carers' advice.
At the workshops, the cases were raised as planned, and the employees again expressed their frustration with regard to difficult family carers. They also guided each other in solving the problems, for example 'Just listen to the family carers; don't do anything, just take the time to listen' (housing service employee). Other advice given by a leader was to sit down and talk with all the residents at a home for persons with disabilities in an effort to demonstrate to the family carers that the employees are involved in the residents' lives.
The family carer representative who was hired told her story about cooperating with the housing service and also described difficult family carers, emphasizing how they should behave in order to forge good relationships with the employees. The employees were nodding along with these statements. Here, the family carer representative confirmed the employees' experiences with difficult family carers.
This emphasis on family carers as difficult, therefore, represents the employees' problems and confirms their superior subject position.

Family carers as responsible
Both sets of stakeholders emphasized that family carers are responsible for performing care work. The FIO talks about the responsibility of family carers as something that the welfare state has taken away from them: '…the welfare state is overloaded when it comes to what it can offer, and there will be nicer and warmer care if we're given back some family care initiatives' (FIO leader). During the workshops, the initiators argued that family carers could perform some of the work in users' apartments, and the above utterance, explaining that family carers and employees need each other, can also be interpreted to mean that family carers should be responsible for parts of the care work. The initiators portrayed these responsibilities in terms of being close to the user and, therefore, able to offer value that the services cannot.
However, according to the initiators, the family carers need help and facilitation: 'They need special leave arrangements, relief care opportunities and recognition for their work' (FIO deputy). The initiators emphasized that it is the employees' responsibility to identify 'how the family as a whole handles the situation and how it has affected them' by asking questions about their daily lives, finances, feelings, spare time, and support networks.
Likewise, the employees expressed that family carers are responsible. Several of the cases submitted in advance by the employees reflected on family carers in terms of their being responsible for care work, as in the following from a distribution office:

How can we collaborate better with family carers after public services have started to help? How can we convey the thought that public services can never replace the care and social contact that family carers provide and make them trust in us [also] taking responsibility?
The employees need the family carers to take responsibility as 'family', not by interfering with how the service is performed. One employee described that a mother who was frustrated and lacked trust in the services needed to be reminded of her role: 'I told her directly that [the resident] needed her as a mother. The mother's attitude changed in about half a year'.
Seen through the lens of the 'family carers as responsible' discourse, the problem for family carers is a lack of help to manage the responsibility, whereas for employees, the problem appears to be a lack of initiative for care work among some of the family carers. The employees are in agreement with the family carers on one point, namely that they are irreplaceable and important to the users and their well-being. The responsibility discourse is coherent with the governmental emphasis on family carers with regard to their responsibility for the end user (Jenhaug, 2018).
To summarize the findings so far, three discourses on family carers are represented by the two parties: family carers as experts, family carers as difficult, and family carers as responsible. These discourses include corresponding problem representations of too little influence, too little initiative, and too many conflicts. 'Family carers as responsible' is a common discourse with regard to family carers being irreplaceable and differing in relation to the problem representation: too little facilitation versus too little initiative.

Employees' stance after the workshops
To learn whether the employees had begun to regulate their identity towards seeing the family carers more as equal partners after the introduction of the innovative idea, the employees' stance following the workshops was analysed. Changes were identified; for instance, they talked about how to avoid being provoked by difficult family carers. To quote one employee at the housing group: 'It is strengthened, the idea that they have something to say, and that you have to think about it all the time, that the information the family carers convey, it is not just to bother me'. Another said: … there's always a reason [unpleasant outbursts by family carers], you have to remember that. If not, you're trapped, and it becomes problematic. Irritating…No matter how stupid it is, whatever they say, you have to just say back 'I'll take it further, I'll look into it'.
Thus, the employees came to a better understanding of why the family carers act as they do, and they saw the importance of maintaining good relationships with family carers. However, the employees' remarks also revealed that, in practice, they do not necessarily see the family carers as experts who possess important knowledge that should influence the service the end users receive any more than it did before. The family carers' interference in how to manage the needs of the end users still seems to annoy the employees, as this comment makes clear: If an employee has trouble with an activity and the family carer accompanies him or her to show the employee how the activity can be accomplished, this is also experienced as irritating by the employees. Regarding this, the employee quoted above added, 'But I think she means well'. Another employee at the housing service said: [The family carer] reads a lot of research from the US. She thinks they've come much further with their research into [the diagnosis] there, and treatment and things, but we don't read those research articles or work that way.
This quote indirectly states that family carers behave like experts, but employees do not acknowledge their expertise. Sometimes they act on it, but not necessarily because they have trust in the family carer:

I could certainly have interrupted and said, 'You know, [the user] is going to eat normal food', but the victory I can take from that is so small compared to destroying the relationship, so…you have to decide if it's worth it. (Employee at the housing service)
When asked about the main message of the workshops, a representative of the distribution office replies '… building good relations for both parties … in that way, you can provide much better support'. To build good relations, the staff at the distribution office implements techniques such as 'empathetic communication' that describes a 'stairway' for how to listen to family carers and when to interact with them. Thus, they are expressing a perspective of good support for the end users that relies on good relationships with family carers, but one that is not necessarily conditional on seeing the family carers as experts and equal partners but rather aiming to build family carers' trust in the services. 'Building good relations', therefore, seems to have different meanings among the stakeholders: being considered experts with influence among the initiators and, for the employees, building trust in the services.
The employees hold on to the discourse of family carers as responsible. Building good relationships and gaining family carers' trust when providing services to end users is helpful when it comes to allocating tasks between the services and the family carers. One employee at the distribution office put it this way: We have experiences of family carers who first say they cannot take on anymore [care work], and then, when we're having a conversation with the family carer and the user and we start getting deeper into it, the family carer says, 'Oh yes, shopping, I can do that, of course', and suddenly there are other tasks they can do…. It's about the way [the service providers] engage with them.
The distribution office staff emphasize the significance of information to and from family carers. 'We give them the information in a considerate way and accept them and … you'll get further then, I think', said one employee at the distribution office. They also allow family carers to call them at any time.
Collaboration with family carers, according to the employees, has improved. The family carers are described as a resource by the housing service whenever they buy items for the user, accompany the user to a medical appointment, or bring the user home for an overnight stay. These statements suggest that family carers should, according to the employees, take responsibility for performing different types of tasks such as care work for their loved ones. However, the employees' attitudes towards the family carers seem to have changed and become more empathetic and understanding of why family carers are sometimes difficult. By mitigating family carers' urges to make decisions or, in other words, be considered an equal partner, the techniques of listening, being available, and providing information aim to avoid deeper conflicts. As problem solutions, these understandings seem to have improved collaboration with family carers post-implementation, according to employees, but not as intended by the initiators. The solutions still reflect difficult and responsible family carers as the represented problem to be solved, contrasting the initial innovative idea based on the problem of too little influence for family carers and protecting the employees' professional identity. According to the definition of service innovation as dependent on the user's value creation, whether or not the innovative idea is implemented in real-world service innovation remains unclear.
However, the differing problem representation and, thus, the transformation of the user-driven innovative idea represent resistance to the initial user perspective of the original innovative idea and pave the way for discussing employee resistance in a power perspective.

DISCUSSION -UNDERSTANDING RESISTANCE TO USER-DRIVEN INNOVATION
This article contributes insights into how employees may resist users' ideas in public service innovation, here discussed in relation to power mechanisms. When the stakeholders describe enhanced collaboration, the three discourses -family carers as experts, family carers as responsible, and family carers as difficult -and their equivalent problem representations all frame the collaboration between family carers and services as deficient and in need of improvement. A common aim among the parties representing change is enhanced collaboration; however, the representations of the collaboration problems differ between the two groups. At the workshops, the initiators' mission, according to the aim of their idea, is to guide the employees towards enhanced collaboration by considering family carers as experts. Nevertheless, the initiators' idea of 'employees considering family carers more like experts' has resulted in 'making family carers trust the services' expertise'. Therefore, the outcome of the innovative idea does not seem to be what was initially planned. Against the backdrop of these findings and the Foucault-based theory of power, I will further discuss and explore how employees are able to resist user-driven innovation.
Implementing the idea through workshops brings the three discourses together. The problem representations of too little influence from the family carers' perspective, too little trust in services from the employees' perspective, and the common problem of family carers having too little responsibility are expressed, while they all aim for enhanced collaboration, thus obfuscating the differing problems. This obfuscation via the overarching concept of enhanced collaboration represents an entanglement of discourses, which Jäger and Maier (2009), based on Foucauldian discourse theory, called a discursive knot. The common emphasis on enhanced collaboration and both parties' encouraging responsible family carers represents the centre of the knot, obscuring the differences in the problem representations. The three discourses coming together and fitting into an overarching concept of enhanced collaboration help the employees interpret the aim of 'giving family carers more influence' as 'helping family carers to trust the services', thereby confirming the employees identities as experts.
In a Foucauldian power/knowledge perspective, the implementation based on one side's interpretation is enabled through subject positions. The employees who are guiding each other at the workshops are an example of what Foucault called the technology of the self; they are reinforcing their subject positions as superior experts despite the initiators' structural position as leaders. Regulating their identity towards seeing family carers as equal partners would conflict with their new role of professionally helping and guiding the user side by regulating the users' (here: family carers') identities, according to Järvinen (2004) and Villadsen (2003). The findings indicate the maintainance of this new role following the implementation, for example when employees emphasize typical governmentality techniques as listening, understanding, and building trust in order to guide the family carers. Holding on to the discourse of difficult and responsible family carers protects and confirms their subject position as professional experts. Thus, this article confirms the findings of Covaleski et al. (1998) and Skålén (2009) that subject positions of professionals overrule the technical, hierarchical power relations established, enabling their resistance to identity change.
An overview of discourses regarding family carers spurs, as well, 'dividing practices' as outlined by Foucault (1982, pp. 177-178) at an overarching level. The category of 'difficult family carers' refers to family carers who are opposed to the way things are typically done and demanding unrealistic services. 'Difficult family carers' conflicts with the positive description of 'responsible family carers' performing care work and, thereby, providing relief for the services. 'Responsible family carers' are encouraged by supervising and trustbuilding, aligning with the findings of schools', dividing parents into two categories, as identified by McKay and Garratt (2013). The governmentality techniques of creating the responsible customer revealed and framed in the private sector's marketing research concerning co-creation with customers (Beckett, 2012;Beckett & Nayak, 2008;Cova & Cova, 2012;Zwick et al., 2008) are here confirmed in public service innovation research. The dividing practices disciplining and leading family carers in order to co-create value seen as advantageous to the public service firm are simultaneously ignoring the family carers' values as worthwhile in service innovation.
Building on knowledge about the co-creation of value propositions as service innovation emphasized by Skålén et al. (2018), this article frames the governmentality techniques found as a practice of 'twisted value propositions'. The user-driven, innovative idea of increased influence as a value proposition based on the problem of lack of trust in family carers' expertise is here twisted into the employees' value proposition of creating trust in the employees' expertise. Twisting value propositions might, therefore, be seen as an act of obscuring the rejection of users' values in service innovation. This article suggests that the act of service innovation with users is a matter of co-creation of value propositions as problem-finding and problem-solving, including complicated identity regulation and resistance actions that, potentially, result in twisted value propositions unfavourable to the user side.
The article, therefore, highlights that user-driven innovation in services might be a risky venture regarding the user side's influence in practice when user values do not correspond to the values that the employees emphasize. Thus, this article, by taking into consideration a power perspective, complicates the promoted importance of including various perspectives among the parties involved in innovation processes.

CONCLUSION
This article contributes to the criticisms of service innovation theory by questioning users' opportunities to exert an influence, in practice, by studying employees' resistance through the Foucauldian lens of identity regulation and power/knowledge.
The article has pointed out that a structural facilitation of user empowerment during service innovation does not necessarily secure equality in practice because of employees' resistance to identity regulation based on subject positions. Different discourses represent problems differently, and an entanglement of the discourses when implementing innovative ideas facilitates interpretations towards solutions that align with the employees' urge to protect their identity as professional experts whose role is to guide the user side of service provision.
The interpretive technique of protecting the professional identity in service innovation is here conceptualized as 'twisted value proposition'. 'Twisted value proposition' is, in turn, framed as a technique based on power relations according to Foucault, advancing the theory of innovation as value co-creation with users.
Relevant for service innovation practice, this article's findings suggest that, if user-driven innovative ideas are implemented as service innovations, problem representations have to be scrutinized and agreed upon -not only on an overarching level but also in detail. The article argues in favour of expanded attention in theory and practice, from how to include users and take advantage of user knowledge and information in service innovation to what acting upon users' ideas will demand of employees and whether or not they will agree to pay the price. Identifying hindrances supports the inclusion of users in positive ways when innovating service. Because modulating employees' identities as experts has been demonstrated to be difficult when users are included in service innovation, further research should consider service innovation as identity regulation. There is also a need to examine the relevance of seeing results of various types of user input as 'twisted value propositions' in service innovation, including a broader range of public services and with regard to co-creative service activities without the aim of innovation.
We know that service innovation is not accomplished by means of one or two interventions, and, most likely, neither is regulation of professional identity. Thus, it is difficult to say whether this project and the changes seen among the employees may represent a starting point for seeing family carers more as equal partners, which would be a radical innovation in public care services.

ACKNOWLEDGMENTS
Many thanks in particular to the FIO centre for being so positive, to the representative of the health department in Oslo, and to all the employees for participating. I am also thankful to Professor Ole Petter Askheim and Professor Per Skålén for giving me a lot of feedback. The project was given ethical approval by the Norwegian Centre for Research Data.