Morspark: Co-production as Expected?

(1)

Morspark: Co-production

as Expected?

How do mutual expectations affect

collaboration between co-producers and

regular producers?

2016

Frank Lekkerkerk, BSc Leiden University, Faculty of Governance and International Relations 9-6-2016

(2)

1 Master thesis Public Administration: ‘Morspark: Co-production as Expected?’

Leiden University, Faculty of Governance and International Relations By: F.J. Lekkerkerk BSc

Supervisor: C.J.A. van Eijk MSc

(3)

2

Preface

This research is the culmination of my Master of Public Administration at the University of Leiden. My interest in society and specifically in the mechanism of collaboration to build a vivid neighborhood made me choice co-production as research topic for my master thesis.

I would like to thank my supervisor C. van Eijk for her advice and support during the research process. I also want to thank all those involved in the Morspark project who participated in this research. Their involvement resulted in a wealth of information. Finally, I want to thank all those who supported me during the process of writing this thesis.

Delfgauw, June 9 2016,

Frank Lekkerkerk.

(4)

3

Summery

Previous research has suggested that more research is needed to embed the concept of co-production in empirical evidence. This study explores how mutual expectations influence the collaboration between regular producers and co-producers. The goal is to get insight in how the interaction between civil servants and citizens take place and to find out how they see each other’s and their own role.

To find answers to the main question: ‘How do mutual expectations affect collaboration between producers and regular producers?’ theoretical insights of co-production scholars (eg. Loeffler, et.al, 2013; Moynihan & Thomas, 2013; Brandsen, Pestoff, & Verschuere, 2012; Ostrom et al., 1978) were combined with theoretical insights on roles and collaboration in the public domain (Van Delden, 2009) and socialization (Rollag, 2012; Kelly et al. 1990).

This qualitative study inspired by the causal-process tracing (CPT) approach was conducted in April 2016 in a co-production project located in the Hoge Mors neighborhood in Leiden, The Netherlands. The project aimed to improve and redesign a public green space called Morspark. Data came from eight semi structured interviews with residents (5) and civil servants (3), and were analyzed using the CPT approach.

The findings demonstrate that in a case were intensive collaboration between public professionals and residents is quite new, expectations tend to be broad and vague. This creates a flexible structure of collaboration in which roles and tasks of participants can emerge. Roles and tasks are distributed on a voluntary basis. This may positively affect collaboration due to motivation, but also may have negative effects like slow progress and role and responsibility ambiguity. The analysis also showed that it may be difficult for residents to approach the municipality to become a partner of the project, especially when co-production is new for civil servants. Then participation of the municipality is likely to be heavily dependent on political embracement and individual civil servants who believe in the project and thus on the expectations of these civil servants. When a municipality becomes a partner of a project the co-production phase starts. The involvement of a municipality makes it possible to combine the financial and organizational strengths of a municipality with the practical knowledge and ideas of residents. When collaboration continues roles and tasks become more solid. Because responsibilities are shared between those involved and not as clear cut, trust is important to make collaboration possible. When collaboration progresses over time and roles and tasks take more solid forms, expectations become more concrete as well. This is likely to effect

(5)

4 collaboration because expectations now may be confirmed or disappointed. This fit or misfit of expectations may damage the trust of those involved in the co-production process. By direct and open information sharing and an open and flexible attitude of project members it is possible to adjust expectations to prevent misfits of expectations and their consequences.

Further research is suggested on the topic of the interaction between regular- and producers, the role distribution among regular- and producers, the definition of the co-producer and the topic of improving legitimacy of policies by involving citizens in the decision making process.

Keywords

Co-production, Public space. Mutual expectations, Roles, Co-producers, Regular producers, socialization.

(6)

5

1. Introduction

The collaboration between government and citizens to improve the quality of service delivery is also known under the heading of co-production (Brandsen et.al, 2012). Well known forms of co-production are for example neighborhood watches, public gardens and neighborhood homecare.

When looking at the current reform agenda, co-production is introduced in many countries (Osborne & Strokosch, 2013). Also politicians are using the concept of co-production in their policies more often nowadays. The Cameron government, a coalition of conservatives and liberal democrats, in the United Kingdom for example wants to empower individuals and communities, encourage social responsibility and create an enabling and accountable state. Prime Minister Cameron calls this the ‘Big Society’. A similar example can be given about the Netherlands. In his yearly speech on the third Tuesday of September 2015, King Willem Alexander spoke about the need for a ‘Participation Society’. Prime Minister Rutte, the leader of the coalition cabinet of the liberal (VVD) and Labour party (PvdA), wants to encourage citizens to take up responsibilities in society.

Although co-production is studied more and more, there are still some gaps in the literature. The role of the citizens in the co-production process, for example, is studied quite extensively (see for example Moynihan & Thomas, 2013; Brandsen, Pestoff, & Verschuere, 2012; Bovaird, 2007; Alford, 2002). The role of the civil servants and the interaction between citizens and civil servants, however, are scarcely studied. In addition to this, the empirical evidence “at the front lines of everyday practice” is thin (Fenwick, 2012). This study aims to further close this gap by using literature about collaboration in the public sphere (e.g., Van Delden, 2009) and intertwine it with co-production concepts. Collaboration can only exist between two or more parties; in the case of co-production these are civil servants and citizens. Because the focus of co-production literature is mainly on the role of the citizens it is

necessary to focus on the role of the civil servants as well. In the end they are both responsible for how the collaboration will work out. Previous research in different research areas shows that role expectations are perceived to be related to collaboration (Van Delden, 2009; Suter et al., 2009). Expectations civil servant have from citizens and the other way around may affect the way they collaborate. Therefore, the focus of this research project is on the collaboration process between citizens (co-producers) and civil servants (regular producers), and more precisely, on the mutual expectations both actors have. In this sense, the main question of the

(9)

8 research is: How do mutual expectations affect collaboration between co-producers and regular producers in a co-production process?

Taking into account the limited amount of literature available on this topic, the

research will have a more grounded, qualitative approach. Evidence will be drawn from semi-structured interviews. This evidence will be used to increase the knowledge of the interaction between civil servants and citizens, and to find out how they see each other’s role.

In this study the case of Morspark will be researched to find answers to the main question. Morspark is a subproject of Morslint, in which citizens collaborate with the

municipality and local organizations to improve a green space in the Hoge Mors district. They collaborate in a project group to design the park. Morslint is a citizen initiative to improve the public space in the Hoge Mors, this is a district located in the city of Leiden in the

Netherlands.

For researchers it is important to take into account that their research is useful for society. It is relevant for the society that civil servants do their jobs as well as possible. This research can contribute to that because it can provide information about the interaction and the roles of the two parties involved in co-production. The findings of this research may contribute to the improvement of the collaboration between civil servants and citizens. The theoretical relevance of this research is a contribution to the co-production literature on this topic. As mentioned before there are some gaps in literature and the findings of this research can help to close these gaps. The qualitative research method helps to create a deeper

understanding of the interaction between the government and the citizens. The introduction of role expectation literature into co-production literature contributes to the theoretic relevance as well.

1.1 Structure of the thesis

To introduce the readers of this thesis to the concept of co-production a brief historic overview and definition will be given in the first part of chapter two. That chapter will also point out that other theories than those published in the co-production field are necessary to find an answer to the main question. Chapter two also elaborates on the roles that involved actors play in a co-production process. The rest of the second chapter displays important factors of collaboration in the public sphere that may be influenced by expectations of the involved actors. In chapter three the research design is explained. It discusses the research approach, the case selection and selection of the respondents for the interviews which were used to collect evidence, how this evidence is analyzed and what the limitations of this study

(10)

9 are. In chapter four a description of the case is given. It shows how the citizen initiative of Morslint and the co-production project of Morspark are related and also gives a historic overview of important events that took place in these projects. The analysis of the evidence is discussed in the fifth chapter. This thesis concludes with the sixths chapter that provides an answer to the main question and a discussion of these findings, it also gives recommendations and suggestions for further research. Appendix A contains a list of interview questions that were asked to interviewees during the data collection of this study.

(11)

10

2. Theory

This chapter discusses literature that is of interest for this research. First a historic overview of the concept of co-production and a widely used definition of co-production are stated. Next, insights on the role of actors involved in co-production are provided. Theory on expected and enacted roles and role expectations is used to provide a framework that can be used to study the impact of expectations on collaboration a co-production process.

2.1 The concept of co-production

The concept of co-production was originally established by Elinor Ostrom and her colleagues in the 1970s (see Ostrom, 1972; Ostrom et al., 1978). Other researchers have continued to further investigate and develop the concept in the 1980s (see for instance: Percy, 1984; Seligson, 1980; Sharp, 1980; Whitaker, 1980). In spite of that, the interest on co-production decreased because marketization entered the public domain in the mid-1980s (Alford, 1998). In the 1990s, however, a new interest emerged, stimulated by authors such as Alford. These authors felt that the concept is of value in public management discussions as an alternative to marketization and the “contract state”. (Alford, 1998: 129). Alford emphasized the

importance of client co-production next to volunteer co-production of public services (Alford, 1998, 2002). The concept of co-production rose again as an essential part of New Public Governance discussions emphasizing networks and partnerships between citizens, public organizations and the third sector as well (Osborne, 2009, 2010; Pestoff, 2012; Verschuere et al., 2012).

The concept of co-production finds itself on the crossing point of different academic disciplines, which makes it an increasingly targeted research object by many scholars. This contributes to different views on what co-production is and who are involved. The way these different scholars operationalize co-production has differs contexts and involve different actors (Brandsen and Pestoff 2006; 2009). It is confusing to use these different understandings of co-production all at the same time; therefore distinctions between those different

phenomena are made by defining them with other terminology.

Citizen participation in the field of service provision, which is often referred to as co-production, is different from the phenomenon of co-management. With co-management different actors are involved. With co-management the “third sector participates, alongside other public and private actors, in managing the growing complexity of delivering diverse public financed services, without any direct citizen or user participation” (Pestoff,

(12)

11 2012:1106). This is not what is understood by co-production in this study, because, as will be mentioned in the working definition below, co-production only involves civil servants and citizens (Brandsen, Pestoff, & Verschuere, 2012:1).

Another distinction that has to be made is between the input (delivering information and ideas) and output (producing services) side of policymaking. Where co-production and co-management are found on the output side of policymaking, co-governance is usually found on the input side. Co-governance includes the third sector and other private actors in the decision-making of public policy for a given sector. These parties are represented and given voice and vote in citywide, provincial and/or national bodies to decide on the policy of a sector (Pestoff, 2012). This is also different than what is understood as co-production in this study, because it involves other actors as well.

As I showed above, production can be distinguished from management and co-governance by the actors that are involved. Below the working definition of co-production that will be used in this study is discussed into more detail.

Following previous work in the field of co-production, particularly in the tradition of Ostrom (1999), co-production can be defined as this widely accepted definition: "the mix of activities that both public service agents and citizens contribute to the provision of public services. The former are involved as professionals, or ‘regular producers’, while ‘citizen production’ is based on voluntary efforts by individuals and groups to enhance the quality of the services they use" (Brandsen, Pestoff, & Verschuere, 2012:1). Also other definitions of co-production are stated in literature. For example a definition by Pestoff (2012) who defines co-production as: “the collaboration between citizens and (semi-) public organizations”. This definition is less specific about the roles that involved actors have and on what basis they participate than the definition stated by Brandsen, Pestoff, & Verschuere (2012:1). Therefore the definition of Brandsen, Pestoff, & Verschuere (2012) of co-production will be used as working definition for the purposes of this research.

A production process involves all the activities that both regular producers and co-producers contribute to the provision of public services. The ‘co-production star’ (Loeffler, et.al, 2013) is a helpful tool to distinct the different activities that exist within a co-production process; it includes co-commissioning, co-design, co-delivery and co-assessment. On the one hand, these activities can be observed as a process wherein these activities follow-up on each other. On the other hand, some activities can be used to improve service delivery as

individual tools. For example, co-prioritizing can be used as an instrument to improve legitimization of policies of the (local) government, other co-production activities are not

(13)

12 specifically necessary to realize this activity. Co-delivery, to give another example, can also be independently applied: public organizations can ask citizens to participate in service delivery process without doing commissioning and design. In Box 1 all the separate co-production activities are displayed.

From: Loeffler, et.al, 2013: 25

As stated in the working definition of this research co-production involves two actors: regular and co-producers. The focus of this research is on the collaboration process between those two parties and more precisely on the mutual expectations both actors had during that process. As stated in the introduction previous research mainly focused on one of both actors

individually, and less on the interaction between the two actors. Co-production literature mainly focuses on the co-producer and on the role of the co-producer has in a co-production process (e.g. Van Eijk and Steen, 2016; Bovaird, et al. 2015; Moynihan and Thomas, 2013; Bovaird, 2007; Alford, 2002; Alford, 1998; Percy, 1984). But the role of the regular producer and the interaction between them are scarcely studied.

The study of Moynihan and Thomas (2013) is a good example that little is known about the role of the regular producer in a co-production process. In their study, Moynihan and Thomas (2013), describe how public professionals have to deal with citizens who are

BOX 1: The Four co-production phases:

 Co-commissioning of services, which embraces :

o Co-planning of policy – e.g. deliberative participation, Planning for Real, Open Space,

o Co-prioritization of services – e.g. personal budgets, ‘community chests’, participatory budgeting, stakeholder representation in commissioning decisions, o Co-financing services – e.g. fundraising, charges and agreement to tax increases.

 Co-design of services – e.g. user forums, service design labs, customer journey mapping.

 Co-delivery of services, which embraces :

o Co-managing services – e.g. leisure center trusts, community management of public assets, school governors,

o Co-performing of services – e.g. peer support groups (such as expert patients), Nurse Family Partnerships, meals-on-wheels, Neighborhood Watch.

 Co-assessment (including co-monitoring and co-evaluation) of services – e.g. tenant inspectors, user on-line ratings, participatory village appraisals.

(14)

13 participating in the public domain. So, basically they are describing the role of the regular producer. Their study, however, is constructed in such a way that it is more about the role of the citizens than about the role of the regular producer.

Other researchers have shown that it is useful to study expectations in a collaboration process. It is well known that collaboration flourishes better when people know and trust each other. This relates more to knowing what the others experience is and what one can expect from someone, than to actually liking the person one is interacting with (Van Delden, 2009). The same argument is made by Suter et al. (2009). Their study into collaboration in health care institutions showed that understanding and appreciating roles and responsibilities, and communicating effectively are two perceived core competencies for collaborative practice at the frontline (Suter et al., 2009:41). So, previous research in different research areas shows that expectations are perceived to be related with collaboration. Another argument to study role expectations is that empirical evidence “at the front lines of everyday practice” could be studied more extensively (Fenwick, 2012). In this light it is useful to study the roles that regular and co-producers play in a co-production process and how they see each other’s roles during that process.

Because co-production literature is mainly on one individual actor and not on how they interact, literature from other disciplines is required to understand what possible answers to the main question may be. The work of Van Delden (2009), titled: ‘Cooperation in the public service’1

gives insights in what is important when conducting research into

collaboration in the public domain. Van Delden (2009) also provides insights in what role the involved actors may play in this collaboration process. His theory however, is not specifically about co-production, but is more about collaboration between different service providers in the public domain. Therefore the theory of van Delden has to be adjusted to collaboration in a co-production process to be applicable for this study. Insights of literature on the individual roles of the co-producer and the regular producer can be helpful to do that. This integration of the different theoretical insights on roles of the two involved actors will be done in section 2.2.

To answer the main question, aimed at understanding how mutual expectations influence collaboration, it is important to understand how both actors see each other’s roles. Literature about role expectations gives some insights in how to analyze mutual role

expectations and what possible effects of role expectations are.

(15)

14 The concept of role has been conceptualized in many ways. According to McKenna (2006) a role could be defined as a set of expected patterns of behavior attributable to a person occupying a particular position in a given social system. For this research it is useful to split two aspects of the role: the expected role, this is the set of behaviors that are expected to be performed by someone in a given position, and the enacted role, this is the actual set of behaviors carried out (McKenna, 2006). It is common that there are always some

inconsistencies between expected and enacted role. In such case, adjustments are necessary to make collaboration possible and to preserve the social system where roles are performed. (Fiordelli et.al. 2014).

This is made clear by an example Fiordelli (2013) gives in her study. Fiordelli (2013:311) looks into collaboration between Medical Residents (MRs) and the immediate colleagues they interact with: Senior Doctors (SDs) and Head Nurses (HNs). One of her findings is that the expected and the enacted role of MR are perceived differently by SDs, HNs and MRs themselves. Each group was inclined to see specific aspects of MRs’ role differently and because of that developed different expectations and interpretations of MRs’ behaviors (Fiordelli, 2013:316). To find a way to collaborate with each other and face the inconsistencies within MR’s role, the three groups developed some adjustments.

To conclude this part of the theory section it is good to make a brief recap. Co-production is aimed to improve the service delivery and involves two parties: co-producers and regular producers. Because co-production literature manly focusses on one of the two parties individually, it is necessary to integrate these insights into other non-co-production theories that focus more on the collaboration and the roles of both actors; this will be done in the next section of this chapter.

2.2 Roles

To be able to do research into role expectations, I will intertwine insights that the theory of van Delden gives about roles of civil servants and citizens with some theoretical insights about roles from previous co-production research. This is necessary because co-production scholars, as explained earlier in section 2.1, mainly focused on the role of citizens and the theory of van Delden is more about professionals collaborating in the public domain. By linking the two, a new and useful framework is crated to be able to conduct the research of this study and enrich co-production literature on this topic. First some insights and

characteristics about the role of the co-producers are given, second the same is done for the role of the regular producers.

(16)

15

2.2.1 Role of the co-producer

Before the role of the co-producers in a collaboration process with regular producers will be addressed first, the question who these co-producers are is discussed. Co-producers are referred to in co-production literature as service users,clients, citizens or patients. In the working definition that is used for this study co-producers are citizens who participate in service delivery on a voluntary basis. These citizens may also be service users or clients of the public organization they collaborate with, but the main characteristic, that makes them

different from other actors, is that they are citizens.

Another element that deserves to be discussed before the role of the co-producers are pointed out, is: Why should co-producers be involved? Co-production scholars give some answers to this question. Cahn (2001) for instance says that co-producers may provide vital ingredients for regular producers to improve the effectiveness of service delivery. He points out that citizens can be the basic building blocksof the missing neighborhood-level support systems, such as families and communities, on which economic activity as well as social development are built up on. The vital ingredients Cahn (2001) writes about are information about the everyday practice and participation in service delivery. Through these two elements regular producers can improve effectiveness of their policies, because they can adjust the service delivery better to the demands and wishes of the users of the service (Moynihan and Thomas 2013), and involving citizens in the service delivery process may cause a reduction of implementation costs (Verschuere, 2012). Through these two elements the regular producers can also increase the level of democracy and involvement. By involving citizens in decision making the policies legitimacy can be improved (Van Eijk & Steen, 2015).

The theory of Van Delden (2009) is very useful for its insights about collaboration in the public domain; his theory, however, is more about the collaboration between different service providers and the role description he presents are therefore more useful for the description of the role of the regular producer. Van Delden does not totally overlook the role of citizens when different service providers collaborate. Van Delden narrows the role of citizens down to coming up with ideas in a process that has a lot in common with a citizen participation process (Van Delden, 2009:140). Van Delden states: “(…) their central role in the co-operation can be described as informing the partners” 2 (Van Delden, 2009:139). This

2_{Original Dutch quote: “Daarom is hun kernrol bij (…) de samenwerking te omschrijven als het informeren van}

(17)

16 view on the role of citizens in a collaboration process is narrower than co-production scholars describe the role of citizens or co-producers.

According to Moynihan and Thomas (2013) the role of residents can be described in three different ways. First they can be seen as citizens that vote and pay taxes. Second they can be seen as customers who make choices between different services they use and pay for as well. A third way to describe the role of inhabitants is as partners. Partners work closely together with civil servants to improve service delivery. Basically these three roles can be described in two ways: the role of informer and second the role of performer. The role of informer involves tasks like: make wishes and goals known. The role of performer involves tasks like: showing up at neighborhood meetings, visualize problems, coming up with new ideas, stimulate professionals to do their job (Moynihan and Thomas, 2013; Delden, 2009). So, when integrating Van Delden’s insights with the co-production context, the following interests and roles can be identified as presented in Table 1

Table 1: Interests and roles co-producers

Interests and Goals  addressing specific problems

 position as partner

Role Behaviors  participating in decision-making

 participating in service delivery

Core Roles Informer: making wishes and goals known

Performer: Participate in service delivery, express

ideas and complaints, answer questions, participate in neighborhood discussions

Effectiveness Positive:

visualize issues

 visualize problems

 input new ideas

 socializing government through coordinating effect of collaboration

Negative:

 contradictions between citizens

 passive, resigned attitude

Inspired by: Van Delden (2009:140)

2.2.2 Role of the regular producer

This section about the role of the regular producer will address three questions. First the question about who regular producers are will be debated, second the question about if co-production does change the role of regular producers will be discussed and, finally, the question about what the role of the regular producer is will be answered.

(18)

17 In literature regular producers are referred to in different terms such as public service professionals, professionals in general and civil servants (Tuurnas, 2015; Brandsen and Honingh, 2013; Moynihan and Thomas 2013). As stated in the working definition of co-production, I will also use this delineation of the concept of regular producers for the purpose of this research.

The introduction and implementation of co-production in the public domain directly influenced the working environment of the regular producer. Co-production compelled the regular producer to divide power, distribute tasks, and share responsibilities with the co-producers. This has implications for the professionals’ (level of) expertise, legitimacy, and autonomy (Van Eijk & Steen, 2015; Brandsen and Honingh, 2013; Moynihan & Thomas 2013). The emerge of co-producers in the service design and delivery phase also effect the role of the regular producer. Because of socialization regular producers have to coop with new tasks, requiring new skills and have to master other kinds of knowledge (Van Eijk & Steen, 2015; Osborne and Strokosch, 2013). So, the role of the regular producer has shifted partly from production of the public service to support of co-producers.

When Van Delden (2009:141) discusses the roles of the actors involved in a

collaboration process in the public domain public domain he describes three different tasks in relation to problem solving: Goal setting, structuring and practical collaboration. According to Van Delden (2009:141) five different actors are involved in this process: the government, directors, managers, professionals and citizens. The government and the directors are involved with goal setting, managers with structuring and professionals and citizens are involved with the practical collaboration. Co-production is a specific kind of collaboration in the public domain, therefor this theory of Van Delden have to be adjusted to co-production. Co-production is "the mix of activities that both public service agents and citizens contribute to the provision of public services.” (Brandsen, Pestoff, & Verschuere, 2012:1). These

activities are most in line with what Van Delden defines as: ‘practical collaboration’. Also the actors that are involved in this collaboration are the same as those who are involved in co-production activities, namely: civil servants and citizens. So, the role description of the professional that Van Delden gives is very interesting.

When looking closely to the role of the manager or the task of structuring in the theory of Van Delden, it is obvious that these also belong in co-production. The role of the manager is to create structures in which partners can collaborate. His task is also about connecting the partners (Van Delden, 2009:133-136). Co-production is not something that emerges out of noting, coordination and fine-tuning are necessary to be able to collaborate (Tuurnas, 2015).

(19)

18 To sum up, two core roles can be adopted from this. First regular producers have to be able to connect with the co-producers. To be able to assist them and work together in a

project, regular producers have to tune their work processes, creating new ideas that are in line with those of the co-producers, and have to intensify interaction with co-producers and other regular producers. By identifying bottlenecks, initiating developments and encouraging promoters, regular producers improve the effectiveness of the service delivery process (van Delden, 2009). Secondly, a core role of the regular producer is ‘to act’ in the collaboration process. This includes performing work-related tasks in service of the collaboration process (van Delden, 2009). These tasks can include support with specific knowledge, financial means, legal assistance or otherwise. This role interpretation of the role of the regular producer shows that it aimed at supporting the co-producers rather than producing a service (Brandsen & Honingh, 2013). So, when integrating Van Delden’s insights with the

co-production context, the following interests and roles can be identified as presented in Table 2.

Table 2: Interests and Roles Regular Producers

Interests and Goals  success collaboration

 relation to employer

 relations with colleagues

 career opportunities

 create something new for society

Role Behaviors  coordinator

 work hard, coop with stress

 adept work processes

 perform work-related task

Core Roles Connecting: tune work processes, creating new ideas,

intensify interaction partners

Acting in collaboration: perform work-related tasks in

order to achieve goal of the project

Effectiveness Positive:

 identify bottlenecks, initiating developments

 encouraging promoters

 Healing stranded relations

 create and improve processes

 adjust approach or work-related tasks Negative:

 over proactive, stimulation dependent behavior

(20)

19

 continue to rely on own professionalism, implicitly counteract, bring no solutions

 indecisiveness

Inspired by: Van Delden (2009:136, 138)

2.3 Expectations and collaboration

Now the concept of production has been discussed and the roles of the regular and the co-producers were debated it is time to focus on expectations that the involved actors have and what is important to focus on when studying collaboration in the public domain. As said before co-production is about interaction between the involved actors, it is about the collaboration process aimed to improve service delivery (Brandsen et.al, 2012). As stated earlier previous research in different research areas shows that expectations are perceived to be related with collaboration (Van Delden, 2009; Suter et al., 2009:41). As explained earlier it is common that there are always some inconsistencies between expected and enacted role. In such case, adjustments are necessary to make collaboration possible and to preserve the social system where roles are performed (Fiordelli et.al. 2014).

These inconsistencies can occur in different aspects of collaboration. Van Delden (2009:97) mentions several aspects of collaboration that play an essential role in a

collaboration process in the public domain. Van Delden’s theory about collaboration focusses more on the collaboration in the public domain in general. Co-production is a very specific form of collaboration in the public domain and therefor the theory of Van Delden in not perfectly applicable for this study. Therefore some adjustments were made to make it fit in a co-production process. Van Delden (2009:97) mentions several aspects that are important for collaboration: Having a solution-oriented work style, understanding each other’s language, establishing a joint work process, demarcation of the project and project tasks are mentioned to be important to a collaboration process. Van Delden (2009:97) also mentions that team building is important when collaboration in the public domain takes place. The presence of a promoter, making disagreements and frictions negotiable, and the emergence of a personal connection between the members are important when it comes to team building (Van Delden, 2009:98). Each of these aspects will be discussed into more detail now.

To make collaboration work it is heavy depended on the way people can find each other in the execution of the collaboration process. Are the involved actors able to bring the collaboration into practice? A solution-oriented work style contributes to the effectiveness of the collaboration process. If involved actors delay or block the process, effectiveness will be

(21)

20 deteriorated. All involved actors together determine the quality and effectiveness of the

collaboration process (Van Delden, 2009:97).

The actors that are involved in the process have to speak each other’s language to be able to collaborate successfully. Regular producers in service of the government are

specialized in their own area and have their own terminology, professional behavior,

distinctive skills and ethics. This gives meaning to their professions, but also creates distance towards others. It is essential for successful collaboration that regular producers and co-producers create a shared terminology (Van Delden, 2009:97). This process of learning each other’s language and creating a shared terminology is referred to in literature as socialization (see for example: Rollag, 2012; Kelly et al. 1990).

This socialization also includes the change of creating the right mechanisms for releasing the potential of user knowledge and to ascertain that the co-production parties have necessary skills to make use of these mechanisms (Osborne and Strokosch, 2013). This right mechanism is not created by forehand but evolve over time. Confrontations are common when new partners start to collaborate. The group must learn to convert their disagreements and frictions into useful solutions. Openness and acceptance are necessary to achieve this. Conflict forces members to develop and test roles, but also functions as a catalyst for cooperation (Van Delden, 2009:97). Socialization will fade to the background when a common mechanism to collaborate is established (Van Delden, 2009:97). This means that involving actors know what is expected from them, having the necessary abilities to function in the process and having the necessary knowledge to interact with other involved actors (Kelly et al., 1990:319). In this way, socialization leads to more accurate role expectations (Kelly et al. 1990:321).

During the process the goals of the project have to be made real. This means that the goals have to be translated into practical, concrete objectives. This demarcation of the project and project tasks is necessary to collaborate (Van Delden, 2009:98).

The last aspect mentioned by Van Delden (2009:98) that is essential for collaboration to succeed is teambuilding. Involved actors are not custom to work with each other, therefore the work of a promoter, a member of the group with a great personal commitment to the project who delivers great effort to get the project started and keep it running, is perceived to be essential for a collaboration process to succeed. Second, openness of involved actors to each other is perceived to be important. Involved actors have to make disagreements and frictions clear to each other and have to learn how to deal with them. Especially within new projects, confrontations are very common. Conflicts, when handled well, force involved

(22)

21 actors to develop and test the work process and also functions as a booster for collaboration. Finally, (professional) commitment between the involved actors is necessary for teambuilding and thus for collaboration. The increase of a professional or personal bond between the participating individuals, not just in the team but also in the one-to-one relationships will make the collaboration more effective (Van Delden, 2009:98).

Having explained what elements of collaboration are important, it is possible to identify possible sources of misfit in expectations and role performance.

Concluding, the following mechanism based on the above discussed theories from co-production literature and other research areas is constructed, this mechanism is visualized in figure 1. Co-production is the collaboration between regular and co-producers that aims to improve the quality of service delivery (Brandsen et.al, 2012). So, the starting condition of such a process is dissatisfaction with the service delivery or the expectation that the service delivery can be improved. The initiative to improve the service can come from citizens or from the government. When such an initiative evolves, participants have expectations about how they and others should behave (Fiordelli et al., 2014). It is common that there are always some inconsistencies between expected and enacted roles. When this happens socialization is necessary to make collaboration possible, and to preserve the social system where roles are performed. (Fiordelli et.al. 2014). Collaboration is successful when the goal of the initiative is realized. Figure 1. initative emerges misfit between ecpected and enected role

(23)

22

3. Research Design

In this chapter the research design is explained. It discusses the research approach, the case selection and selection of the respondents for the interviews which were used to collect evidence, how this evidence is analyzed and what the limitations of this study are.

3.1 Research approach

The evidence for this research will be collected through qualitative research; more precisely through semi-structured interviews. This qualitative approach is chosen because the goal of the research is to increase the knowledge about the interaction between regular and co-producers, and to find out how mutual role expectations play a role in this collaboration process. As mentioned earlier, literature on the interaction between those actors is limited. Therefore, qualitative research, contrary to quantitative research, fits better to the goal of the project and to the state of the art.

To find an answer to the main question the causal-process tracing (CPT) approach will be used as a starting point (Blatter and Haverland, 2012). The CPT approach is designed to reveal how “the temporal interplay among conditions or mechanisms leads to specific outcomes” (Blatter and Haverland, 2012:27). This is in line with this research because this study aims to uncover how mutual expectations influence the collaboration process between regular and co-producers. Also the storytelling style of presenting the date and the fact one unique case is studied into depth are elements that are derived from the CPT approach. The historical aspect of the CPT approach is an element that is not conducted in this study. Instead of doing observations over time, a time frame of the collaboration process is reconstruct from the collected data. A problem with this is that participant bias weakens the evidence, but because of time limitations reproduction of the time frame is a way to find answers to the main question and accomplish the goal of the research.

3.2 Case Selection

The case under investigation in this study is a project in the Hoge Mors neighborhood in Leiden, the Netherlands. With this project citizens wanted to improve a neglected park in their neighborhood. Because this study uses the CPT approach as a starting point, also the case selection has its roots in the guide lines the CPT approach offers.

According to this approach this one case is selected because of its collaboration between co-producers and regular producers. For the CPT approach the number of cases is

(24)

23 less important, because the researcher has to deal with a given amount of resource, increasing the number of cases reduces the possibilities and probability of finding sufficient empirical evidence to provide a convincing explanation for each individual case (Blatter and Haverland, 2012: 100).

Next to this, the most important criteria for case selection are accessibility and availability. It is important to satisfy these criteria to be able to reach deeper insights into subjective perceptions and motivations of important actors, as many sources as possible should be open for analysis (Blatter and Haverland, 2012: 102). Because the case is very recent participants are still able to remember and answer questions about their role expectations. Also because this case has a local scope, participants are contacted easier

compared to high level civil servants, this improves the accessibility of this case. Also the fact that public greenery policy has a rather low political sensitivity helps to improve accessibility as well. Participants are more likely to answer questions about subjects that have a low political sensitivity, than on subjects that are very controversial.

This case is also selected because it has theoretical relevance for the broader co-production literature. Often used cases of co-co-production that are studied relate to safety or health care (Van Eijk & Steen, 2016; Van Eijk & Steen, 2015; Bovaird, Ryzin, Loeffler & Parr, 2015; Bovaird, 2007). The Morspark project relates to a different area, namely the green sector in the public sphere, as such extending our insights to another policy area. The case also ensures that there is enough interaction to study. The project team includes some officials of the municipality and some external professionals, next to that about fifteen citizens are actively involved as well. The way in which those actors collaborated with each other is explained in section 4.2.

To collect data for this study, to groups of actors will be interviewed. The first group, earlier described as regular producers, consists of the professionals or civil servants. They work directly or indirectly for the municipality. The second group that I will interview consists of co-producers or citizens. These co-producers are involved in the co-production process on a voluntary bases. In total five civil servants, and eleven residents were asked to participate. In the end, a total of eight semi-structured interviews were conducted with three civil servants and five residents.

The civil servants that participated were involved in the co-production process as project manager, as civil servant maintenance public space and as architect. The two civil servants who did not participated in this study were involved as project assistant and architect. All the co-producers that participated in this study were involved in the project as residents. In

(25)

24 the project also several local public professionals participated in the Morspark project. In this study the focus is on the role of the regular and the co-producer, therefor the public

professionals were not approached. The architect that participated in this study (RP3) was also one of the imitators of the Morspark project. When the municipality got involved in the Morspark project this architect functioned like one of the architects of the municipality, therefore he will be referred to as RP3.

Involvement Description Code

Regular producer Project manager RP1

Regular producer Maintenance public space RP2

Regular producer Architect RP3

Co-producer Resident CP1 Co-producer Resident CP2 Co-producer Resident CP3 Co-producer Resident CP4 Co-producer Resident CP5 3.3 Data Collection

This section is about how the data that is necessary to answer the main question is collected. First, the approach to potential interviewees is explained, second the setup of the interviews is discussed.

All potential participants were approached by email. That email contained information about the research project, what the researcher expect from the participant and what the participant could expect from the researcher. This made sure that the requirement of informed consent is met. Appointments were made by mail or telephone. All interviews were conducted between April 18th and April 22nd 2016.

The interviews were built up in such a way that the participant felt comfortable. The interview starts with some general questions about how they got involved in the project and what their role was in the project. After the more general questions, the questions about the enacted and expected role were asked.

For the purpose of easier, more structured and consistent analysis of the data, the interview will be conducted in a semi structured way. Furthermore, this approach also allowed the researcher to ask additional questions if necessary. In this way it was possible to compare the answers and find more in-depth evidence.

(26)

25 To collect the required data, the interviews with both the co-producers and regular producers, was divided in four parts. The first part was aimed to gain trust and make the participant feel comfortable. This part contained more general questions about the participant and the Morspark project. The second part was aimed to get an understanding about the roles that the involved actors had in the project. These questions focused on how the other party behaved during the co-production process. The third part focused on what the expectations of the participants were. The best way to collect evidence on role expectations is at the start of the process. However, taking into account the time frame of the research project, it was impossible to find a case that could be analyzed from start to finish. Furthermore, it should be noticed that this way of collecting data also puts pressure on the respondents involved,

because they have to be interviewed two times. Therefore, respondents were only interviewed once and their role expectations were reconstructed. Consequently, the answers of the

interviewee about role expectations can be biased. Interviewees might not remember precisely wat they expected of the other party in the first place. The researcher was aware of this during the interviews and asked follow-up questions to ensure the evidence about role expectations would be as pure as possible. The fourth part was about the collaboration and contained questions about the concepts of collaboration that were discussed in theory section 2.3. Finally, two questions were added to make sure all relevant items were discussed and to give the participant the opportunity to give further comments. The interview questions are added in Appendix A.

3.4 Data analysis

The goal of the study is to increase the knowledge about the interaction between regular and co-producers, and to find out how mutual role expectations play a role in this collaboration process. As mentioned before, the method that will be used to accomplish these goals has its roots in the CPT approach. The most important basic characteristic of the CPT approach is that the search for answers is based on ‘configurational thinking’ especially on “the

assumption that explanations should begin with the assumption that a plurality of causal factors work together to create an outcome” (Blatter and Haverland, 2012:81). Another basic characteristic is that the CPT approach “as a technique of drawing causal inference takes advantage of the fact that causality plays out in time and space” (Blatter and Haverland, 2012:81). As a result of these characteristics observations that determine the temporal order by which a process develops (‘comprehensive storylines’), the empirical observations that clarifies the pathway how a cause leads to an effect (‘smoking guns’), and empirical

(27)

26 information that specifies the underlying action-formation mechanisms that link causes and effects (‘confessions’) are the main pieces of evidence to answerser to the main question (Blatter and Haverland, 2012:81).

The following section describes what evidence is necessary to answer the main question of this study. First, it is necessary to get a general overview of the co-production process, its initial conditions and its outcomes. This overview will serve as a foundation on which the evidence about how the causal mechanism works can be built. In the CPT approach observations that determine the temporal order by which a process develops are called

‘comprehensive storylines’ (Blatter and Haverland, 2012:111). This contains events like: who got involved when, what events took place when and where. Second, this overview also contains information that reveals how expected and enacted role developed through the collaboration process. This contains information about what the expectations of the regular and co-producer were and how they behaved (what they did during the process). An overview of the collaboration process will be created by comparing the story lines that interviewees give. When an overview is created, ‘turning points’ can be identified from the data. Because interviewees will remember important events and mention these in the interviews, the number of occurrence of an event says something about the importance of an event.

The following section is about how the evidence was derived from the transcripts. The data analysis was split into two phases. In the first phase the transcripts of the conducted interviews were prepared. Unfortunately the information provided through interviews does not automatically translate itself into evidence. Information becomes evidence by a process of linking the information to the theory through coding. Trough coding it is possible to link that what the interviewees have said to the theory. Because this study has a quantitative approach detailed labels and concepts were constructed during the analysis of the evidence. This was made easier by the use of semi-structured interviews.

The evidence about the different concepts that are mentioned in the theory section were identified in the following way: ‘Comprehensive storyline’ observations were derived from the interviews by searching for: Names, dates, organizations, events, goals, locations and other things that occurred when I processed the interviews; The evidence about the expected role was revealed by tracing the transcripts for words that link to opinions, motivations, expectations, need, or comparable verbs that are linked to interviewee expectations or about expectations of other actors; To find evidence about the enacted role the transcripts were examined for verbs like: do, go, talk, organize, etc. These verbs had to be linked to actions of the interviewee or other actors involved in the process; To find evidence about misfits the

(28)

27 transcripts were examined for: disagreements and discussions about role distribution; To find evidence about socialization the transcripts were examined for: agreements about role

distribution. An overview of the terms that were used to structure the interviews is stated in table 3. In the second part of the analysis the individual pieces of evidence were restructured and combined to formulate an answer to the main question. (Rubin and Rubin,

2011:201). The result of this process is presented in chapter five of this thesis.

Table 3.

Reconstruction collaboration process Names, dates, organizations, events, goals, locations, and labels that occur when transcripts are processed.

Expected role Words or phrases that link to: opinions, motivations, expectations, and needs, or comparable verbs that are linked to the interviewee or about other actors.

Enacted role Verbs like: do, go, talk, organize, etc. that are linked to actions of the interviewee or other actors involved in the process.

Misfit Disagreements and discussions about role distribution.

Socialization Agreements about role distribution

3.5 Validity and reliability

Every research method has its strengths and weaknesses; this has consequences for the validity and the reliability of this research. These consequences are discussed in this section.

In-depth case studies have a high internal validity, because they describe the case very closely. This in-depth approach makes it possible to encounter alternative factors that were not specified by forehand, but that played a role in the collaboration process and that increases the internal validity of this study.

Because the interviews are conducted while the project had already started,

participants might answer questions about role expectation with a bias because they do not exactly remember what their expectations were. This participant bias reduces the internal

(29)

28 validity of the evidence. The conducted method of structured interviews helped to limit these biases by asking follow-up questions.

Another bias that threatens the validity is that interviewees might give socially desired answers, because they like the project they are working on or the persons they are working with. To limit this bias it was made sure that the interviewee felt as comfortable as possible. To achieve this, the interview was conducted on a location of the interviewee’s choice. At the start of each interview it was made very clear to the interviewee that there were no right and wrong answers. Also follow-up questions were used limit the possibility of a biased answer to the question. The fact that public greenery policy has a low political sensibility also reduced the chance of participants giving social desired answers.

The external validity of in-depth single case studies is weaker than results of studies that rely on a large number of cases, because results of in-depth single case studies are only generalizable to the population of cases. This means that the findings of this study can only be applied to cases that are similar to the Morspark project (Blatter and Haverland, 2012:40).

The reliability of this study is strengthened by the use of semi-structured interviews. Also the fact that the interviews are recorded and are literarily transcribed makes it possible for other researchers to redo the analysis on order to test the results of this study.

(30)

29

4. Case Description

This chapter contains a description about the citizen initiative Morslint, and the Morspark project which is a sub-project of Morslint. A historic overview is stated at the end of this chapter

4.1 Morslint

The Morspark project is a plan that is part of a bigger citizen initiative Morslint that aims to develop a pedestrian route to improve and refurbish various places in the public space in the neighborhood of the Hoge Mors. In that way the initiators hope to create a neighborhood that is pleasant for pedestrians and to increase cohesion in the district. Next to the municipality and the residence of the Hoge Mors neighborhood several third sector organizations support the Morslint initiative, these include: Libertas Leiden, GGZ Rivierduinen, Radius foundation, Care group ’s Heeren Loo West-Nederland, Housing Corporation Portaal, Popma & ter Steege, Wijken voor Kunst Leiden (B&Wbesluit 150611:3) .

The Morslint initiative has one main goal and three sub goals. The main goal is to create a pleasant living and working environments in the High Mors on the basis of

participation and self-reliance of residents and users of the area. The three sub goals have a physic, social or economic character. Physic goals include: improvement of accessibility and coherence of the district by improving and connecting existing pedestrian routes and

improving the public space by utilizing and developing existing qualities. Social goals are: enhancing involvement of residents and users of the neighborhood to their daily living and working environments and bringing together initiatives, needs and capacities of residents, users and local social organizations. Economic goals are: increasing the effectiveness of investments (including existing services) in the district by consistency of interventions and stakeholder involvement and increase the future value of the neighborhood (B&Wbesluit 150611:5).

After a period of negotiation talks, the municipality adopted the subproject Morspark in June 2015. The municipality has established a so-called project assignment to be able to get the municipal organization started with the subproject Morspark (B&Wbesluit 150611:5).

Residents and professionals in the Hoge Mors district signalized in 2012 that the construction of new housing and facilities gives a boost to the neighborhood. They agreed that more was possible and necessary to realize this for the whole district. From that moment some of these residents and local professionals work together to further shape the improvements in

(31)

30 the district. This led to the creation of the citizen initiative of Morslint in December 2013. The initiative included several subprojects: Morspark, Diamantwand, Lelylaan, Crescendeohof, weerklankenplein and ‘Verblijven langs het Lint en Bewegen langs het Lint’ (B&Wbesluit 150611:5).

4.2 Morspark

Morspark is a subproject of Morslint in which residents collaborate with the municipality and local organizations to improve a green space in the Hoge Mors district. The project aims to realize the following results: the creation of a neighborhood wide supported design of the park, the design of a management plan which identifies agreements on joint usage and

management of the park, a estimation of the financial costs of the plan, financial coverage for the estimated investment, communication and participation through a communication plan, approval by the municipal council and finally the redesign and construction of the park (Project opdracht Morspark:5).

Figure 2: The project area. It contains roughly the green space between: Opaalstraat (above), Saffierstraat (top right), Topaaslaan (bottom), Maansteenpad (left) en Turkooislaan (bottom left).

From the moment the Morspark project became an official project of the municipality in June 2015, the municipality started to look for a project manager who was going to be lead the project on behave of the municipality. In November the manager started with acquaintance with the residence and partners who were already involved in the project. The manager also wanted to get a general idea what progress was made previously. In January 2016 the project really took off.

(32)

31 Three different consultation structures were created to structure the process. The first consultation level focusses on design. On this level a private architect, who was one of the initiators of Morslint and an architect of the municipality are involved. The private architect is responsible for the sketch and draft design, because of the close contacts with residents and the involvement in Morslint. The responsibility for the definitive design lies with the municipal architect who has knowledge of municipal spatial planning standards. These architects meet once a moth. In January the sketch design was finished, in March the draft design was completed and according the planning the definitive design is expected in May. These designs are presented to the residents at a so-called residents evening. This second consultation level is aimed at testing if the designs are really in line with what the residents want. All residents of the Hoge Mors neighborhood are welcome to these evenings. A

residents evening is planned after a design is finished by the architects. In total three evenings are planned. The last and most interesting consultation level for this study is the project group in which a dozen active residents and local partners participate. In this project group different topics like playground, greenery and exercise are being discussed. At these meetings

residence and partners can express their views and ideas on these topics. This project group also meets once a moth.

4.3 Historical overview

Below an historic overview is shown. The overview consists of important events that took place in the years before the interviews were conducted.

 In December 2013 the development plan Morslint prepared by an initiative group.

 In the spring of 2014 the project group worked on building support for Morslint by creating a temporary Morslint and a presentation by PS Theater.

 In July 2014 a sketch design ‘Park De Hoge Mors’ was offered to alderman Van Gelderen. The municipality is requested to participate in the project group.

 On October 9, 2014 the design of Morslint is presented to alderman Laudy.

 On October 27, 2014 the design by a civil servant is discussed at the periodic consultations of greenery management (a municipal organization responsible for the maintenance of the greenery in the public space) in the presence of the aldermen Damen and Laudy. They indicated that they are basically in favor of municipal support of the initiative. It was agreed that the civil service can prepare a proposal for the municipal council.

(33)

32

 During the period of November 2014 to November 2015 further communication and coordination between the municipality and the initiators of Morslint to create a project plan for the Morspark project takes place.

 On July 22, 2015 a definitive project assignment is disclosed by the municipality. The Mayor authorizes alderman Laudy to sign the agreement on behalf of him.

 On November 10, 2015 the municipal council approves the financial support of the Morspark project.

 During the period of January 2015 to May 2015 the project group had monthly meetings to discuss the different topics.

 In January 2015 the sketch design was presented at a residence evening.

(34)

33

5. Analysis

This chapter is about how the case relates to the theory of this thesis. Two phases can be identified when looking at the developments of the Morspark project. The first phase of the project can be seen as a citizen initiative and the second phase can be identified as co-production. In the first phase the initiative developed from a first idea to a plan were the municipality was one of the necessary partners to create and implement the plans. The second phase can be described as co-production, where civil servants and the group residents

collaborated to design the park.

5.1 The start-up phase

It is necessary to look closely to how the project developed before the municipality made it an official project in the summer of 2015, because expectations already existed and emerge in that first phase and some structures and role distributions already emerged as well. This section will discuss two major themes that influenced the structure and expectations of the collaboration of the co-production project. First, the effect of the very start of the project will be discussed. Second, the effect of the process of involving the municipality in the project is analyzed.

5.1.1 First Ideas, expectations and collaboration

The idea for the Morspark project emerged over time and was not created by one single person or actor. An architect who was involved in a different project in the area came up with the idea to improve the public space of the entire neighborhood and asked residents to come up with ideas. The residents came up with the idea of a pedestrian route. Later this idea was transformed into a plan that linked all the green areas in the neighborhood. The theory section of this thesis points out that co-production aims to improve the quality of service delivery (Brandsen et.al, 2012). Therefore, a dissatisfaction with the service delivery or the expectation that the service delivery can be improved can function as an initiator for such a process. In this case the opinion of the architect was one of the starting points of the initiative. He thought it were possible to improve the use of public space. Also some residents of the neighborhood came up with ideas and joined his initiative.

“The Idea of the pedestrian route came from a group of residents who live in the

neighborhood, the idea of the making improvements to the public space came from us. Later we thought of the idea to link the green areas in the neighborhood, which happened back in 2013.” (RP3)

(35)

34 Van Delden (2009: 97) points out that generally spoken actors involved are not used to work together, which is also the case for the Morspark project. Therefore the work of a

promoter, a member of the group with a great personal commitment to the project who

delivers great effort to get the project started and to keep it running, is considered essential for a collaboration process to work properly. Within the group of involved actors such a promoter can be identified. A group member enacted behaviors that correspond to the role of promotor: “After we really tried to get people on board and didn’t get any response, we designed a plan by our own and argued: ‘Apparently we have to come up with something specific, otherwise people will not understand it. So my role was to boost, visualize and propose opportunities.” (RP3)

The public professional points out that the citizens were not that much involved at the first developments of the plan. They shared some ideas but were not participating to put things on paper. They needed something specific to be triggered to participate. Several residents pointed out that first expectations were vague and unclear, but they became clearer after the process went on. A resident responded to a question about what the expectations were of the project: “Well, at the start of the project it was very vague. (…) But during the process I liked it more and more (…).” (CP1) Another resident said that it was not clear what could be expected, the resident denied to know specifically what was going to happen from the beginning: “No, but that was also due the fact that things became clearer when people came up with ideas.”(CP4)

The theory points out that an important element of collaboration is the creation of a certain mechanism to collaborate. This mechanism is not created beforehand but evolves over time (Van Delden, 2009:97). The group did not have a clear idea about how to structure the process. “I expected that it has to be possible to improve the quality of the public space. It was interesting to see if this would be achievable by some kind of collaboration.” (RP4) It was difficult to distribute roles and tasks among interested residents. Role distribution among group members was executed on one’s own initiative or on request by one of the group members. “She stood up herself and started to do that role all by her own. You have to let these things grow in a certain way, or something like that. It is a search for what kind of structure is necessary, where you say: ‘This is something for you!’ and where you leave room for people to wait for someone to take the initiative by their own.” (RP3)

The role distribution that emerged was later adopted in the co-production project. The role distribution resulted in a mechanism of different groups who focused on different themes within the overall plan. The way the project manager would later organized the project is

Morspark: Co-production as Expected?