Journey before destination: A study on how past experiences can shape collaboration progress in four regional energy transition networks in the Netherlands

(1)

Leiden University

Master thesis

Faculty of Governance and Global Affairs

Public Administration: Public Management and Leadership Supervisor: Maarja Beerkens

Name: Steven van der Plas Student Number: S1508024 Tel: 0630114460

Email: Stevenvanderplas12@gmail.com

Journey before destination

A study on how past experiences can shape collaboration

progress in four regional energy transition networks in

(2)

(3)

Abstract

This thesis explores the way network maturity affects the level of collaboration in organisational networks. To this end, a comparative case study was conducted into four regional networks on energy transition in the Netherlands. Two of these networks have a distinct networking history on energy transition while the other two have no compatible past networking experiences. The findings show that the level of collaboration in the two mature networks was directly influenced by past networking choices and engagements. Not only did these networks rely on existing governance structures to support their collaboration efforts, they also exhibited high levels of regional project initiation, external communication and financial resource pooling. The two unevolved networks demonstrated a lower level of partnership implementation aspects and used more straightforward governance structures to facilitate their efforts. Surprisingly, this thesis did not find a link between participation and network maturity, as all observed networks actively included and relied heavily on the input of a broad range of governmental and non-governmental actors.

The results reaffirm the notion that it takes time and a certain amount of collective experience as a network to successfully implement collaboration and to reap its benefits. Being able to build upon previous policy decisions and commitments not only has practical advantages for networks, but also helps them in gradually creating the network infrastructure to support their activities. The findings also show that without sustained commitment, these earlier experiences can become incompatible with future networking efforts, which compromises their usefulness. However, not all collaboration aspects are reliant on the maturity of a network. For instance, the experiences of the observed cases show that networks are able to involve and activate large groups of societal actors in a relatively short amount of time, once issues become salient and there is a need for a joint solution.

(4)

Preface

So here I am at the end of the journey, with this massive end product about such an interesting topic. Getting to know the different networks and the dynamics behind them was extremely insightful. I came across many aspects and intricacies that unfortunately could not be discussed in this thesis. However, I am very optimistic about the potential for future research possibilities, which I partially outline in the final subchapter in this document. Writing this thesis took longer than expected and it ended up being a bigger project than I originally anticipated. I could never have finished the work in this capacity were it not for a select group of people who helped me with the thesis itself. First and most important, many thanks to my supervisor, Maarja Beerkens. Even though it took some time, you patiently read my pieces and made my work better with your corrections and feedback. Additionally, my results would never have been as clear without the help of a group of interviewees who took the time to answer my questions. I am very grateful that you decided to help me in a time where conducting interviews was deemed difficult or even impossible.

I also have to give my thanks to a group of people very dear to me who helped me stay sane throughout the year, particularly during the weeks in lockdown, when my days mostly consisted of eating, sleeping and writing my thesis. Thanks to the lovely people at the board of the student association Leiden United for your support these last few months. I had the pleasure of being its president for the last year and it has been an amazing journey together. Despite the difficulties of the last few weeks, you still managed to motivate me enough to keep working and your support ultimately helped me finish this final product of my time as a student. I am also particularly grateful to my friend Bob, who managed to read and correct my entire thesis when others could not. Your comments might just have saved me, literally.

(5)

List of abbreviations

AAG Algemene Adviesgroep

BO-RES Bestuurlijk Overleg Regionale Energie Strategie BRG Bestuurlijke Regiegroep

EU European Union

IPO Interprovinciaal Overleg LTO Land en Tuinbouw Organisatie NAO Network Administrative Organisation NMU Natuur en Milieufederatie Utrecht PHO Portefeuillehoudersoverleg

PJ PetaJoule

PS Provinciale Staten

RES Regionale Energie Strategie RRV Regionale Ruimtelijke Visie RSW Regionale Structuur Warmte TWH Terrawatthour

UVW Unie Van Waterschappen

VNG Vereniging van Nederlandse Gemeenten VV Verenigde Vergadering

(6)

1. Introduction

1.1. The power of networks and collaboration

In modern society, public agencies are often confronted with challenges that reach further than the boundaries of their own jurisdiction and problem-solving capacity. Such problems can be, but are not limited to ‘wicked’ problems, which are characterized by their complexity and far-reaching consequences. The hierarchical nature of contemporary public organisations makes it difficult to effectively coordinate on providing solutions that completely address these challenges. In response to these coordination failures, there have been numerous instances of the forming of networks or collaboration initiatives between public, semi-public and private organisations to create effective policies where individual agencies failed to address a complex problem. In such cases, legally autonomous organisations work together to achieve not only individual but common goals, which potentially makes collaborating or networking a viable solution to address recurring societal issues. However, in public contexts, these networks can become extremely complex structures themselves, requiring vast explanations and theories into how they work. Managing these networks to increase effective interagency collaboration has proved challenging, because of the varying contexts, participants and network-level goals, there is no obvious one size fits all solution to guarantee positive outcomes for all parties involved. Consequently, public interagency collaboration is hardly the panacea for every situation where a problem goes beyond the jurisdiction of a single agency.

Given their importance, networks have garnered increasing societal and academic attention since the beginning of the 21st_{century. Especially with regards to the way participants interact} between each other, how they are governed and how networks can lead to positive outcomes. Collaboration is widely regarded as the most intense way of interaction between organisations in a network (Mcnamara, 2012). Complex governance structures are created and organisations share goals, pool resources and jointly manage the network. Collaboration is therefore often the goal in situations where intense interaction between multiple public agencies is needed. However, collaboration is not easy to attain, since it requires the right circumstances and a certain amount of time to succeed (Mcnamara, 2012). Moreover, mature networks with a networking history commit to collaboration more readily than relatively newer networks that did not have time to fully develop the trust and institutional structures required for collaboration

(9)

1.2. Illustrating local energy transition networks in the Netherlands

A recent and prominent example of networking attempts to promote interagency collaboration can be found in the policy field of local energy transition in the Netherlands. Since the signing of the Paris climate agreement, climate change and the need for energy transition have been identified as two massive challenges for the 21st century by the Dutch government (Ministry of Economic Affairs, 2016). Addressing this issue has always been on the national agenda, but despite the development of national policy documents, strategies and covenants, there was only sporadic involvement from municipalities in this whole process. The role of local governments in this energy transition process is potentially massive, considering the fact that a wide array of partners rely on local governments for coordination and because the development of solar, wind and heat energy are intrinsically local and regional questions (Rijksoverheid, n.d.). What further complicates addressing the issue of energy transition is that public responsibility is interwoven in a complex web of local, provincial and national jurisdictions. The exact role of the municipality remained open to interpretation, which led to a variety of formulated public problems, solutions and targets (Rijksoverheid, 2018). The result was that energy transition only received proper attention in a distinct number of municipalities. Furthermore, regional cooperation between municipalities on the topic of energy transition occurred only sporadically before 2016 (Rijksoverheid, 2018).

This status quo of limited and differentiated local interaction in the policy field of energy transition in the Netherlands changed when the Dutch government introduced plans for a comprehensive climate agreement in 2018 (Klimaatakkoord, 2019). This agreement was the product of negotiations between the three layers of the Dutch government, companies and stakeholder organisations. The aim of this agreement is to reduce CO2 emissions by 49% in 2030 through 600 measures related to energy transition and sustainability, divided between the policy areas of urban environment, mobility, industry, agriculture and electricity (Klimaatakkoord, 2019, 2-3). The agreement has been passed by a broad coalition of political parties in both houses of the Dutch Estates General. A key element of this climate agreement is that local and regional levels of government have to be heavily involved in its implementation, specifically when it comes to the generation, storage and transportation of green energy and heat energy. To this end, municipalities had to work closer together and with non-governmental actors to coordinate the means, targets and solutions for energy transition.

(10)

To ensure that municipalities across the whole of the Netherlands would contribute to the climate agreement, a reformation of the way local energy transition policies were created was necessary. Therefore, the country was divided into 30 energy regions, which are essentially networks of municipalities, provinces, water boards and non-governmental actors working together to create a comprehensive regional energy transition strategy for the coming thirty years (Rijksoverheid, 2018). These regions have to implement the sustainable heat energy and electricity aspects of the climate agreement on the regional level. Energy regions are not necessarily a new layer of government, situated between local and provincial authorities, since they are created with the input of the municipalities and other layers of government (Rijksoverheid, n.d.). Instead, they directly connect with or are a continuation of already existing local networking attempts. Incorporating these bottom-up elements has two aimed advantages. First, by stimulating the municipalities to network, the asymmetries of the old system of energy transition will be gradually broken down (Klimaatakkoord, 2019, 222). Second, it uses the valuable ties and knowledge of the regions where networking on the topic of energy transition is already happening (Rijksoverheid, 2018).

The aim is to let every energy region create and implement its own Regional Energy Strategy (RES), which is a framework that outlines what the energy transition goals are for that region and how the different layers of government will work together with stakeholders to achieve this. A RES describes how much and through what means a region is going to develop sustainable electricity and heat energy in 2030. Considerable decision-making authority is delegated to the energy regions. In fact, they can even include other energy transition policies in their RES, such as sustainable mobility, industry or agriculture (Klimaatakkoord, 2019, 222). A RES is both an extensive policy document that outlines a region’s sustainable energy policy for the coming decades and an agreement between municipalities, water boards, provinces and regional and local partners. Theoretically, the energy transition efforts of all regions combined will make up a significant portion of the national climate agreement. In this sense, the Dutch government has effectively decentralised part of its energy transition policy to the local and regional level. Each of the regions has to submit a conceptual RES in May 2020, which is a precursor to the RES 1.0, which will be finalised in 2021. Furthermore, most regions have created a starting document in 2019, marking the official start of the RES creation process and outlining the ambitions and governance aspects to make this happen.

(11)

The creation of the energy regions is the most recent instance of networking practises being used to tackle a societal problem that requires interagency collaboration. What is striking is that this process is occurring simultaneously in 30 regions across the Netherlands. The energy regions and the networks behind them broadly share the same purpose and objective, namely to stimulate collaboration on energy transition and to create comprehensive policies. However, there are also distinct differences in the level of success of collaboration across the different regions. Hence, the question arises why these differences between the energy regions are occurring in this fashion. What is clear is that the framework of the energy regions has been mandated in a top-down fashion by the Dutch government. On the other hand, existing networks between municipalities, companies and other relevant local actors are used as a catalyst for making progress in the energy regions. The result is that some energy regions already have an extensive and relatively mature network while others have been created from the ground up with the help of the Dutch government, which means that they are relatively unevolved. What is unclear however is whether this difference in maturity matters and what its consequences are for the development of collaboration in the regions. The maturity of a network and how it has evolved directly relates to the existence or absence of previous networking experiences of the regions. To investigate the relationship between the former and the latter, the following research question is posed:

“How do past networking experiences affect the level of collaboration in regional networks in the policy field of energy transition in the Netherlands?”

The research question will be answered by comparing four energy regions on their progress so far towards establishing a comprehensive RES. Two of the chosen networks have a distinct prior history of cooperation on energy transition, meaning that they are relatively established. The two other observed networks have no such prior history and are relatively novice in cooperating on energy transition. The question arises whether this difference in maturity leads to different processes and ultimately to a discrepancy in collaboration. The analysis will be conducted through the use of a qualitative multiple case design, since this allows for the analysis of the context that the energy regions have operated in and for a detailed look into the factors that directly cause collaboration to be facilitated or halted. Furthermore, comparing the findings from multiple energy regions against each other will yield reliable and valuable insights. The

(12)

research attempts to apply existing theoretical foundations to a more recent phenomenon to test, update and create theoretical insights.

1.3. Societal relevance

Collaboration has increasingly become a necessity to solve contemporary societal problems. With the advent of multi-level governance and the perceived importance of broad societal consensus for public decisions, collaboration between public agencies has become a necessity for governments to survive. Exploring how public agencies can collaborate together or with non-governmental actors on topics such as energy transition will lead to additional insights on how an approach can be created to correctly address such problems. Local governments are becoming increasingly important in The Netherlands, especially following the considerable decentralisation efforts in the healthcare sector. Local authorities are getting more responsibilities and are being widely involved in executing national policies, with mixed results (De Koster, 2019). While local authorities are usually more effective in providing tailor made approaches, decentralisations have always gone hand in hand with severe cutbacks, restricting capacity to coordinate. Therefore, given the popularity of decentralisation measures in the Netherlands and the pressure to succeed, it is vital that the workings of local collaboration are explored to the fullest.

Energy transition is one of the biggest challenges that the modern world faces today, which means that there is significant debate on how it should be implemented and what it should cost. Besides the political nature of the issue, there is also the fact that implementing effective policies requires a comprehensive approach from both governmental and societal actors. In the Netherlands, the issues of climate change and energy transition have become more widely recognized in the last years under the current and previous Dutch cabinets (Ministry of Economic Affairs, 2016). Despite the wide recognition of climate change and its effects, there has equally been an increase in public scrutiny regarding the costs of energy transition and what measures are warranted. Given its divisive position in public discourse, it is important to fully uncover how this process can be optimized.

(13)

Furthermore, energy transition is an issue of international significance. The progress of the Netherlands in pursuing its energy transition goals has to be viewed in the context of several international treaties. For years, the Netherlands has not been able to attain its own targets when it comes to producing renewable energy resources and reducing greenhouse gas emissions. Yet, the Dutch Prime Minister Mark Rutte has continuously emphasized the importance of more ambitious climate targets from an EU-wide perspective, claiming that the effects of the Paris climate agreement will not be enough to stop the rise of global temperatures (NOS, 2018). Therefore, the success of the energy region program is also vital for Dutch international standing. On the local and regional level, energy transition policies impact a wide range of stakeholders, as energy policies are also interlinked with other policy areas, such as the use of space in the region (De Graaf, Hildebrand, van Schaick, Somsen & Westervoorde, 2018).

1.4. Scientific relevance

The academic debate on addressing cross-cutting issues has been widespread and many scholars have researched how collaboration can be reached. For example, Mu, de Jong and Koppenjan (2019) note that interagency collaboration is dependent on many factors, which are prominent in differing strands of literature. Their attempt at creating an encompassing model to assess and explain interagency collaboration is far-reaching, but they note that their results are different from those of other models in the literature. Previous networking experiences have been theorised to affect collaboration, with Human and Provan (2000) noting that network evolution affects the creation of “embedded relationships” in networks. However, a direct link between network evolution and collaboration has not been researched in this manner. Collaboration is further discussed by authors such as Mcnamara (2012), who developed typologies of different kinds of collaboration. However, they do not go into the root causes of the presence or absence of collaboration. While these typologies of collaboration and coordination are a good way to assess the level of collaboration in particular cases, they have not been linked to the causes of the collaboration itself.

Additionally, the research provides the opportunity to research and compare multiple instances of interagency collaboration. Not only does a comparative case study design offer many advantages in terms of generalisability and reliability of the findings, but it is also very well-suited to the specific context of local energy transition policymaking in the Netherlands. The

(14)

structure of the energy regions means that there are dozens of networks in the same policy area with similar goals, scope and context, which means that a comparative study using the most similar systems design would be appropriate. The research on networking, interagency collaboration and jurisdiction-cutting problems relies on case-studies and the comparison of networks that do not share these contexts (Ansell & Gash, 2008). Comparing multiple cases in such a fashion could bring more clarity to the black box that collectively makes up the factors that influence public agencies successfully collaborating.

(15)

2. Theoretical framework

The theoretical framework section will summarize and review the literature on important theoretical topics that this thesis will use in its empirical research. There is a lot of ambiguity regarding collaboration as an academic concept. Therefore, it is vital to first distinguish what collaboration actually means and how it relates to other concepts, such as cooperation and coordination. Once these concepts are properly defined for later operationalisation, the theoretical framework turns to the value of network evolution and how its different forms are compatible with collaboration as defined in the literature. Following from this synthesis of the literature, concrete expectations will be formed to further guide the rest of the research.

2.1. Distinguishing between different stages of organisational interaction

The literature on public interagency problem-solving is extremely diverse and uses many different frameworks and terms categorize and explain public cooperation. The literature on interorganisational theory often uses concepts such as cooperation, coordination and collaboration interchangeably, which compromises conceptual consistency, communication and the comparability of research (Mcnamara, 2012, 390). What further complicates this is the presence of the concept of policy integration, which either refers to the method of attaining greater collaboration or to the specific end goal of having the policies of multiple organisations perfectly aligned into a comprehensive approach. Meijers and Stead (2004, 3-6) capture this conceptual ambiguity perfectly by illustrating the vast amount of concepts that are used and by explaining that most of them refer to the same phenomena.

What becomes clear from the studies of the literature on interorganisational policymaking is that there is a certain amount of hierarchy between different modes of organisational cooperation. Certain concepts refer to deeper interactions while others are only used to denote sporadic or informal cooperation. For example, Meijers and Stead (2004, 5) note that cooperation refers to the weakest form of integrated policymaking while coordination is much more far-reaching (Meijers & Stead, 2004, 5). Interactions between organisations become more numerous, formal and comprehensive (Meijers & Stead, 2004, 5). In their model, integrated policy-making is the end stage, where dependencies and ties between organisations are the strongest. In other models, such as Mcnamara’s (2012), collaboration refers to the most

(16)

integrated stage of joint policy-making, which again points to an overlap of concepts despite them having the same meaning. Thus, Meijers & Stead (2004, 5), Mcnamara (2012) and Keast, Brown and Mandell (2007) all identify differences between cooperation, coordination and collaboration in terms of how strongly the actors are tied together and what the outputs of these different ties of associations are. Briassoulis (2004) contributes to a similar definition, instead focussing on policy integration as a main definition instead of collaboration. Candel and Biesbroek (2016, 214) note that the hierarchy between cooperation, coordination and collaboration is used by many authors in the literature, despite the existence of small and relatively inconsequential differences between the models. The demarcations between these conceptions are blurry, as the relationships between actors in networks are mentioned as crucial factors in determining the mode of interaction (Rayner & Howlett, 2009, 99-100 ; Candel & Biesbroek, 2016, 216-217).

The most important question is how exactly cooperation, coordination and collaboration relate to each other and what they exactly entail. Particular attention is given to Mcnamara’s (2012) conceptual model, since it is one of the most comprehensive and operationalised models in the current literature and because it resonates with findings from a wide array of authors, most notably Tatcher (2007), Keast et al. (2007), Mandell, Keast and Chamberlain (2016) and Mu et al. (2019). She differentiates between cooperation, coordination and collaboration on ten elements from the interorganisational literature: Design, formality of the agreement, organisational autonomy, key personnel, information sharing, decision-making, resolution of turf issues, resources allocation, systems thinking and trust (Mcnamara, 2012). The literature that this framework is based on will also be highlighted, ensuring that complete definitions are being made.

2.1.1. Cooperation

Cooperation is widely defined in the literature as the most basic form of interaction between public agencies. This was not always the case, as the early interorganisational literature often used cooperation and coordination interchangeably (Keast et al., 2007, 13). Meijers and Stead (2004, 5) note that cooperation entails two organisations working together to accomplish their own goals. No joint policies are being made, but there are interactions between low-level civil servants of both organisations. The organisations that are ‘cooperating’ remain completely

(17)

autonomous and the level of sophistication and formality of the interactions is relatively low. Keast, Mandell, Brown and Woolcock. (2004, 368) support this, stating that cooperation entails a limited connection between organisations and that the interactions are of low intensity. In their interviews with public servants, Keast et al. (2007) found that most of them saw cooperation as a potential starting point for deeper integration. Mandell, Keast and Chamberlain (2016, 329) add that cooperation is a form of interaction that poses the least risk for organisations. Mcnamara (2012) provides the most comprehensive account of cooperation, stating that cooperation works within organisational structures, is mainly informal and mainly involves public servants at the lowest levels. Furthermore, trust is not required to make cooperation work, but it may develop when cooperation continues for an extended period of time (Mcnamara, 2012, 391). Lastly, organisations remain virtually autonomous and decision-making happens independently, since not much interaction is needed to facilitate cooperation (Mcnamara, 2012, 392). Concludingly, Keast et al. (2007, 18) describe cooperation as getting the right relationships to increase the performance of both organisations while remaining completely autonomous.

2.1.2. Coordination

Coordination is the second stage of policy integration according to the academic literature. Coordination is marked by more frequent and intense communication between two or more organisations (Keast et al., 2007). On the other hand, Meijers and Stead (2004, 3) note that coordination is often used as an umbrella term for other concepts related to integrated policy making. Nonetheless, Meijers and Stead (2004, 5) note that coordination goes further in terms of integration than cooperation. Coordination is about realising certain individual or collective goals using formal agreements and interactions between high-level personnel. Mcnamara (2012) adds to this, stating that coordination is typically realised through hierarchical structures. Furthermore, organisations become more dependent on each other and high-level leaders work closely together to build relationships of trust (Mcnamara, 2012, 392). Agreements are generally formalised and decisions are being made through centralised hierarchies, even though network action is required to facilitate coordination. Finally, the parties might enlist the help of neutral facilitators or independent brokers to resolve issues or further coordination (Mcnamara, 2012, 392).

(18)

However, a notable difference between Mcnamara’s (2012) conception of coordination and those of other authors, such as Peters (2018) and Meijers and Stead (2004, 5) is that Mcnamara still emphasizes individual goals when discussing coordination. Her argument is that coordination can be described as an instrumental process with formal linkages and interdependencies. Keast and Mandell (2012, 12) also note that coordination happens when there is a certain goal or mission that needs to be achieved. Coordination as described by Mcnamara (2012) and Tatcher (2007) is far from the “integrated policy-making” that authors such as Rayner and Howlett (2009) talk about. Coordination does not mean that an integrated strategy is formed or that organisations effectively pool their resources and undergo deeper ties, but it means that they are working toward goals that the actors agree upon with a medium level of interaction (Mcnamara, 2012). Coordination is more easily formed and that it requires fewer resources to sustain. Therefore, there is a distinct idea in the literature that coordination is not only easier to produce and sustain, but that it is also more desirable in certain situations (Keast et al., 2007; Mu et al., 2019).

2.1.3. Collaboration

The literature defines the most integrated state of interactions between organisations as collaboration. Keast and Mandell (2012, 12) describe collaboration as a form of relationship characterised with high levels of interaction and interdependency. Specifically, actors have to rely on each other to be effective. In that sense, the interdependency of the actors goes beyond resource or informational needs, since a collective commitment is required for collaboration to form and work. Mcnamara (2012, 391) notes that collaboration is a type of interaction between participants to pursue complex collective goals. A shared interest is also necessary to solve interconnected problems. Mu et al. (2019) note that collaboration is more feasible when there is a mutual understanding of interdependency between the participating organisations. The idea here is that the organisations understand that the network as a whole can achieve objectives that individual participants cannot, which warrants the pooling of resources and the shifting of goals towards a collective purpose (Mcnamara, 2012; Keast et al. 2004).

Mcnamara (2012, 392) emphasizes that collaboration works through shared power arrangements, which take form through formal and informal agreements. Through these formal and informal channels, communication is very frequent while decision-making is participative

(19)

in nature. Conflicts are also resolved through deliberation and interaction. Collaboration is thus identified among the most extensive forms of interaction that can occur between organisations in a network. Theoretically, collaboration offers great advantages to the participating organisations: resources can be pooled most effectively to address boundary-spanning issues, while new ways of working together can be explored (Mandell et al., 2016). However, this intensity has a dark side, since collaboration is difficult to develop and maintain (Keast et al., 2004; Mcnamara, 2012). Mandell et al. (2016, 335) support this, stating that collaboration worked in their two case studies, because cohesion and the deliberate leveraging of language were used to develop supportive social infrastructures. Mu et al. (2019, 583) note that organisations involved in collaboration are required to make large changes in their operations, which signifies the necessary commitment and associated risks that the participants need to take. Summing up, the literature describes collaboration as the form of interaction with the highest ceiling of collective achievements. However, it is also not straightforward to initiate or sustain. It is vital to uncover why collaboration succeeds in some cases while it fails to do so in other situations.

2.2. Collaboration determinants

The main question that authors such as Mcnamara (2012) leave unanswered is in what situation which form of the cooperation-coordination-collaboration typology is more desirable or achievable. In the previous overview, collaboration has consistently been described as the most intensive form of interaction between participants, with authors such as Keast et al. (2007) and Mcnamara (2012) stating that it takes time and effort to build the relationships necessary to engage in successful collaboration. The literature on public administration has many models that determine collaboration success. Most of these models are extremely comprehensive, since they explore many causal links and relations. For instance, Mu et al. (2019, 584) aggregate factors from the wider public administration literature into a single framework, which results in them listing factors such as, power disparity, standard operating procedures, meta-governance, perceived interdependence, issue salience, policies and legal frameworks as determinants for collaboration. This is in line with other summaries of the literature. Meijers and Stead (2004) summarize a similarly wide array of factors that can influence collaboration, which relates to Mcnamara’s (2012), Tatchers (2007) and Keast et al. (2007)’s conception of collaboration. Lucidarme et al. (2015) distinguish between antecedent conditions and management factors that

(20)

Harvey (2009). Antecedents include the purpose of the partnership, the environment, network structures and the planning of the partnership. Management factors include a wide array of partnership attributes, communication and decision-making factors. This is in line with Tatcher’s (2007) distinction between collaborative infrastructure, procedures and leadership. In short, there are many models in the academic literature that attempt to completely account for many aspects that can influence collaboration as a process. The focus in this framework is on the structural factors, which relate to the way the participants govern the network influence the actual outcomes of collaboration.

Contrary to most extensive models in the burgeoning literature, this theoretical framework will focus on one particular conceptual area: the effect of past networking experiences on collaboration success. In both the literature on collaboration and the literature on networks, the idea that the organisations have a history of interacting or networking is viewed as an important variable. Mcnamara (2012), Keast et al. (2007) and Tatcher (2007) all mention the importance of long-term interactions for the forming of new collaborative structures. They note that time is necessary to develop the structures, bonds and commonalities that drive collaboration between organisations. However, in Mcnamara’s (2012) definition of collaboration, time as a factor is only paid lip service, even though other models include such a temporal element (Ansell & Gash, 2008; Mu et al, 2019; Candel & Biesbroek, 2016). In the networking literature, similar findings can be discerned, as much research has been done in the evolution of networks and how they mature over time (Provan & Kenis, 2008; Provan & Milward, 2001; Span, Luijkx, Schols & Schalk, 2012). In the next section, the knowledge from both the collaboration and networking literature will be synthesised in order to create expectations regarding the effect of previous interactions on collaboration success.

2.3. Network maturity and collaboration

In the collaboration literature, previous interactions between organisations are deemed important for a number of reasons. First, previous interactions determine the initial level of trust when a partnership originates. Ansell and Gash (2008) note that a history of cooperation or conflict between the organisations can influence the collaborative process significantly. They note that a history of cooperation can lead to high levels of trust and social capital between organisations, which will in turn lend itself to increased commitment and smoother

(21)

communications through self-reinforcing virtuous cycles (Ansell & Gash, 2008, 553-554). On the other hand, a history of antagonism will have the opposite effect, as it can lead to dishonesty and a lack of understanding between participating organisations. However, collaboration can still work in those cases, if the organisations are interdependent and if there are efforts and incentives to create social capital and trust.

Bryson, Crossby and Stone (2006) provide another influential model for understanding collaboration. Their model focusses on initial conditions, the process of collaboration and structural effects. It affirms the model of Ansel and Gash (2008), as they both focus on endogenous factors, such as legitimacy-building and structural or governance factors in the collaboration process. Furthermore, both models emphasize the positive influence of existing relationships or networks. Bryson et al. (2006) state that linking mechanisms, such as legitimacy-building, trust and a shared problem definition make collaboration easier to achieve. Mu et al. (2019, 598) instead point to the importance of exogenous interventions during the collaboration process. Mu et al. (2019) also point to the importance of earlier interactions, stating that early participation in policy design can help agencies clarify goals, interests a shared understanding of problems can lead to an increase in perceived interdependence, which makes collaboration easier to achieve. Ansell, Sorensen and Torfing (2017, 476-477) follow the same reasoning, stating that leadership and previous interactions can lead to a joint ownership of problems, even though in some cases this led to a lot of talk but no actual joint measures. Mu et al. (2019, 586) also note that a failure to solve complex problems on their own can lead agencies to seek collaboration, since it will be regarded as the only way to achieve high level solutions.

The literature on policy networks also notes the value of previous interactions and the existence of previous cooperation patterns. The age of the network and its collective experience is referred to as network evolution in this body of literature. Provan and Milward (2001, 416) state that network evolution is an important determinant to assess network effectiveness. For instance, a network can evolve to include more organisations that contribute to solving a complex problem (Provan & Milward, 2001, 416). Including more organisations across more relevant subsystems is considered as very important to creating integrated strategies (Candel & Biesbroek, 2016). Effective and mature networks can thus ensure that the pooling of resources, information and

(22)

clients happens frequently and seamlessly throughout their organisational fields (Provan & Milward, 2001, 416). Human & Provan (2000, 359) note that networks can evolve in significantly different ways even though they might have certain similarities or are situated in the same sector. Additionally, they found that networks that are created from the ground up have different dynamics than networks that are artificially created. Most importantly, they found that complex network governance structures can help in building legitimacy and embedded relations between network actors, which is in line with findings from the collaboration literature (Human & Provan, 2000, 341-359). Provan and Kenis (2008) support these notions, stating that networks become more formalised over time as demands and need for network-level competencies change. In short, it is much easier to develop embedded relationships when the network is supported by a formal governance structure, which needs time to develop.

Span et al. (2012) also note that there is a relationship between governance roles and performance in local public networks. They note that network evolution is an important mediator in this relationship (Span et al., 2012, 191). Specifically, they note that top-down governance is more likely to happen in young public networks and that more mature networks will shift towards collaboration easier (Span et al., 2012, 192). This logic is based on the findings of Mintzberg (1979), who notes that networks favour top-down decision-making to create more stability. In networks that have matured more, such a lead-organisation role becomes less necessary to create stability. Most importantly, the findings of Span et al. (2012) confirm that the general findings about the relationship between network maturity and collaboration, illustrated in Provan and Milward (2001), Human and Provan (2000) and Provan and Kenis (2008) are also applicable to local networks, where local governments either act as top-down leaders or facilitators. The question arises if there is a relation with collaboration and whether these findings also work for regional networks, where local governments are participants, but not necessarily the lead organisations.

In short, there are numerous links in the literature between the maturity of a network, previous interactions between organisations and the subsequent ease of attaining collaboration. Broadly, the findings from the literature can be summarized as follows: organisations that have previously interacted with each other and have a history of cooperation are more likely to

(23)

collaborate, since there is more trust and social capital. Additionally, the way a network matures influences the likelihood of attaining collaboration. Mature networks more often create complex governance structures, develop deeper ties across more subsystems and are more likely to be governed in a participative manner rather than a top-down governance style.

2.4. Forming expectations from the literature

This theoretical framework has first defined collaboration and set it apart from two other majorly used conceptions in the literature, namely cooperation and coordination. After isolating collaboration as a concept, its relationship with a history of cooperation and network evolution has been laid bare in the literature on collaboration and networking. From this, concrete expectations on this relationship can be formed.

First, the general findings of the literature on network evolution and collaboration all point to one central idea: the maturity of a network matters in creating deep and lasting relationships. Whether it is the development of ‘social capital’ mentioned by Ansell and Gash (2008), the creation of Human and Provan’s (2000) ‘embedded relationships’ or simply fostering ‘trust’, network evolution is mentioned in all of these cases as a main determinant (Provan & Kenis, 2008; Span et al., 2012). Despite these differences in terminology, the idea that time is necessary to develop collaboration or one of its elements is prevalent in the literature (Mcnamara, 2012). These networks have more affinity with coordination, which is associated with less intense forms of interaction and a smaller timeframe to achieve results. Hence, the first expectation is as follows:

Expectation 1: The maturity of a network directly impacts its ability to achieve collaboration.

Second, the literature has unveiled more in-depth propositions regarding the relationship between consistent relationships and the attainment of collaboration. Mature networks more readily adopt collaboration in a number of ways. Primarily, mature networks are able to create and support complex governance structures, which facilitate more practical collaboration aspects such as financial resource pooling, information sharing and joint project development.

(24)

Provan & Human (2000) discuss this most in-depth, but other major networking articles, such as Provan & Kenis (2008) note that over time, networks move towards more brokered and formal forms of governance. These decision-making structures are important for collaboration, as noted by Mcnamara (2012) and Keast et al. (2007). Mcnamara (2012) notes that such structures have to be created by the participants to address collective interests. Transparency and consensus are also mentioned as crucial in this regard (Mcnamara, 2012; Keast et al., 2007).

Mature networks create what Human and Provan (2000) understand as internal legitimacy, meaning that the participants are more aware of each other and their roles in the network. Additionally, Provan & Milward (2001) note that mature networks can address complex problems more easily by involving more organisations from different subsystems. This sentiment is widely held in the collaboration literature, with both Ansell and Gash (2008) and Bryson et al. (2006) noting that networks collaborate more easily when there is a certain cooperative history. In addition to organisational benefits, mature networks also have higher levels of trust and its members possess more awareness of each other (Provan & Milward, 2001; Mu et al., 2019). Ansel and Gash (2008) also note that governance structures can become ambiguous in the context of collaborations, since collaborations are dynamic and complex themselves. It takes time and persistent network interactions to properly delineate the roles of the network participants in these complex sets of interactions. On the other side of the spectrum, unevolved networks are connected by Provan and Kenis (2008) and Span et al. (2012) to top-down governed networks with a simpler lead-organisation governance model. Mature networks have the capability to develop a governance structure that not only coordinates network activities and competencies, but also provide inclusion and participation for all organisations involved, boosting what Mcnamara (2012) defines as the design, decision-making, systems integration and trust elements of collaboration. From this, the second expectation can be formed:

Expectation 2: Mature networks are able to support more complex and participative forms of governance than unevolved networks.

(25)

Lastly, network maturity also affects the timeframe and speed in which collaboration can be realised. A mature network is able to mobilise a wider array of actors and resources to solve a complex interagency problem than a novel network. Additionally, the actors within mature networks have a history of cooperation and often are involved in joint goal-setting and information seeking. Bryson et al. (2006, 46) note that a common problem definition is a linking mechanism that makes collaboration easier to succeed. Candel and Biesbroek (2016, 222) view policy goals as one of the dimensions of policy integration, noting that in situations of collaboration, policy goals are embedded within an overarching strategy. Creating such strategies and implementing them happens quicker in situations where actors are already aware of their interdependencies and where they have a common problem definition. What contributes to this, is that mature networks can mobilise more actors across more subsystems (Provan & Milward, 2001). This has major implications for networks that seek to build and implement such an integrated strategy, such as the energy regions in the Netherlands. The third expectation builds on the practical outcomes that these findings from the literature imply:

Expectation 3: Mature networks are more effective at creating and implementing joint strategies than unevolved networks.

(26)

3. Methodology

In the methodological chapter, several practical aspects of the thesis are clarified. First, the design of the research and its main variables will be explained, together with some of the strengths and logical weaknesses of the chosen design. Then, the variables will be operationalised into indicators in order to make them measurable. How exactly they will be measured will be detailed in the research methods and analysis section below, where the research methods and the unit of analysis will also be fully outlined. Other important aspects which will be addressed are the validity and reliability of the research. The case selection will also be touched upon, together with a brief introduction of the cases and their contexts. Lastly, the data collection and management sections will give a detailed overview of the process of collection and handling of data according to the design and methods previously described.

3.1. Research design & methods 3.1.1. Research design & variables

In order to test the expectations made in the theoretical framework and the answer the research question, a research design of qualitative multiple-case study is chosen. What this means is that multiple cases will be researched and compared in order to arrive at general conclusions on the observed processes in the cases. In essence, the networks of four energy regions will serve as the cases that will be compared in order to test the effect of network evolution on the attainment of collaboration. ‘Past networking experiences’ is the independent variable here and collaboration is the dependent variable. Important here is that two of the observed networks are mature or evolved networks with a history of cooperation prior to the creation of the energy regions and that the two other observed networks do not have such a history. The rest of the wider context of the energy regions remains the same: all of these networks have the goal of creating a conceptual RES and RES 1.0 before the deadlines of June 2020 and March 2021 respectively. Other potential confounding variables, such as network size and the province that the networks find themselves in are either accounted for or are theorised to have a negligible impact on network evolution. Therefore, the cases can be compared using the most similar systems design, since the unique opportunity has risen to compare networks in different stages of maturity, but with comparable sizes, goals and socio-economic and political contexts. The main objective is to research what the networks have achieved and what processes led to this

(27)

point. By comparing these processes between the energy regions, the research question can be answered to a satisfactory degree.

A qualitative multiple case study design has many advantages. According to Toshkov (2016, 264-265), using in-depth case analysis within a comparative case addresses issues of measurement, confounding and reverse causality. Additionally, the effect of network evolution on a concept such as collaboration and its elements requires a qualitative approach. Collaboration is a concept that can only be measured using thick description. The factors that influence collaboration and its elements are numerous and work in subtle ways. Context is thus very important to understand both the cases and the mechanisms behind them. The literature also displays the merits of a qualitative approach. Mcquire and Agranoff (2011, 274) note that the effectiveness of networks can only be studied by providing accurate descriptions of networks, through accounting instances of network management and by describing the broader socio-economic and political contexts.

In addition to the advantages of its qualitative approach, the chosen design also uses its comparative aspects to infer causality. Analysing not one, but several cases with very similar backgrounds and circumstances means that the validity of the results will be much stronger. Additionally, combining intense within-case analysis with a more comparative aspect means that the research not only tests existing propositions from the networking and collaboration literature, but also provides the opportunity to observe the inner workings of the cases to generate new theories or avenues of research.

3.1.2. Operationalisation 3.1.2.1. Collaboration

The first variable that needs to be operationalised is collaboration. The theoretical framework section has already extensively discussed the conceptual differences between collaboration, coordination and cooperation, but a more comprehensive operationalisation will now be provided (Keast et al., 2007, 25; Mcnamara, 2012, 391). First, this thesis differentiates between the more structural and practical elements of collaboration in networks, which are identified

(28)

The first dimension of collaboration, governance structure, is measured through three aspects: the nature of the decision-making structure of the network, the formal inclusion of governmental actors in the network and the inclusion of non-governmental actors in the network. The nature of the decision-making structure of the network is measured by observing the way decisions are being made on the executive level within the network and how this translates to the civil servant level. What is of particular importance here is whether decisions are being made on the basis of consensus or whether all executive decisions are made by a small number of actors. Collaborative networks are more consensus-based than coordinative networks. The formal and informal inclusion of governmental and non-governmental actors is measured by taking note of the governmental and non-governmental actors in the bodies that make up the decision-making structure of the network and what their exact roles are. Of note here are the ratio of governmental and non-governmental actors included in the formal decision-making structure in either an executive or participatory role. For instance, an energy region can still involve stakeholders outside of its decision-making structure through participatory tracks. Collaborative networks include both governmental and non-governmental actors in their decision-making structures or allow for the representation of non-governmental actors in a representative body, while coordinative networks rely on a smaller pool of actors to make decisions or fail to include non-governmental actors.

The second dimension that makes up collaboration is partnership implementation. Partnership implementation entails the more practical outcomes of the network in terms of progress towards collaboration efforts. Partnership implementation is operationalised through shared strategy forming, the methods of internal and external information sharing and financial and resource pooling. Shared strategy forming revolves around the contents of the policy documents that the regions have produced on energy transition and whether the network simply coordinates the different sub-goals and capacities of the actors or whether the documents describe actual joint projects that require resources from each actor for a combined result on each sub-goal. Collaborative regions engage in concrete projects and create specific scenarios for energy transition activities while coordinative networks have more affinity with solely exchanging capacities, interests and solutions. The methods of information sharing will be measured by looking at the internal communication within and between the working groups and the steering

(29)

documents and interviews. Collaborative networks communicate using a wider variety of methods to inform stakeholders within the network than coordinative networks, who do not inform certain actors. Lastly, financial resource pooling is measured by looking at the official planning of the networks and the way the partners commit financial resources to projects and energy transition efforts. In collaborative settings, the network actors pool resources to joint projects. This does not happen in coordinative networks, where the actors merely coordinate on how to invest their own financial resources or when these resources are solely obtained by a higher level of government. The Dutch national government finances the energy regions to a certain degree, but contributions from the involved actors are also necessary to support the energy regions in their activities (Regionale Energiestrategie, n.d.).

3.1.2.2. Networking experiences

As for the independent variable, past networking experiences, this research takes a processual approach to network maturity and operationalises it by referring to the history of cooperation of the networks. This is operationalised by outlining the direct effect of past energy transition networking attempts on the level of collaboration each of the networks. Collaborative networks are expected to rise organically and invest capacity and resources for an extended amount of time in the network. Coordinative networks are built solely using formal structures and typically involve higher levels of government. The existence of organic cooperation on climate change in the energy regions before start RES programme thus gives a strong indication towards the evolution process of the network. Important to this process are the activities of the energy regions before the start of the climate agreement and the RES process. Lastly, this research measures the perceived value of these early initiatives by looking at the starting documents of the RES processes and the conceptual RES of the regions and documenting the value of these early initiatives to the current networking process. An entire overview of the variables, dimension and indicators can be found in table 1.

(30)

Table 1

The operationalisation of collaboration and past networking experiences outlined.

Variable/dimension Indicator Research methods

Collaboration

Governance structure Presence of consensus-based or top-down decision-making structure to govern the region

Documents, interviews

Roles assigned to governmental actors in the formal decision-making structure

Documents, interviews Roles assigned to non-governmental

actors in the formal decision-making structure

Partnership implementation Initiation of concrete regional projects

on each sub-goal identified by the region

Documents

Methods of communication between governmental and non-governmental actors in the region

Extent of the pooling of financial and human resources by the actors in the region

Past networking experiences

Organic versus mandated creation of the network.

Documents

Existence of regional initiatives on energy transition before the start of the National Program RES.

Perceived importance of previous regional initiatives in the starting document of the RES process and the conceptual RES of the region.

(31)

3.1.3. Research methods and analysis

The research methods section outlines how the research question: “How do past networking

experiences affect the level of collaboration in regional networks in the policy field of energy transition in the Netherlands?” is going to be answered using the proposed design of a

qualitative multiple case study. First, the level of collaboration will be described in each network section, leading to an overview of the structural and implementational aspects of collaboration within the network. Then, the previous experiences of the networks will be described as outlined in the operationalisation section, leading to an overview of how the actors in the network started working together and how this has influenced the current progress.

The research will be carried out using document analysis and, when available, interviews with local and regional policymakers. Various strategic documents that describe the overview of the structure of the partnerships by the network itself, the responsible Network Administrative Organisation (NAO) or the participants will be used to study the structural aspects of collaboration. The implementation aspects will be studied using implementation plans, descriptions of joint projects, scenarios and local, regional and provincial environmental policy documents. Qualitative semi-structured interviews will be conducted with policymakers that are directly involved in the energy regions in either a policymaking or implementation role to serve as an additional source of information and to create triangulation. Due to measures related to the outbreak of the COVID-19, interviews could only be scheduled for the cases of Holland Rijnland, Amersfoort and Alblasserwaard (See appendix 1). In this appendix, the general information about the interviews and the respondents is visible. Only the researcher and the first and second reader of the thesis will have access to the full transcripts, including the full names and functions of the respondents. The unit of analysis is the level of collaboration and the evolution process of the four chosen energy regions and the unit of observation are the local, regional and provincial strategic documents, implementation plans and the qualitative interviews with policymakers involved.

The chosen research design and methods attempt to guarantee as much as possible both internal validity and external validity. Internal validity refers to the measurements that the chosen methods capture and whether the examined relationships are indeed causal. The measurement validity in this research is addressed by a thorough examination of historical documents

(32)

together with qualitative semi-structured interviews. These give a direct look into the structural and implementation aspects of the networks and give insight on how these aspects were created and how the networks either have evolved or were created. The use of direct sources from the networks and its participants is extremely important to safeguard the internal validity, in order to get a deeper understanding of the ways network maturity influences collaboration. This allows for the pursuit of direct causal relationships and reduces the possibility of spurious findings in the relationship between network maturity and collaboration. In terms of external validity, which covers how generalisable the findings of the research are, the comparative aspect that this research offers has more advantages than a research with a single case study design (Toshkov, 2016). By studying four networks, the proposed design creates the opportunity to compare evolved to unevolved networks and to compare the networks in both categories against each other to find out the mechanisms behind the evolution processes of the networks.

3.1.4. Case selection

This research analyses and compares the evolution of the networks behind four energy regions in the Netherlands. Two of the networks behind the energy regions have been selected because of the presence of existing network structures and a history of interaction between the participants, which potentially indicates that they have used these previous interactions to support their collaboration efforts. The two other networks are defined as unevolved, as they distinctly lack this prehistory and are relatively new in terms of age. What follows is an overview of the four networks that will be studied.

The first observed network with a prior history of interaction is the energy region of Holland Rijnland. Holland Rijnland has a significant history as a formal network between 13 municipalities in a sub-region of the province of South-Holland. As a network, Holland Rijnland promotes local interaction in the policy fields of housing, mobility, sustainability, energy, tourism and education (Holland Rijnland, n.d.). The network is governed by an NAO, which is an organisation with the specific purpose of managing the network. This NAO is funded and administrated by the municipalities, which in turn are represented in the NAO in various ways. Before Holland Rijnland became an energy region, it already created its own energy agreement in 2017 with various societal parties. In this agreement, the partners pledged

(33)

themselves to the following goals: saving 2,6 Petajoule (PJ) energy, generate 2,5 PJ of sustainable electricity, research sustainable heat energy options and other forms of energy such as biomass and blue energy. As part of this agreement, the region created an implementation plan, consisting of 6 implementation lines, to coordinate the efforts of each municipality. Therefore, when Holland Rijnland was selected as an energy region, it was already developed with a governance structure and was already heavily coordinating on implementation. The energy agreement is expected to be the basis for the RES that the region is expected to develop. To start its RES process, Holland Rijnland has developed a RES starting document in 2019, which is called ‘Van Regionaal Energieakkoord Holland Rijnland naar Regionale Energiestrategie (RES) Holland Rijnland’. In April 2020, Holland Rijnland developed its conceptual RES.

The second observed network with a compatible networking history is the energy region Amersfoort. The municipalities of Amersfoort, Baarn, Bunschoten, Eemnes, Leusden, Soest and Woudenberg have been working together for decades in a way similar to Holland-Rijnland to promote intermunicipal cooperation. The municipalities officially entered a covenant in 2009 to officiate these networking activities, pledging to work together on region issues through periodic strategic agendas. A strategic agenda covers a period of three to four years and determines the policy themes that the region is focussing on for that period. The last strategic agenda, ranging from the period of 2019 until 2022 was created using a participative process where residents and stakeholders played a significant role. Energy was among the topics of this new strategic agenda. Since then, the region has worked together with societal partners, most notably network operator Stedin, to create a comprehensive governance structure together with support from broad societal partners. The outcomes of the strategic agenda have been used as input for the conceptual RES and the region has made the distinction between executive, operational and participative layers that the RES process entails. In the beginning of 2019, the region began working towards a definitive RES with a starting document and in March of 2020, its conceptual RES was finalised and sent to the municipal countries for ratification.

The third network is the energy region of Midden-Holland in the province of South-Holland. Midden-Holland consists of the municipalities of Gouda, Waddinxveen, Zuidplas, Krimpenerwaard and Bodengraven-Reeuwijk. Of these five municipalties, four are situated in