The Dutch approach unravelled : an analysis of the role of power, expertise and interests in the NCICD-program in Jakarta, Indonesia

(1)

The Dutch Approach Unravelled

An analysis of the role of power, expertise and interests in the

NCICD-program in Jakarta, Indonesia

Master Thesis Environmental Geography

June 27th, 2018

Student

Name: Maaike Prins

Student Number: 1694882

E-mail: maaikeprins95@gmail.com

Supervisors

Supervisor: Dr. J. Scholtens

Second assessor: Dr. J. Maiyo

University of Amsterdam

Faculty of Social Sciences

Nieuwe Achtergracht 166

1018 WV Amsterdam

(2)

2 Preface

In front of you lies my Master Thesis “The Dutch Approach Unravelled”. On its own it is the result of six months of hard work, but for me it is the product of four dynamic years of Human Geography education. I started my ‘geography career’ at the University of Utrecht, where I have developed an interest in the combination of sustainability and development. This interest brought me at the University of Amsterdam for the second part of my geography career: A Master. I specialized in Environmental Geography and during the past year I have gained an understanding of the chances and limitations in terms of climate change and a sustainable society. Above all, I learned to stay critical, to focus on the story ‘behind’ the story and I have kept asking myself: “Why are things the way they are and how did they come this way?” This critical perspective and my fascination for climate change and development geography brought me back to the core of my Dutch identity: The fight against water. Although the process of writing this thesis has been a journey on its own, I am thankful for the challenge it has given me, the interesting people I have met and the inspiration it has given me for my future. Even though the journey of this thesis has come to an end, I feel that my journey in the Dutch water sector has just started.

Of course, I could not have written this thesis without the help of my respondents. I have greatly enjoyed the interviews, which not only taught me a lot about the NCICD-program, but also showed me around in the Dutch water sector. Therefore, I would like to thank my respondents for their participation, the interesting interviews and the time they have invested in my research.

Furthermore, I would like to thank my supervisor Joeri Scholtens for his supervision. I have enjoyed our cooperation and I want to thank him for the time he has put in the understanding of my research, thinking along and the productive feedback sessions. I have appreciated his challenging questions and the fact that he pushed me to take the research to a higher level, which I think has made this thesis a worthy end of my Master’s degree. Above all, I have enjoyed his enthusiasm for my research, which has kept me motivated to the end.

Last but not least, I would like to thank Sander and my dear friends for supporting me throughout the process. I think I could not have done it without our coffee breaks at the library, the motivational meme sharing and the countless talks about our researches. Finally, a big thanks to my family for their endless support, mood-boosting diners and sympathy during these past few months.

Maaike Prins Utrecht, June 2018

(3)

3 Abstract

As the result of climate change and human-induced impacts, many delta cities around the world are sinking. Jakarta is one of these vulnerable cities, facing the joint challenge of land subsidence and rapid urban development. The Indonesian government invited the Dutch government, given its global reputation in relation to ‘living with water’ and flood protection, to develop a plan to save Jakarta from flooding in the future. Supported by the Dutch government, a group of Dutch actors developed the so-called National Capital Integrated Coastal Development (NCICD)-program, which supposedly does not only save Jakarta from flooding but also facilitates social-economic development of the city. However, critics have argued that this program has been developed with a hidden agenda of the government and the consortium. This study combined the apolitical Policy Network-approach with the more critical approaches on expertise and framing in order to answer the following question: “How has the NCICD-program been constructed by the perceptions, interests and expertise of the Dutch policy network?” Drawing on 14 semi-structured interviews with a diversity of actors related to the Dutch network working on the NCICD-program and an analysis of 36 documents, this research argues that policymaking in the NCICD-program has been a political process. Overall, four key arguments illustrate the political character of the program. First, the NCICD-program has been a dynamic process, because the differential resources (e.g. decision-making power, agenda-setting power, expertise) available to the various actors in the NCICD-network created a hierarchy in terms of their ability to influence the content, shape and development of the program. Better said, these resources have determined the decision-making process, rather than it being a neutral process to find an optimum solution for Jakarta. Second, the network mainly existed of technical expertise, which gave the program a technical character. As a result of this emphasis on technical expertise, other types of expertise were hardly consulted, hence I argue that the program has been technically locked-in from the start. However, this contradicts with the Dutch Delta Approach, which is also promoted as a participatory approach. Third, I argue that the consortium has mostly been directed by their interests to promote the Dutch water sector, for which they allegedly needed a prestigious program to sell their expertise. However, it is not clear what type of expertise the consortium exactly hoped to sell: The participatory Dutch Delta Approach or the technocratic approach that was deployed in the NCICD-program. As the Dutch Delta Approach is not fully delivered it is questionable to what extent the approach could be transferred to different countries, mainly because the Indonesian context turned out to be quite decisive for the design of the program. Fourth, expertise and interests proved to be important aspects to consider as they have been turned into different problem- and solution framings, leading the actors to possess different ideas about the ‘ideal’ program. All in all, this research has shown that power, expertise and interests have been influential for the development of Jakarta’s flood protection. As these factors are possibly contradicting the interests of the local communities, I state that the policy network should be expanded in order to involve more diverse types of expertise, which would allow to be more sensitive to the local context and to work towards a high-quality solution with public support. This would arguably be more beneficial both for the network, the effectiveness of the plan, and eventually for Jakarta’s flood protection, because an inclusive approach is likely to be more successful on the long-term and in terms of good publicity. If the Dutch are able to achieve this, Jakarta’s flood protection could potentially go hand in hand with an improved reputation of the Dutch water sector.

(4)

4

5

7. Discussion: Analysing the Construction of the NCICD-program ... 52

7.1 The Importance of Resources ... 52

7.2 Expertise as an Influential Resource ... 53

7.3 The Constructive Power of Interests and Frames ... 54

7.4 The “Forgotten” Importance of the Context ... 55

8. Conclusion... 56

8.1 The NCICD-program: A Politically Constructed Product? ... 56

8.2 Looking Ahead: Implications and Recommendations for Further Research ... 57

Bibliography ... 59

Appendix A: List of interviewees ... 64

Appendix B: List of collected documents ... 64

Appendix C: Topic List Interviews Government ... 66

Appendix D: Topic List Interviews Consortium ... 67

Appendix E: Topic List Interviews NGOs ... 68

Appendix F: Topic List Interview O1 ... 69

Appendix G: Topic List Interview O2 ... 70

Appendix H: Topic List Interview O3 ... 70

Appendix I: Codebook Analyse Documents ... 71

(6)

6 1. Introduction

1.1 Problem Statement

Climate change is increasingly viewed as one of the biggest threats to society, in terms of the potential damage to the economy, the environment and national security (Francesch-Huidobro et al., 2017). Scientists have observed changes in, for example, the global mean temperature, the global mean sea level and the distribution and intensity of precipitation, in all of which there exists a global variation. Furthermore, it is likely that low-lying coastal areas (e.g. deltas and coastal plains) will be extremely vulnerable to impacts such as flooding and coastal erosion (IPCC, 2014; Moser et al., 2012). This is risky, because the attractiveness of coastal deltas has led to a growth in population. Globally deltas are home to about 600 million people, whom are mostly living in Asian megacities (Vörösmarty et al., 2009; Renaud et al., 2013). Moreover, urbanization has transformed the natural deltaic processes in such a significant manner that many deltas are now experiencing land subsidence as the result of a decrease in sediment supply from the rivers to the coast (Moser et al., 2012; Francesch-Huidobro et al., 2017). This is arguably a challenging transformation, as research has shown that about two-third of the cities with a population of more than 5 million people are already located in areas not higher than 10 meters above sea level (Moser et al., 2012), and it has been calculated that the subsidence rate of coastal cities range between 6-100 mm/year (Erkens et al., 2015). Additionally, the IPCC projects that by 2100 the global mean sea level rise will be about 3-10 mm/year. The joint occurrence of subsidence and sea level rise is called relative sea level rise, causing coastal cities to ‘sink’ (Deltares, 2015; Syvitski, 2008; Syvitski et al., 2009). Although human activities have been the prime drivers of change in deltas during past decades, there is a high agreement on the fact that climate change will exacerbate the problems in deltas (IPCC, 2014). Therefore, climate change mitigation measures alone will probably not be enough to reduce the vulnerability of these sinking deltas, hence extensive adaptation measures need to be taken (IPCC, 2014).

One of those vulnerable sinking cities is Jakarta, the capital of Indonesia and the 3rd_{biggest megacity worldwide}

(citypopulation.de, 2014). It has been estimated that North-Jakarta sinks about 7.5 to 17 cm/year (Consortium, 2014) and at certain locations this rate is even up to 20-28 cm/year (Abidin et al., 2011). According to Deltares, this means that 25% of the city is likely to be inundated by 2025, if no significant measures are taken (Brinkman & Hartman, 2008). Currently, the city has already experienced an increase in extreme floods, with the 2007-flood being the greatest flood in 300 years. The flood inundated about 40% of the city, caused 76 lives and over 400,000 people to flee the capital. On top of that, the structural damage was calculated to be about $544 million (Brinkman & Hartman, 2008; Consortium, 2014; Oude Elferink, 2017). After this extreme event, the Indonesian government asked the Dutch government to help them analyse the causes of the flood and assist them in the development of appropriate measures to avoid such harm in the future (Brinkman & Hartman, 2008; Oude Elferink, 2017; Sherwell, 2016). In my opinion, it is not surprising that they flew in the Dutch. First of all, Indonesia is a former colony of the Netherlands and since the decolonization the Netherlands is still assisting Indonesia in, for example, their water issues, which is emphasized in three Memorandum of Understandings (MoUs) between the two countries (Bakker et al., 2017; Win, 2017). Second, the Dutch earned global recognition as experts in living with water (Colven, 2017, Win, 2017, Kimmelman, 2017). The Netherlands is located around three deltas, of which the largest is the Rhine-Meuse delta (Wesselink et al., 2007). Historically, the country has experienced subsidence and has been prone to relative sea level rise (Renaud et al., 2013), but technically engineered structures, such as dikes, small-scale flood defences and larger-scale dams, are protecting the country from flooding (Wesselink et al, 2007). This enables half of the Dutch population to live below sea level (Ginting, 2012) and turned the Netherlands in “an undisputed bastion of international expertise on fighting the seas and land reclamation” (Sherwell, 2016).

Upon request of the Indonesian government, the Dutch government sent out a tender on the market, which resulted in the formation of a consortium of Dutch engineers and consultants and since 2012 they have worked on a plan to save Jakarta from flooding: the National Capital Integrated Coastal Development (NCICD)-program. The consortium presented their first Masterplan in 2014, which is currently being revisited in the

(7)

second-NCICD-7

phase. Although the program is still under development, its current design contains the idea of an outer seawall that closes off Jakarta Bay, similar to the Dutch Afsluitdijk. In addition to flood-protection, the program would also facilitate social-economic development of the city (Consortium, 2014). However, the program is facing a strong opposition from critical non-governmental organizations (NGOs) and local (fishing) communities who argue that the program is in fact proposing a ‘pseudo-solution’, which does not tackle the real problem, being groundwater extraction (Leitner et al., 2017; Bakker et al., 2017; Win, 2017; Oude Elferink, 2017). Furthermore, Colven (2017) and Bakker et al. (2017) claim that this program is quite lucrative for the Dutch government and the Dutch water sector. First of all, they argue that the Dutch organizations create a market for themselves in Jakarta. Second, they reason that when the program turns out to be a success, the Dutch will further strengthen their position in the global water sector. Consequently, other deltas might ask the Dutch for help too, since Jakarta is not the only sinking delta that needs help. Additionally, Oude Elferink (2017) states that a lot depends on this plan for the Dutch. She notes that this is the prestigious project “to show the world that the Netherlands is still the undisputed king of water defences”. And this might exactly be the time for the Dutch to show their expertise, as the call for adaptation measures will probably grow considerably during the following years. In fact, the IPCC has emphasized that the benefits of protecting deltas at the global scale are currently larger than the social-economic costs of inaction (IPCC, 2014). As a result of this emphasis on early adaptation, the Netherlands can turn climate change into an opportunity (Kimmelman, 2017).

Although the Dutch have been criticized for their supposedly self-interested involvement, the export of the Dutch Delta expertise is exactly one of the aims of the Dutch government. To support their International Water Ambition (IWA), which aims to ensure water safety and security worldwide and to export the Dutch knowledge and technology, the national government has developed the so-called ‘Top Sector Policy’ (Dutch: Topsectorenbeleid) for the water sector. Moreover, the government developed ‘The (Dutch) Delta Approach’, which summarizes the successful approach of the Dutch which they wish to export. Hereby they recall their historical expertise in the sector as part of their success: “Over centuries, the Dutch learnt to protect their delta,

making it an ever safer place to live while exploiting its potential to the fullest. Dutch water management skills and knowledge is therefore exported all over the world by water managers, engineers and scientists” (NWP, 2017,

p. 6). Besides, the water sector is one of the priority sectors in the Dutch Official Development Assistance (ODA)- policy. The directions of the Dutch development policies changed radically in 2014, when former minister Ploumen introduced the new “Aid and Trade”-agenda. She argued that development cooperation should no longer be aimed only at eradicating poverty, but should also contribute to successes for Dutch companies abroad (Ministry of Foreign Affairs, 2013). Coinciding with these policies, the Dutch water sector is currently also active in the deltas of, amongst others, Vietnam, Bangladesh and Myanmar (Holland Trade and Invest, 2015; Holland Trade and Invest, 2018; Gies, 2017).

In sum, with the assistance of the Dutch government, a consortium of Dutch actors is currently developing a flood-strategy to protect the city of Jakarta. Although the city is facing a serious threat, which could potentially use the recognized Dutch water expertise to solve the problem, critics have argued that the NCICD-program is, under the guise of ‘saving Jakarta’, primarily used to further Dutch private sector interests and promote and safeguard the Dutch international reputation on flood protection. The Dutch government and the consortium are allegedly working together, in order to promote the Dutch water sector and create lucrative opportunities in the future, either in Jakarta or in other sinking cities. This thesis departs from this controversy regarding the Dutch involvement and aims to unravel how the Dutch have actually constructed the NCICD-program and which factors have played a role in the development of the program, thereby paying specific attention to the power of expertise and interests.

(8)

8 1.2 Academic and Social Relevance

So far, scholars have explored Dutch water management in multiple ways. Most research focuses on flood policy in the Netherlands and its history (see Wesselink, 2007; Wesselink et al., 2007; Disco, 2002, Klijn et al., 2012). Furthermore, a few scholars studied the Dutch involvement in the NCICD-program in Jakarta. For instance, Colven (2017) studied the program from a techno-political perspective and argued that it unites political and economic interests, engineering expertise and (post)colonial relations. In her research, she incorporated both an Indonesian and a Dutch perspective. However, her analysis lacks an in-depth exploration of the role of the

different Dutch actors in the program and considered the actors acting as one homogenous group instead. Yet I

argue that it is important to unravel the policymaking processes of the program and distinguish between the actors, as they supposedly have different roles and different behaviour leading to a process which is not as straightforward as Colven describes. Moreover, Leitner et al. (2017) researched the program from a political ecology perspective and looked at the concepts of urban ecology of security and environmental injustice, which they applied to the local communities. Other (critical) research has been done by a coalition of the Dutch NGOs SOMO, TNI and BOTH Ends. Together they published a report, which explains how the Dutch government is entangled in the Masterplan. Their main critique has been that the Dutch government has overemphasized the promotion of the Dutch businesses at the expense of inclusive development in Jakarta. In their opinion the NCICD-program does not offer a sustainable solution for Jakarta’s problems in social, environmental and financial terms (Bakker et al., 2017). Other studies related to the program are mostly concerned with the land subsidence itself and how it is related to urban development (e.g. Brinkman & Hartman, 2008; Abidin et al., 2011; Saputra et al., 2017).

The studies mentioned above provide a rich analysis of the NCICD-program, yet arguably still lack an in-depth analysis of the Dutch involvement and how different internal networks, interests and ideas are combined into

one program. Contrary to the existing literature, this research will not look at the (possible) outcomes of the

NCICD-program for the local communities, but at the Dutch input instead. Furthermore, this thesis will contribute to the discussion on how to understand policy processes by combining the apolitical network approach with the more critical approaches on expertise and framing.

This new combined perspective is not only interesting from an academic point of view, but it is also socially relevant. First of all, deconstructing the development of the NCICD-program by the network of Dutch actors will help outsiders (both people in Indonesia as in the Netherlands) to understand the program better, especially how and why the Dutch companies are engaged in such work overseas. Furthermore, this thesis will contribute to an understanding of the politicized processes behind the program and how these processes constructed the NCICD-program, which could provide a lesson for the other Dutch ‘water’ programs. Moreover, it is relevant to see how the expertise, which is central in ‘The Dutch Delta Approach’, is produced through different networks actors in the case of Jakarta, and how this expertise aligns with the interests and ideas of the actors. As Jakarta is not the only delta the Dutch are involved in, this would also provide a lesson about the ‘export’ of the Dutch Delta expertise and how the Dutch are seizing the opportunities of rising seas and sinking cities.

1.3 Setting the Scene: A Brief Introduction to Jakarta’s Challenges as a Delta

Jakarta is the capital of Indonesia, located on the Western part of the island Java (see figure 1.1). The city lies on a flat coastal area, where thirteen rivers flow their way through the city into Jakarta Bay. Jakarta enjoys a tropical climate, with most rain falling during the monsoon season (November-May). A combination of the low-lying topography and this tropical climate has always made Jakarta vulnerable to flooding during the rainy season (Cybriwsky & Ford, 2001; Delta Cities, 2017). This vulnerability was already apparent during the Dutch colonization (1619-1949), in which the waterfront had turned into a dirty ‘swamp-like’ area with heavy polluted waters (Cybriwsky & Ford, 2001). At that time, the waterfront was turned into an area for the urban poor and Kusno (2001) argues that this area is nowadays still considered as “the other space” of the city, home to the urban poor, fishermen and Chinese communities.

(9)

9

Currently, Jakarta is experiencing a rapid population growth as result of economic development. The population of Jakarta grew from 8.4 million to more than 10 million in the period 2000-2015, and it is expected that this will grow up to about 16 million by the year 2020 (Bakker et al., 2017). Besides, de metropolitan area of Jabodetabek (the area consisting of Jakarta, Bogor, Depok, Tangerang, South Tangerang and Bekasi) is home to more than 30 million people and is considered to be the 3rd_{biggest megacity worldwide (citypopulation.de, 2014). According}

to Abidin et al. (2011), population growth and urbanization have led to several environmental problems, such as water pollution. Besides, urbanization and economic development have increased the vulnerability of the city to flooding, which became clear after the flood in 2007 (Abidin et al., 2011). However, currently the Indonesian government and the Dutch consortium are trying to turn the tide and address Jakarta’s vulnerabilities.

Figure 1.1: Map of DKI Jakarta and Greater Jakarta and their location in Indonesia

1.4 Structure of the thesis

This thesis commences with the theoretical framework (chapter 2), which presents the key concepts of this research and discusses the literature on governance, policy networks, expertise and framing. Eventually, these concepts are brought together in a conceptual model. Hereafter, chapter 3 builds on this conceptual model and presents the research questions of this study. Furthermore, the chapter introduces the methodology and its possible limitations. Consequently, chapter 4-6 discuss the main findings of this research, which are divided in an overview of the NCICD policy context (chapter 4), an analysis of the different aspects of the network, e.g. its actors, their interest and the power structure (chapter 5) and analysis of the different problem and solution frames which have constructed the content and shape of the NCICD-program (chapter 6). Next, chapter 7 brings the main results together and tries to pinpoint why the NCICD-program should actually be seen as a political process. This chapter also revisits the theoretical framework and discusses whether the findings align with the existing body of literature. Finally, chapter 8 concludes this research by answering the research questions and outlining the theoretical implications and the implications for the Dutch Delta Approach. The chapter finishes with a few recommendations for further research.

(10)

10 2. Theoretical Framework

This chapter aims to provide a coherent overview of the literature in three academic fields, which will be applied to deconstruct the NCICD-program in Jakarta. I will introduce the Policy Network Theory (section 2.1) and combine this with debates about the politics of expertise (2.2) and the politics of framing (2.3). Although these academic concepts are mostly used separate from each other, I argue that they are complementary relevant to understand the construction of the NCICD-program in a political and constructivist manner. Here, the Policy Network Theory will be used as a basis to describe the structure of the network. As policy outcomes are supposed to be the consequence of a political process as well, this theory is extended to the debates about expertise and framing. These debates are brought together in section 2.4, which provides a synthesis and visualizes it in a conceptual model.

2.1 Policy Networks

2.1.1 Switch from Government to Governance

Before the 1970s, the majority of scholars claimed that states were governed by strong, central governments based on traditional top-down programs (Hajer et al., 2004). However, since the 1970s critical scholars started to challenge these ideas. Instead, they reconsidered the role of the state in policymaking, as they saw a growing interaction between state and society (Carlsson, 2000; Hajer et al., 2004). The reconsideration of the traditional-style of government was boosted by political scientist Rhodes, who studied the government of Margaret Thatcher in Great-Britain during the 1980s. During Thatcher’s first years as prime minister, the country was experiencing an economic recession. To stimulate the national economy, Thatcher introduced a set of reform measures based on the liberal ideals of free markets, privatization and deregulation. Due to the privatization of public services and the increased influence of the market in the public sector (marketization), the delivery of public services got fragmented. In turn, this created the need for organizations to cooperate in order to deliver these services (Rhodes, 2000; Rhodes, 2007). So, as stated by Teisman (1992), the introduction of neo-liberal concepts in the policy domain had led to fragmentation (or specialization) among actors, which had stressed the need to work together to produce policies and deliver services. Consequently, this switch also introduced non-state, private actors to the policy domain.

As scholars observed the increased complexity of the societal and political realm, they increasingly disapproved of the idea of a central, top-down style of government (Mandell & Steelman, 2003; Sandström, 2008). Instead, they argued that the idea of more cooperative, bottom-up policy arrangements was more applicable to current-day policymaking in Western Europe (Hajer et al., 2004; Al-Habil, 2011; Blanco, 2015). As the central government was no longer seen as the prime agent in policymaking, but just as ‘another’ player next to various public and private actors, scholars argued there has been a switch from government to governance (Hajer et al., 2004; Sandström, 2008, Al-Habil, 2011; Blanco, 2015; Rhodes, 2000). Governance is characterized by actor, multi-level and cross-sectoral types of policymaking, thereby including actors from different scales and backgrounds (Rhodes, 2007; Hajer et al., 2004; Lienert et al., 2013; Van den Brink & Meijerink, 2006). Furthermore, Sandström argues that these multi-actor structures had moved beyond a formal hierarchy to a more horizontal structure (Sandström, 2008). Besides, the switch to governance captures the idea that the top-down approach of the sovereign states, which relies on regulations and formal authority, had been replaced by decentred governance based on shared goals, interdependencies and negotiations (Rhodes, 2000; Sørensen & Torfing, 2005).

So drawing on the typologies of different scholars, I understand that governance differs from government in terms of the number and type of the involved actors, their responsibilities and the means of governance. Notably, the switch from government to governance is an academic concept and this switch is not considered to be something actively pursued by the state. Instead, I see governance as a concept used by academia to simply categorize the changes in the public sector since the 1970s. So the switch talked about in this section summarizes the dominant narrative about the changing role of the state (Hajer et al., 2004). In contrast, critical scholars are more nuanced about, for example, the decrease in power of the state, which will be discussed in the next section.

(11)

11 2.1.2 Hollowing-out of the State, or not?

In section 2.1.1 I argued that the switch to governance coincided with a decrease in power of the central government as it allegedly has to share its ‘traditional’ power and responsibilities with other public and private actors. Following this line of reasoning, Rhodes refers to the ‘hollowing-out of the state’, as one of the consequences of the switch to governance. This idea basically means that the government is assumed to have lost the power to govern effectively by command and has to rely more on diplomacy (Rhodes, 2007). Whereas Rhodes focuses more on power when talking about ‘the hollow state’, Al-Habil (2011) reasons more from the perspective of production. In his view, the state is hollowing-out because the government has to rely heavily on external actors for the delivery of services. Although both scholars have a different perspective, in my opinion they share the notion that marketization had led to a fragmentation in the delivery of services which in turn decreased the influence of the central government in politics (the process of hollowing-out). They both acknowledge that the central government no longer has the prime role in policymaking, as it has to share this with other actors which increases the need to cooperate or negotiate.

However, this idea of hollowing-out has been challenged by several critical scholars. Neglected by Rhodes at first, critics argued that the state’s power did not necessarily decrease, since governments had just changed from harder instruments of control to softer instruments such as bargaining. Rather than referring to a decline of the state, critics refer to an increase in the control of the state but with different means (Rhodes, 2007; Rhodes, 2016). After considering these arguments, Rhodes changed his position and acknowledged that hollowing-out might not be the correct term to capture both the ‘rolling back’ (to a minimalist state) and the rolling out of the state (to the inclusion of other actors in governance) (Rhodes, 2016).

Acknowledging the discussions about the hollowing-out of the state, Sørensen and Torfing (2005) argue that although the state still plays a key role in the governance of different levels, it is increasingly ‘de-governmentalized’ since it no longer has the monopoly to govern “the general well-being of the population” (p. 195). This specifically captures the idea that policymaking is no longer the domain of the state, but that it has become a multi-actor activity. Importantly, these scholars do not argue that the role of the state has completely declined, rather that the traditional view of government is no longer applicable. All in all, this discussion leads to the question how the power between the government and the private companies is divided and to what extent this new role of the central government is applicable to the NCICD-program?

2.1.3 Policy Networks as Mode of Governance

In section 2.1.1, I mentioned that the shift to governance implies that more actors, from various backgrounds, are included in the policy realm. To capture the pluralization of policymaking, scholars came up with the concept of ‘networks’ to describe how policies are made (Blanco et al., 2011; Rhodes, 2000). The concept is not completely new, because in the 60s/70s scholars were already talking about issue networks when discussing the interaction between government and society (Hajer et al., 2004). Carlsson (2000) and Sandström (2008) argue that the bottom-up network approach has been formulated to reconceptualise the more conventional, top-down view of policymaking, which did not capture the complex structure of society in which political power is in fact diffuse and policies are supposedly no longer created within the formal structures of the government (Carlsson, 2000; Sandström, 2008). Therefore, critical scholars came up with the idea of policy networks to describe the interaction between state and society and to capture how actors are ‘doing governance’.

As opposed to the previous hierarchical decision-making structures of government, governance is identified by horizontal networks of public, private and non-profit organizations (Al-Habil, 2011; Hajer et al., 2004; Mandell & Steelman, 2003; Adam & Kriesi, 2007). Several authors came up with a definition for (policy) networks. For example, Carlsson (2000) defines policy networks as “clusters or complexes of organizations connected to each other by resource dependencies and distinguished from other clusters or complexes by breaks in the structure of resource dependencies” (p. 504). Furthermore, Rhodes (2007) refers to a policy network as “a set of formal

(12)

12

and informal institutional linkages between governmental and other actors structured around shared interests in public policymaking and implementation (p. 1244). Lastly, Klijn (1997) defines networks as “more or less stable patterns of social relations between mutually dependent actors which form themselves around policy problems or clusters of resources and which are formed, maintained and changed by a series of games” (p. 30), in which games refer to a continuing set of actions between different actors around issues of shared interest, which are guided by formal and informal rules. Thus, important elements to define networks are the large number of actors, the involvement of actors from different levels and sectors, the non-hierarchical way of policymaking and the mutual dependence on the basis of resources (Carlsson, 2000; Klijn et al., 1995). In this case, I prefer the definition given by Rhodes (2007), since it is broader formulated than the definition of Carlsson and less abstract than the definition of Klijn (1997).

To study governance I also draw from Hermans & Thissen (2008), who argue that policymaking should no longer be thought of as actors pursuing an optimum solution. Conversely, they state that it should be seen as a ‘social process’, in which aspects such as power, relationships, resources, interests and strategic behaviour play a big role as well (Hermans & Thissen, 2008). Taking these aspects into account, they argue that most policies produced through networks are the result of actor’s resources, perceptions and values, which are discussed in the next section.

2.1.4 Resources, Perceptions and Values

First of all, Hermans and Thissen (2008) argue that resources play an important role in the formation of networks and policymaking. Drawing on the ideas of fragmentation and specialization, scholars argue that actors no longer possess all the resources themselves to achieve their objectives, so that these resource-constraints force them to interact with other actors and exchange resources (Teisman, 1992; Van den Brink & Meijerink, 2006; Sandström, 2008; Rhodes, 2007). According to Sørensen and Torfing (2005) and Blanco (2015), the specialization leads to an increased efficiency in policymaking, as many actors can together identify problems and opportunities at an early stage and produce flexible responses which answer to the complexity of the problems. In other words, relationships in networks are connected to the possession of resources and the asymmetry which exists therein (Klijn, 1997). Here, resources are referred to as “[the] practical means and instruments that actors have to realize their objectives” (Hermans & Thissen, 2008, p. 809). Resources can be divided into ‘hard’ and ‘soft’ resources; hard resources refer to, amongst others, money, expertise and formal decision-making power (Van den Brink & Meijerink, 2006; Klijn et al., 1995; Arts & Van Tatenhove, 2005), and soft resources refer to the ‘act’ of diplomacy, communication, framing and bargaining (Blanco, 2015; Arts & Van Tatenhove, 2005).

Although Sandström has argued that policy networks are non-hierarchical, Kashwan et al. (2018) state that resources such as decision-making power come with a difference in access to power. Here, power refers to “the ability of groups (…) to structure the rules” (Kashwan et al., 2018, p. 6), and the scholars argue that actors try to impose change by using power ‘from above and below’. This means that power is not only used by the powerful actors to dominate over the less powerful actors (‘power over’), but that it is also used to create new opportunities and greater control for the ‘other’ actors (‘power to’). Furthermore, Kashwan et al. (2018) recognize three types of power: overt power (decision-making power), agenda-setting power and discursive power, which are put together in the ‘Power in Institutions Matrix’ (see table 2.1).

(13)

13

Table 2.1: The Power in Institutions Matrix

Source: adapted from Kashwan et al. (2018).

Second, Hermans and Thissen (2008) argue that perceptions are important determinants for policymaking in networks. Perceptions embody the images actors have about the world around them (‘reality’), on the basis of which they understand actions and other actors (Hermans & Thissen, 2008). Other important perceptions are the perceptions actors have about the problems and solutions. Eventually actors need to develop a shared perception of the problem and solution for the network to be effective (Mandell & Steelman, 2003).

Third, Hermans and Thissen (2008) state that values are important elements of policy networks. They describe values as the actors’ internal motivations, which provide information about the drivers of the actors (Hermans & Thissen, 2008). Values are based on actors’ interests, which are the benefits an actor expects to derive from the program/network, and on the actors’ objectives, which tend to be more specific and summarize the goals an actor would like to achieve (Hermans & Thissen, 2008). In turn, values are also related to the problem perception discussed above (Mandell & Steelman, 2003). Moreover, values are related to resource dependencies, since it is assumed that actors cannot achieve their own interests and objectives without the help of other actors. Hence, they need other actors and their resources to achieve their more ambitious goals (Teisman, 1992).

In brief, policymaking in networks occurs through different actors who all have their own perceptions and values, and control part of the resources which are needed to achieve their own interests and broader objectives. Since it is likely that actors have competing perceptions and values, policymaking in networks is characterized as a bargaining or negotiation game in which actors seek to maximize their own interests and objectives (Sandström, 2008; Klijn, 1997). Reviewing this debate on policy networks, I argue that the NCICD-program should be deconstructed on the basis of perceptions and values, and how these align or compete with each other. In addition, the idea that policy networks are founded as a result of resource constraints leads to the question whether resources have also been determining for the NCICD-network and the relationships within it? The power matrix of Kashwan et al. (2018) is a good addition to the Policy Network Theory and allows to question whether the resource-constraints in the NCICD-network have also led to a hierarchy in terms of power, as oppose to what is argued by Sandström, that networks are per definition non-hierarchical. Nevertheless, according to Mandell and Steelman (2003) and Adam and Kriesi (2007), other (contextual) factors, such as the type of issues and the political, institutional and cultural context, also determine policy outcomes.

Power Over (by more powerful actors)

Subtypes Decision-making power Agenda-setting power Discursive power

Structural/Institutional power

Controlling international finance, investments, and geostrategic decisions via multilateral and multinational global financial institutions.

Elevating some agendas to the public domain, while excluding others in the pursuit of dominant interests.

Shaping of preferences and values to pre-empt grievances.

Power To (by less powerful actors)

Co-optation Incentivizing behaviour and

outcomes desirable to more powerful actors.

Crafting and popularizing agendas that appeal to multiple constituencies.

Promoting values and discourses germane to the outcomes of interest to dominant actors.

Counter-power and Resistance

Mobilizing socially and politically to mount an overt challenge to the status quo.

Promoting policy proposals and agendas with the specific goal of countering dominant agendas.

Sponsoring other-regarding values and discourses to overcome the barriers of collective action.

(14)

14 2.1.5 Critiques on Policy Networks

The previous sections outlined the main arguments related to the Policy Network Theory. However, critical scholars have challenged this theory and criticized policy networks for their neoliberal, non-democratic nature. According to Bianco (2015), there exists a debate “between those who observe the evolution towards the network paradigm as an opportunity to democratise urban governance and those who point to the highly asymmetrical nature of urban political power and its increasing concentration as a consequence of neo-liberalization” (p. 124). As follows, those who argue that policy networks make governance more democratic emphasize their supposedly horizontal structure and praise the involvement of both public, private as non-profit actors. I would argue that the arguments of the criticist can be divided in two groups: Those criticize the elitist structure of networks and those that question the accountability of the networks.

First of all, critical scholars claim that policy networks are not democratic, because they are assumed to be characterized by an elitist model of power (Blanco et al., 2011; Rhodes, 2007; Blanco, 2015; Klijn & Skelcher, 2007; Sørensen & Torfing, 2005; Stone, 2003). Klijn and Skelcher (2007) argue that networks should be seen as centres of power and privilege, whereby structural advantage is given to private interests in policymaking, which eventually creates a community of strategically powerful actors. Additionally, Blanco et al. (2011) and Sørensen and Torfing (2005) state that networks empower those actors who have crucial resources, hence it is argued that these actors are more powerful than others. As a consequence, the networks are only restricted to these few key players, making them elitist. Furthermore, Blanco (2015) argues that this concentration of power in de hands of a small business elite is a fundamental part of the neoliberal hegemony and reflects the structure of an unequal society. To sum up, this view disputes the arguments that networks entail a more democratic style of governance. In contrast, it states that networks are far from democratic, as they are allegedly restricted modes of governance characterized by an unequal, hierarchical structure.

Second, critical scholars have questioned the extent to which policy networks are in fact accountable. Al-Habil (2011) wonders whom the actors are accountable to if the state is indeed hollowed-out. If the government included external actors to the policy domain to execute their ‘traditional’ tasks, do these external actors also have a responsibility to the citizens? Rhodes (2000) and Hajer et al. (2004) argue that these non-state actors cannot be held accountable, since networks function under a significant degree of autonomy and are self-organizing. Therefore, I would argue that these actors are only accountable to themselves, which allegedly makes them prioritize their own interests above the public interests.

The arguments presented in this section are important to take into account within this research. Following Sørensen and Torfing (2005), who argue that one cannot state beforehand whether a network is democratic, as this depends on the specific network and its context, I wonder to what extent the NCICD-network is democratic and how this is influenced by its context. Although Sandström (2008) claims that networks are non-hierarchical, the critiques presented in this section lead to the question whether or not the NCICD-network is dominated by a power structure and how the different actors function within this structure.

2.2 The Politics of Expertise

2.2.1 Introduction to Expertise

Expertise is an interesting concept. Experts and expertise are words that people use regularly, but there are no clear standards for it. Generally, people are considered experts when they have substantial or authoritative knowledge in a particular field (Newman, 2014), i.e. people are ‘made’ experts when others think of them as people possessing a considerable amount of knowledge in one specific area. Gobet (2015) stated that “an expert in a given domain is somebody who obtains results that are vastly superior to those obtained by the majority of the population” (cited by Gobet & Ereku, 2016). This definition of expertise is quite objective, and does not say anything about its relation to the people using the expertise. For long, this relation was not considered important and academia was dominated by discourses that produced images of expertise as apolitical and knowledge was

(15)

15

supposed to objectively inform politics and find solutions for political challenges (Kleinschmidt et al., 2009). However, famous scholars such as Foucault, Gramsci and Jasanoff have questioned this supposedly apolitical nature of knowledge. Instead, they argue that there is no such a thing as apolitical, neutral expertise (Bock, 2014). Similarly, Bocks states that treating expertise as an external, independent variable is an incomplete understanding of political realities. In section 2.1.4 I argued that within networks, expertise forms an important resource in policymaking. The fact that expertise is allegedly political is relevant for this study, as literature shows that it is likely that expertise is mobilized in such a way that it fits the actors’ values and perceptions. In turn, this has substantial consequences for the policies the actors are responsible for. This section will elaborate more on the supposedly political nature of expertise.

2.2.2 Politics and Expertise

Bock argues that expertise gained importance in policymaking, because knowledge is regarded as a tool to reduce “the uncertainty of the fragmented world we live in” (Bock, 2014, p. 382). That is why different kinds of experts (from universities, think tanks, consultancy firms etc.) are brought in by public sector organizations to support problem definitions, policy solutions and political developments (Stone, 2003). As a result, expertise is considered political, because it boosts credibility and establishes legitimacy of political arguments (Bock, 2014; Van Buuren & Warner, 2014). So actors use expertise to increase their success, which turns expertise into a powerful (political) resource (Turner, 2001). Additionally, it is assumed that actors ‘mobilize’ their own expertise, thereby picking those experts whose arguments support the actors’ own political ideas and interests (Van Buuren & Warner, 2014; Norgaard et al., 2009; Kleinschmidt et al., 2009; Stone, 2003). In my view, this implies that interests influence the kind of expertise is used. For example, politicians who believe in climate change will not use the arguments of climate sceptics to illustrate their arguments. Rather, they quote pro-climate change experts to support their arguments about mitigation measures. In this sense, Stone (2003) notes that knowledge is used as a weapon to serve material interests.

In this context, expertise is allegedly used to promote one’s perceptions and foster one’s interests. That is why I suppose that different perceptions and values shape different definitions, explanations and solutions of problems. To illustrate, Woods (2006) argues that “technical ideas shape politics not because they are ‘the best ideas’, but only where they meet social, organizational or political needs of the moment of key actors” (cited by Mosse, 2007, p. 7). In other words, expertise is mobilized to accomplish one’s goals and interests. Due to this politicization of expertise, Forsyth (2010) argues that one should always consider who benefits from science and how problems are explained, as he states that this is likely to be biased.

Furthermore, scholars have criticized the inequality related to the production of expertise. As I argued before, expertise could be treated as a resource, and as with most resources it is not equally distributed among people. That is why scholars, such as Bock (2014) and Turner (2001), criticize expertise. They claim that the unequal distribution of expertise enables experts to influence non-experts who lack the expertise to understand the problems and solutions themselves.

2.2.3 Influencing Expertise

Critical scholars have not only claimed that politics is influenced by expertise, they also argue it to be the other way; that the production of expertise interacts with politics (Mosse, 2007; Forsyth, 2010; Suhay & Druckman, 2015; Pielke, 2006). For instance, Suhay and Druckman (2015) argue that scientific processes do not occur in a moral vacuum, but are influenced by a range of human values. Likewise, Forsyth (2010) and Mosse (2007) argue that science does not provide an objective picture of the world, but only describes those aspects on which it focuses. In the same way, Pielke (2006) claims that scientists balance findings with their own political and ideological perspectives, which limits them to develop creative policy options. In fact, Suhay and Druckman (2015) argue that values and perceptions can function as some kind of filter, because these factors determine whether one perceives information as credible.

(16)

16

What I draw from this is that experts mobilize their own expertise by selecting only those arguments that fit their frame of reference and do not harm their interests. Suhay (2017) illustrates this with the example of the climate change debate, in which most scientists state that there is enough evidence of climate change to support immediate mitigation measures. However, there are also scientists who argue that there are still too many uncertainties to impose the costs of regulation on the citizens. Suhay claims that these scientists are often connected to industries which are likely to lose profit if they need to adapt. In other words, the personal backgrounds of the scientist determine their position in the climate change debate. Pielke (2006) disapproves this political character of expertise. Instead, he argues that scientists should function as ‘honest brokers’, i.e. scientists should present a range of policy alternatives with their advantages and disadvantages. In turn, it is up to political actors to negotiate about these alternatives to come to one policy.

Furthermore, it should be noted that there are all kinds of expertise, from the more technical kinds to the more political kinds. Arguably, the way environmental problems are approached is the product of how experts ‘frame’ the problem, which allegedly determines the knowledges needed to solve the problem (more about framing in section 2.3). It is argued that when, for example, engineers define problems as technical, they create demand for their own expertise (Wesselink et al., 2013; Sörlin, 2013). In this regard, Fischer (1993) claimed that when experts increase their influence in politics, one can speak of a ‘technocracy’: Governance processes dominated by technically trained knowledge elites, who have gained more powerful positions in decision-making. In his view, through the process of technocratization experts could become the driving force behind political processes, thereby increasing the importance of studying the politics of expertise.

In short, I follow the arguments of several scholars who argue that expertise is politicized. First, actors mobilize particular kinds of expertise to boost their credibility, whereby they strategically choose experts who support their values and interests. Second, experts are influenced by their values and political perspectives, resulting in a biased production of knowledge. So not only does politics influence what expertise is used, it also influences what kind of expertise is produced. In particular, different interests shape different kinds of expertise, which in turn shapes different explanations of problems and solutions. Since expertise differs among people and through time, I argue that expertise could be seen as a construct, created by both experts and political actors. This debate on the politics of expertise is interesting, as knowledge is considered to be an important resource in the Policy Network Theory, which therefore stresses the importance of combining these two academic debates in this research. Hence, I will study how expertise has constructed the development of the NCICD-program. Moreover, the debate on expertise claims that values and perceptions are intertwined with expertise. Accordingly, this leads to the question how values and perceptions have influenced the NCICD-program and network?

2.3 The Politics of Framing

2.3.1 An Introduction to Framing

According to Arnscheidt (2008), ‘frames’ determine how we understand the world around us and explain our version to others. Arnscheidt refers to frames as walls beyond which we do not look, i.e. everything that falls beyond the walls we miss. Hereby frames are supposed to be informed by people’s knowledge, beliefs and values. Furthermore, Rein and Schön (1993) define framing as “a way of selecting, organizing, interpreting, and making sense of a complex reality to provide guideposts for knowing, analysing, persuading and acting” (p. 146). Moreover, Entman (1993) states that framing is “to select some aspects of a perceived reality and make them more salient in communication text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation and or treatment recommendation” (p. 52). What I draw from this is that frames are a kind of glasses through which one looks at the world, which informs one’s version of it and helps to communicate this version of reality. I would argue that ‘the act of framing’ can be used as a persuasive, political tool to promote frames in order to convince others of one’s own perspectives. Moreover, the act of framing is considered to be used by actors to sell a specific project or policy, by which actors frame issues in such a way that they construct a convincing narrative in which their initiative is part of the solution to a better future (Van

(17)

17

Buuren & Warner, 2014). In this way, frames determine what are considered facts and which arguments are left irrelevant, by which frames thus determine the boundary of a problem, the solutions and the allowable alternatives (Wesselink & Warner, 2010). According to Arnscheidt (2008), this makes the concept of framing useful in the context of politics, as policies and laws supposedly reflect certain problem- and solution-related frames. In other words, policies are subjective and shaped by frames. Furthermore, she argues that actors consciously or unconsciously choose to neglect frames, which fall either beyond their ‘walls’ or do not fit their interests.

So frames (and the act of framing) allegedly reflects ones interests (Arnscheidt, 2008; Van den Brink & Meijerink, 2006). In this context, Miller (2015) researched the politics of framing in the case of China’s South-North Water Transfer Project. Although this mega-infrastructure project is very much contested, Miller argued that the Chinese government used water scarcity- and ecological benefits-frames in order to sell the project. These frames functioned to deflect attention away from the more human-induced causes of water stress in the North of China and helped to sell the project as a quick solutions to all its problems, whereas critics argued that the government should come up with more long-term solutions to deal with the underlying causes of water scarcity in the region. Furthermore, Miller et al. (2017) conducted a larger scale research into the politics of framing in relation to big infrastructure projects, and were able to identify frames such as ‘the scarcity-argument’, ‘the economic benefits-argument’ and the ‘there’s no other alternative’-argument to justify the capital-intensive projects. Thus, this supports the ideas that actors use the act of framing to promote one’s own problem and solution perceptions, and thereby ones interests.

2.3.2 Frames and Discourses

Overall, the previous section argued that frames are used to promote one’s perceptions and interests, which in turn determines the boundaries of problems and solutions. That is to say that different perceptions allegedly lead to different explanations of problems and solutions. Hence, I argue that the problem or solution does not exist, but that multiple explanations can co-exist and are related to people producing the frames.

Miller (2015) states that frames correspond to the discourses behind them. Discourses are defined by Hajer as: “a specific ensemble of ideas, concepts, and categorizations that are produced, reproduced and transformed in a particular set of practices and through which meaning is given to social and physical realities” (cited in Smith & Kern, 2009, p.81). According to Hajer and Versteeg (2004), discourse analysis takes a critical standpoint towards the ‘truth’, as it is assumed that there is no such as thing as the truth, rather there are multiple truths. Furthermore, Rein and Schön (1993) refer to ‘policy discourses’, which are “the interactions of individuals, interest groups, social movements and institutions through which problematic situations are converted to policy problems, agendas are set, decisions are made and actions are taken” (Rein & Schön, 1993, p.145). Taking both the definition of Rein and Schön and Hajer into account, it is clear that discourses determine the problematization of phenomena and the formulation of their solutions (Hajer & Versteeg, 2004). Furthermore, discourse analysis helps to understand how certain relations have been produced, as the construction of phenomena also involves the construction of who is supposed to be responsible for a certain problem or who has the authority to solve it (Hajer, 1993).

The analysis of discourses was introduced by Foucault (1971) and has been used by scholars from different disciplines within social science since, for example in policy analysis. The use of discourse analysis in policy studies increased after the publication of the book ‘The Argumentative Turn in Policy Analysis and Planning’ by Fischer and Forester (1993), in which the authors argued that policymaking should be seen as subjective, because they assume that policies are the product of socially constructed problems and solutions. Additionally, Arnscheidt (2008) supports the use of discourse analysis in policy studies, since it enables one to view policies as the outcome of a struggle about how the problem is defined, who is assumed to be responsible and who should have the authority to solve it.

(18)

18

Based on the above discussion, I argue that policy discourses are the outcomes of the interactions between different groups of people, with their own ideas, through which problems and solutions are defined. In turn, these perspectives are ‘sold’ through the strategic use of frames. Reviewing this debate, I recognize that the NCICD-program has potentially been shaped by the ideas and interests of the network-actors, which makes me wonder how the different interests and perspectives of the multiple network actors are being united into one policy outcome. In addition, Hajer and Versteeg (2004) argue that one should also pay attention to the historical, cultural and political contexts, as these contexts allegedly influence one’s ideas. The analysis of the context of the NCICD-program will thus also allow to deconstruct the NCICD-program.

2.4 Synthesis Theoretical Framework

In this chapter, I have explored the three central concepts of this research: Policy networks, politics of expertise and politics of framing. Although the academic concepts are mostly used separately, I argue that they should be used to complement each other when studying policymaking. In my view, the combination of these concepts leads to a better understanding of policymaking. I have visualized my understanding of how these concepts tie together in figure 2.1.

First, academia identified policy networks as the way actors organized themselves to ‘do governance’, which has been described as the inclusion of multiple actors from different backgrounds in the policy domain. Studying networks could therefore give information about the policy outcomes it is responsible for. However, I argue that an evaluation of the actors in an apolitical manner is not sufficient to really unravel how a certain policy has been constructed. Rather than treating policymaking as an apolitical or neutral process, in which different types of expertise are brought together to use the best available body of knowledge to come to a solution for the greater good of all, this research emphasizes the ‘political’ character of policymaking and focuses on how the backgrounds of the different actors have shaped the policymaking process.

Second, the theoretical framework showed that policy networks should be deconstructed by looking at resources, values and perceptions. Together with the way the policy network is organized, these factors are assumed to be influential for policymaking, hence for policy outcomes. So, resources partly determine the way a network is organized, because resource-constraints are argued to shape the relations between the actors. Although some scholars argued that policy networks are non-hierarchical, critical scholars have claimed that they are elitist, as the network is only limited to key players with key resources. As resources are argued to relate to power, I assume that resources are shaping the power structure of the policy network and thereby also influence the design of the NCICD-program. Types of resources are, for example, decision-making power, agenda-setting power, money and expertise.

Third, as said, expertise is one of resources and therefore one of the key factors that determine the policy outcomes of the network. However, the discussions in this chapter have shown that expertise is allegedly not neutral. On the contrary, scholars have argued that expertise is strategically used and produced to fit one’s own ideas and interests. Combining discussions in the framework allows me to assume that expertise is influenced by values, with people mobilizing expertise in such a way that it fits one’s own values (i.e. interests). Furthermore, I argue that expertise is also likely to shape perceptions, as expertise functions as some kind of filter that determines how a problem needs to be solved (supposedly with one’s own expertise/the available expertise). In other words, if expertise is influenced by perceptions and values, it is likely that different perceptions and values shape different problems and solutions. Therefore, I argue that the combination of these elements is essential for a complete understanding of how the NCICD-program is being developed.

Fourth, combining the discussions of the political character of expertise with the academic debate on the politics of framing shows that expertise is likely to be sold through framing, which is a persuasive tool to promote one’s perspectives and to sell a project as a solution for a better future. Similar to the political character of expertise,

(19)

19

framing determines the boundary of a problem and its possible solutions. Hence, I assume that experts sell their expertise in such a way that they shape the problems and possible solutions.

Fifth, literature has also shown that frames reflect discourses, which are the ideas people have about the world that inform their actions. Drawing on the theory of policy networks and discourses, I assume that one’s ‘personal’ discourses are reflected in values and perceptions, as the theory has shown that values and perceptions are also influencing one’s actions in the policy network. On this note, Forrester and Fischer (1993) have argued that policymaking should be treated as subjective, as they claim that it is the product of socially constructed problems and solutions. In other words, the problem or solution does not exist. Better said, it is framed as the problem or

the solution as it fits one’s values and perceptions. So own values and perceptions are sold through frames, which

therefore influences policymaking.

Six, these previous points show why it is important that policymaking should not be considered as an apolitical or neutral process in which complementary types of expertise are brought together to come to an optimal solution for the affected people. Instead, I claim that it should be treated as a political process in which, based on the network’s power structure, certain types of expertise and problem- and solution perceptions are emphasized to foster interests, which in turn direct the policymaking process (in this case the NCICD-program). Nevertheless, discussions on the formation of networks and discourses have shown that contextual factors, such as the historical, political, cultural and institutional contexts, are influential for policy outcomes as well. On this note it is to assume that the entire NCICD-network and –program is shaped within its specific contexts, making it an important point of study as well.

The Dutch approach unravelled : an analysis of the role of power, expertise and interests in the NCICD-program in Jakarta, Indonesia