environmental science & policy 14 (2011) 675684

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/envsci

Co-producing knowledge: joint knowledge production

between experts, bureaucrats and stakeholders in Dutch
water management projects
Jurian Edelenbos, Arwin van Buuren *, Nienke van Schie
Department of Public Administration, Erasmus University Rotterdam, The Netherlands

article info
Published on line 23 May 2011
Keywords:
Knowledge coproduction
Coproduced knowledge
Participation
Network governance

abstract
This article analyzes the process of knowledge co-production between experts, bureaucrats
and stakeholders in two Dutch water management projects. The methods used for coproduction are analyzed, along with the impact of the resulting knowledge on the decisionmaking process. Based on the cases, it is concluded that knowledge co-production between
experts and bureaucrats is not very problematic, because of discipline congruence and
institutionalized relations between the two in Dutch water management. Knowledge coproduction between stakeholders on the one hand and experts and bureaucrats on the other
is more problematic and leads to problems of legitimacy in knowledge production and
decision-making.
# 2011 Elsevier Ltd. All rights reserved.

1.

Introduction

Citizen involvement is increasingly being implemented in

complex water management projects in the Netherlands
(Edelenbos et al., 2010), and also abroad (Rinaudo and Garin,
2005; Petts and Brooks, 2006). Through the active involvement
of stakeholders, new insights, information and knowledge are
brought into evaluation and decision-making processes. The
field of water management has traditionally been dominated
by water professionals from governmental and research
organizations (Lintsen, 2002). A closed and highly interconnected network of water authorities as well as public and
private knowledge institutes has developed as a result (Petts
and Brooks, 2006). With the introduction of stakeholder
involvement, this traditional emphasis on expert knowledge
is now faced with competition from the local knowledge of
citizens. In this new situation, conflicts between expert
knowledge and stakeholder knowledge can easily occur
(McClean and Shaw, 2005). Experts and stakeholders hold

different perspectives, values and motivations, and this leads

to negotiations and struggles in the creation of knowledge that
is to be used in policy-making (Van Buuren, 2009).
This article is interested in the process of co-producing
policy relevant knowledge for the purposes of evaluation and
decision-making between bureaucrats, experts and stakeholders. The much utilized division between expert and
stakeholder knowledge (Rinaudo and Garin, 2005; Petts and
Brooks, 2006) is elaborated upon by the introduction of a third
form of knowledge: bureaucratic knowledge (McClean and
Shaw, 2005), i.e. the knowledge held by bureaucratic officials
(civil servants) about political and administrative procedures
and processes. The worlds of experts, bureaucrats and
stakeholders are rather different (Hunt and Shackley, 1999),
and these differences may hinder the knowledge co-production process between these three domains. Conflicts may arise
and interactions between experts, bureaucrats and stakeholders come to a halt in co-producing acceptable knowledge.
We approach knowledge co-production as the interaction
process between experts, bureaucrats and stakeholders aimed

* Corresponding author. Tel.: +3110 4082635; fax: +3110 4089099.

E-mail addresses: Edelenbos@fsw.eur.nl (J. Edelenbos), Vanbuuren@fsw.eur.nl (A. van Buuren), vanschie@fsw.eur.nl (N. van Schie).
1462-9011/$ see front matter # 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.envsci.2011.04.004

676

environmental science & policy 14 (2011) 675684

at exchanging, combining and harmonizing elements like

facts, interpretations, assumptions and causal relations from
these different knowledge domains. It involves discussion and
negotiation, but in the end leads to a common knowledge
ground, which is authoritative for actors from the different
domains (cf. Ehrmann and Stinson, 1999).
The goal of this article is to describe and analyze the
process of co-producing knowledge among experts, bureaucrats and stakeholders. This article starts off with the
following research question: what is the nature and level of
co-production between experts, bureaucrats and stakeholders in the generation of knowledge for water related
spatial decision-making? This article focuses explicitly on
two themes: (1) the organization of the co-production of
knowledge, and (2) the impact of the co-produced knowledge
on decision-making.
In order to answer the research question, comparative case
study research was conducted on two water management
cases in the Netherlands: one in the Arnemuiden area and
another around Gouwe Wiericke. In both cases, water was an
important issue in the decision-making process. Moreover, in
both cases, attempts were made, in different ways, to relate to
or integrate stakeholder, bureaucratic and expert knowledge.
These cases can be seen as characteristic exponents of the
new water governance approach towards water management in the Netherlands, which is aimed at involving citizens
and stakeholders in the decision-making process (Edelenbos,
2010; Van Buuren et al., 2010).
A range of methods for data collection were used in our
case studies. We conducted participatory observation, which
enabled the researchers to follow the process in detail, in realtime (for a period of more than two years) and to make inside
observations of various meetings. In addition a number of
methods were employed in the process, including document
analysis of a variety of written resources (notes, policy
documents, reports, etc.) and interviews with stakeholders,
experts, bureaucrats and authorities. In-depth semi-structured interviews were held with a dozen key respondents in
each case, representing the different groups of experts,
bureaucrats and stakeholders. By combining more participatory and reflective roles in the research team, the researchers
were able to combine participation and critical distance.
In Section 2, the theoretical focus and analytical framework
of the study is presented. Section 3 describes the two cases,
that of Gouwe Wiericke and the area around Arnemuiden. In
Section 4, the two cases are analyzed and compared on the
level of co-production between the three knowledge domains.
Section 5 provides conclusions and a discussion.

2.

Knowledge in complex decision-making

Knowledge in complex decision-making processes is often

disputed (Van Buuren and Edelenbos, 2004). Nowadays, in
network societies, decision-making has become a social issue
and the mobilization of different sources of knowledge as input
for decision-making has increased tremendously (Pielke, 2007).
The provision of knowledge is no longer solely the domain of
societys elites. The vocal, self-assured, highly educated and
well-read citizen of today has obtained much more insight into

the nature of scientific knowledge (Nowotny et al., 2002).

Knowledge has become public property (Bernstein, 1991).
As a result of this, the relationship between scientific
advice and the policy process has changed as well (Pielke,
2007). Scientists have been knocked off their pedestal and
have lost their a priori authority. The deficiencies of sciencebased appraisal, particularly in complex and uncertain decision
contexts, are not only evident to the public but also increasingly are
questioned by experts who are challenged by counterclaims (Petts
and Brooks, 2006: 1046). A knowledge market has emerged,
with many suppliers (and customers), competing in their
attempts to prove the significance and authority of their
knowledge (Edelenbos, 2005).

2.1.

Expert, bureaucratic and stakeholder knowledge

In complex decision-making, there are a variety of actors who

produce and rely on knowledge that differs in terms of both
content and orientation (Eshuis and Stuiver, 2005), and that
emerges within different institutional and social contexts
(Hajer and Wagenaar, 2003). In general, three types of
knowledge can be identified within the context of decisionmaking processes (Rinaudo and Garin, 2005; Hunt and
Shackley, 1999):

- scientific (or expert) knowledge;

- bureaucratic (or administrative) knowledge;
- stakeholder (lay, practical, non-scientific or professional)
knowledge.
Scientific knowledge is mainly developed by experts, which
in the Netherlands usually are independent of government
(Van Buuren and Edelenbos, 2004). The validity of this type of
knowledge is based on scientific models and methods, and on
the rigorous quality checks of the peer review process (Irwin
et al., 1999). Of course, there are differences between the
natural and social sciences in terms of their use of different
premises, methodologies, norms and values (Nowotny et al.,
2002). This article focuses mainly on technical expertise, due
to its focus on water management.
Bureaucratic knowledge is heavily intertwined with administrative and governmental practices. Within complex
decision-making the participating decision-makers and governmental representatives bring in and mobilize this type of
knowledge to underpin their arguments. It stresses the
political and strategic use of knowledge, and has less of an
emphasis on the substance or intrinsic value of knowledge
(Kingdon, 1984). Bureaucratic knowledge is also based on
professional and scientific grounds, but it has less strict checks
and balances compared to scientific peer review (Lintsen,
2002).
Stakeholder knowledge is grounded in the experiences of
stakeholders, or is related to context or location (Eshuis and
Stuiver, 2005). This type of knowledge concerns local
experiences and insights and is strongly entwined with the
day-to-day activities of people. It is derived from the practices
in which people (inhabitants, entrepreneurs, etc.) are involved. Expert and stakeholder forms of knowledge continue
to be approached differently. Expert-knowledge generation is

677

environmental science & policy 14 (2011) 675684

Table 1 Overview of the differences between expert, bureaucratic and stakeholder knowledge.
Expert knowledge

Bureaucratic knowledge

Stakeholder knowledge

Norm for knowledge production

Warrant for useful knowledge

Scientific validity
Positive peer review
and prospects for publication

Social validity
Level of fit with the business,
local experiences and interests

Core business

Scientific research: systematic

and objectified observations
Validating scientific hypotheses;
expanding the knowledge domain

Policy usefulness
Appropriateness with regard
to standards and warrants of
bureaucracy, and political use
Rule-following behaviour:
bureaucratic practices
Political-administrative
support for proposals

Criteria for success

institutionalised and exclusive and shared through peer-reviewed

processes, whereas lay knowledge is embedded in the world around
and directly impacting on individuals (. . .) (Petts and Brooks, 2006:
1046). Moreover, whereas technical experts often strive for
universal prescriptions (depending on the specific discipline),
stakeholder knowledge is contextual and local (Petts and
Brooks, 2006; Wynne, 1991; Irwin et al., 1999).
These three types of knowledge are essentially part of
different practices. These practices can be characterized by
their different language, their different norms and values,
their different systems of warrants and their own rules of the
game, procedures and criteria (see Table 1).

2.2.
Coproduced knowledge: where experts, bureaucrats
and stakeholders meet
Many scholars, mainly those in the social sciences, stress that
knowledge production is a process of social construction
(Latour, 1999; Knorr-Cetina, 1999). In this social construction of
knowledge, the worlds of experts, bureaucrats and stakeholders are combined and become interconnected (Woolgar,
2000). The way in which this connection is organized affects
the legitimacy of the knowledge used in making policy
decisions and thus the legitimacy of decisions. The proper
organization of this connection is stressed in literature on, for
example, joint fact-finding (Ehrmann and Stinson, 1999),
participatory policy analysis (Hoppe, 1999), collaborative
dialogues (Innes and Booher, 1999), collaborative analysis
(Busenberg, 1999), interactive social science (Caswill and
Shove, 2000), interactive knowledge (Lindblom and Cohen,
1979), cogeneration of knowledge (Petts and Brooks, 2006), and
civic science (Backstrand, 2003). Knowledge in these
approaches becomes a serviceable truth: a state of
knowledge that satisfies tests of scientific acceptability and supports
reasoned decision-making, but also assures those exposed to risk that
their interests have not been sacrificed on the altar of an impossible
scientific certainty (Jasanoff, 1990: 250).
Some authors argue that both expert and lay knowledge
should be used in the production of knowledge (Petts and
Brooks, 2006; Rinaudo and Garin, 2005; Yearley, 2000; Petts,
1997). Backstrand (2003) has identified experts, policy-makers
and citizens as relevant knowledge providers. In such an
approach, there is explicit recognition among traditional
decision-makers that others can fruitfully contribute to the
identification of problems and their solutions, especially when
decision stakes or uncertainty about information is high
(Gallopn et al., 2001; Ravetz, 1999). This requires a more open
approach to what constitutes authoritative knowledge and

Daily life, private business,

defending certain societal interests
Support for ones own interests
and agenda

expertise than the technocratic approach. This raises an

important question of how to balance the different sources of
knowledge within processes of knowledge coproduction
(Backstrand, 2003: 25).
In the interactions between experts, bureaucrats and
stakeholders, knowledge is produced that cannot only
withstand scientific standards (scientific validity), but that
also fits into the system of the bureaucracy (policy relevance)
and has societal relevance or is recognized by stakeholders
(social robustness; Edelenbos et al., 2010). When knowledge is
scientifically valid, socially robust and useful for policymaking, it can be termed coproduced knowledge. Knowledge
that lacks scientific validity turns out to be negotiated
nonsense; knowledge that lacks input from stakeholder
knowledge becomes superfluous knowledge (de Bruijn and
Heuvelhof, 1999); and knowledge that lacks policy relevance
will end up unused because it is politically and administratively inappropriate.

2.3.

Analytical framework

This article aims to provide insight into processes of

knowledge co-production in complex water related decision-making (Fig. 1). The roles and contributions of expert
(1), civil servant/bureaucrat (2) and stakeholder (3) knowledge
in this process are studied, as well as at which moments and
how (methods, process, and instruments) knowledge between
these three domains is co-produced.
The extent to which co-production occurs between expert,
bureaucrat and stakeholder knowledge (4, 5, 6, Fig. 1) is
examined. In addition, the nature of this co-production, if it is
present, is studied, along with the arrangements (models,
instruments, etc.) that are used to organize the processes of
co-production. Finally, the extent to which the three knowledge domains contribute to the process of decision-making is
studied.
Co-production refers to the nature and level of interaction
between the three different forms of knowledge. We define
and approach knowledge co-production as the ongoing
interactions between experts, bureaucrats and stakeholders
in developing usable knowledge that crosses different actor
domains (expert, bureaucrat and/or stakeholder). It involves
exploration, discussion and negotiation on the relevance of
the different knowledge domains and the different aspects
they represent, but in the end leads to a certain common
ground that can be used as a starting point for assessment and
decision-making. In order to measure the level and nature of
this interaction, a three-point scale is used. This scale

678

environmental science & policy 14 (2011) 675684

Fig. 1 Overview of analytical framework.

measures the extent to which co-production of knowledge

takes place. In the case of major interaction, the three
forms of knowledge achieve a joint knowledge base. In the
case of minor or absent interaction, the three forms of
knowledge exhibit little or no co-production and the nature
of the interaction takes the form of (hard) political
negotiations and fights, in which certain forms of knowledge (expert, bureaucratic and/or stakeholder) are not
coordinated or integrated. The following three-point scale
qualitative of nature and is used to measure the level of
interaction between expert, bureaucratic and stakeholder
knowledge. To order to assess the level of interaction on this
scale we use qualitative criteria, which we operationalized
in the following way:
 Major interaction/two-way communication: joint activities
are truly developed to prepare for or to actually develop a
common knowledge base. A major degree of interaction
occurs when stakeholders, experts and bureaucrats display
an open and communicative attitude and actively exchange
information and knowledge with each other, are receptive to
other insights and sources of knowledge, and meet on a
regular and organized basis from the start till implementation of the decision. Their knowledge differences are
displayed through a process of communication, negotiation
and mutual adjustment. There is explicit attention for
bringing together knowledge from different domains by
using certain methods and procedures. Methods and tools
are aimed at realizing an open dialogue and equal

participation of actors with open access for all types of

knowledge.
 Medium interaction/one-way communication: joint activities are developed strategically or symbolically with no real
intention to create a common knowledge base. A medium
degree of interaction occurs when stakeholders, experts and
bureaucrats do meet with one another, but do not actively
express their own information and knowledge. Moreover,
they are not really receptive to other sources of knowledge
and do not succeed in creating a common knowledge base.
There is much difference between the knowledge assumptions and values, which are displayed in one-way communication to the other actors. Interaction is mainly oriented
on convincing the other of own assumptions and epistemological values and does only on parts lead to processes of
harmonizing and combining knowledge. Interaction is more
focused upon consultation: actors can react upon each
other, but a reflexive dialogue is absent.
 Minor or absent interaction: no real joint knowledge coproduction activities are developed. In this level of interaction, the actors show no explicit intention to create a
common knowledge base, and there is too much disagreement and/or mutual misapprehension that leads to diminishing or even stopping of interactions among experts,
bureaucrats and stakeholders. Furthermore, a minor or even
absent degree of interaction occurs when knowledge is
developed in a solitary manner, i.e. when actors show no
intentions to include other values and perspectives, and
when interaction takes the form a priori excluding certain

environmental science & policy 14 (2011) 675684

values and perspectives. No common and supported

knowledge base is created.

3.

Case descriptions

This section describes the two case studies on the integration

of stakeholder, bureaucrat and expert knowledge, namely
Gouwe Wiericke and the area around Arnemuiden. Both case
descriptions are briefly introduced. Subsequently, the three
themes of our conceptual framework are discussed.

3.1.

Case 1: Gouwe Wiericke

The Dutch peat soil meadow areas have significant problems

with their water management. Due to soil surface decline,
water management is a difficult and expensive job in this area
and low ground water levels (which are required to enable
agriculture in these areas) are difficult to realize. The reason
for this is that low water levels fasten soil drop and threaten
the overall water quality, because of the intrusion of salt water
(TNO-NITG, 2000; Arcadis, 2002).
These problems have given rise to the development of farreaching proposals on the part of the Water Board Rijnland in
2000. This water management authority has conducted
research on the feasibility of water retention areas in the deep
polders of Gouwe Wiericke West to solve a couple of water
problems. These studies, conducted by renowned consultancy
bureaus, have shown that water retention in this polder would
be a catch all solution for the problems to do with water quality
and quantity (Water Board Rijnland, 2000).
These studies formed the starting point for a short but
intensive decision-making process initiated by the Province of
South-Holland and the Water Board in 2004. An Environmental Impact Assessment (EIA) was the first step of the formal
planning process to ensure the realization of these retention
areas. When the EIA was launched, fierce resistance from the
inhabitants of the polder was made public, as the general
public absolutely disclaimed the knowledge disseminated
about the problems in the polder (RLG, 2007). The authorities
realized that neglecting them would seriously complicate
future discussions and the implementation of their ideas and
thus, decided to form a Working Group made up of some of the
angriest inhabitants.
In the first few months after its initiation, the Working Group
was unable to do very much as the researchers tasked with
conducting the EIA were busy working on their research, which
was clearly demarcated by principals from governmental
bodies. Stakeholders were not allowed to change the research
focus of the EIA. Coproduction was, in actual fact, not permitted.
Ultimately, the EIA was not finalized at all. During the research
process, it became undoubtedly clear that retention areas
would be too expensive and would also generate too many
technical risks. The Working Group was given the option of
delivering an alternative proposal with more room for agriculture based upon their own local knowledge about the potency of
the area (Royal Haskoning, 2005).
The Working Group eagerly accepted this opportunity and
within a few months, produced an initial rough sketch of the
physical nature of the area (fall 2005). This sketch was the

679

basis for more thorough analyses of the problems and

potencies of the area, as well as the optimal mix of functions.
Experts and stakeholders worked together in the development
of the Working Group proposal, in joint field excursions and
joint design sessions (Van Buuren, 2006).
The Working Group proposal was enthusiastically presented to the various authorities involved. However, the
proposal was not able to convince the officials and governors
who were responsible for decision-making but were only
slightly included in the interactive process. They did not agree
on this proposal, as they believed it contributed too little to the
realization of their ambitions and they could not recognize
their initial ambitions (Province of South-Holland, 2006). The
Water Board and Province decided (Spring 2006) to implement
some small elements of the proposal and postponed the
ultimate decision to a new policy initiative: a Peat Meadow
Contract for Gouwe Wiericke.

3.1.1. The role of stakeholders, bureaucrats and experts in the

process
An interesting alteration of roles in the process described
above can be observed. In the first round, experts dominated
the policy process and expert knowledge was the main
building block for making policy decisions by the Water Board
Rijnland and the Province (Project group GWW, 2004).
Stakeholder knowledge was neglected for a long time and
they were locked out of the decision-making process.
However, after the disappointing results of the EIA, stakeholders were given a much more central role in the process.
They were allowed to develop their own proposal. Experts
were given the role of critical reviewers and providers of
supplementary data. A process of coproduction was organized
(Project group GWW, 2005).
The bureaucrats in this process seem to have had a more
subordinate role in both rounds. Their knowledge of the
ambitions of their administrators, the procedures that their
organizations follow and the conditions put forward by other
policy departments was used to fine-tune the proposal, but
as can be concluded later it was not sufficient to prevent the
debacle that occurred.

3.1.2. Models and instruments used to generate information

and knowledge
In the first round of policy making, the mathematical and
technical methods of the experts dominated knowledge
production. After the installation of the Core Group (composed
of farmers and representatives of local stakeholder groups),
these models were supplemented with field visits and lay
knowledge of the farmers from the polder, although these
were mainly symbolic. As a result, the stakeholders remained
disappointed with the extent of their input in the final EIA
report.
In the second round, a multi-criteria analysis (MCA) was
used to judge the Core Groups proposal and to compare it with
two other scenarios (c.f. Beinat and Nijkamp, 1998). The MCA
which did not have many limitations placed upon it with
regard to form and content was able to bring together expert,
bureaucratic and stakeholder knowledge. The resulting
comparative and quantitative table was seen by all the actors
involved as a good instrument for communicating the

680

environmental science & policy 14 (2011) 675684

proposal put forward by the Core Group but did not convince
the authorities and the bureaucrats who informed them (DHV,
2005).

3.1.3. Use of information in the project and decision-making

stages
Although the added value of the planned retention areas was
convincingly demonstrated (at least in the eyes of the
authorities), its feasibility was not studied at all. The outcomes
of the EIA led to the decision to cancel the retention areas and
to begin an alternative exploration. The Core Groups proposal
brought together a variety of facts and insights, but the
participants appeared not to be able to convince the authorities of the added value and long-term sustainability of the
proposal.
In the second round, only a limited range of bureaucrats
and their knowledge was involved. Emphasis was placed on
the coproduction between experts and stakeholders. Bureaucrats strategically remained at a distance. This is an important
explanation for why the political decision-makers did not
approve the proposal of the Core Group. In general, authorities
especially from the Water Board Rijnland put a higher value
on their internal bureaucratic knowledge as compared to
stakeholder knowledge. They questioned the effectiveness of
the Core Group proposals and were not convinced about the
trustworthiness of the data used (Van Buuren, 2008).

3.2.

Case 2: the area around Arnemuiden

Located in the Zeeland delta in the Southwest of the

Netherlands, redevelopment of a rural area near the city of
Middelburg resulted in severe resistance on the part of the
public, which was not involved in the development of the
plans for the area. From the 1990s onwards, the governmental
planning process for this area reflects a turbulent history
with plans being proposed and successively obstructed.
Governmental parties aimed at initiating a revival of water
recreational functions and a large scale housing project in the
area with these plans, while the local stakeholders mostly
inhabitants of a neighbouring town called Arnemuiden
preferred to maintain the current rural state of the area. As
both ends of the rope hardened their positions, this ended up
creating a situation of deadlock.
In order to mediate these opposing positions, at the
beginning of 2006 an interactive process was set up, that
was aimed at producing a joint vision of the area.1 The aim of
this process was to provide advice on the reorganization of the
relevant area through a participatory process that involved the
relevant (governmental and non-governmental) parties and
organizations as well as local actors, stakeholders and NGOs
with a stake in the issue. This (non-binding) advice was
presented to the city council of Middelburg at the end of 2007.
Currently, the city council is (still) pondering the plan for the
area.
The process started at the beginning of 2006. Stakeholders
were gathered in an Advisory Group. Their task was to develop
1
This project was funded by Living with Water, a Dutch knowledge impulse programme aimed at the study and implementation
of changes in water management.

scenarios for reorganizing the area. Experts on a variety of

subjects and from different backgrounds were gathered in an
Expert Group. They were positioned in a supportive role in the
process, i.e. supporting the stakeholders in the development
of feasible scenarios and answering their questions.
In a series of workshops and meetings, the Advisory Group
formulated scenarios for the redevelopment of the area. Based
on the results of initial interviews and policy analysis, the
Advisory Group developed four dream scenarios, representing the ideal futures of the area irrespective of formal and
technical constraints. The Expert Group reacted to these
scenarios, after which the Advisory Group adjusted them
accordingly. The Advisory Group then valued the scenarios,
revealing the highly valued elements of the scenarios, which
were then gathered into two new scenarios. The Group further
specified these new scenarios and the Expert Group again
discussed the results. Based on the expert comments, the
Advisory Group again adjusted and developed the scenarios
into the eventual advice, proposing to develop the region in an
integral fashion following either of the two scenarios developed.

3.2.1. The role of stakeholders, bureaucrats and experts in the

process
Bureaucrats set the projects aims and limitations. Apart from
existing policy documents and prior (local) agreements, the
project and its outcomes were kept as open as possible.
Stakeholders were actively involved in the process from the
very beginning. Their perspectives were central to scenario
development in the Advisory Group. Experts and bureaucrats
were gathered in an Expert Group. This group was involved in
the process at a later stage, only after the stakeholders had
formulated their initial ideas, and playing a merely reactive
role.
Communication between stakeholders and bureaucrats/
experts was mediated through a Process Team, composed of
members of the scientific and bureaucratic organizations
initiating the project. Former governmental plans for the area
under consideration showed a dominance of bureaucratic and
expert views on the area, and were not recognized by the
stakeholders. The Process Team hoped that placing some
distance between the experts/bureaucrats and the stakeholders in the project would create more room for the
development of stakeholder views and knowledge in the
development of scenarios and would prevent further domination by expert views. The aim was to develop a body of
coproduced knowledge through combination of the findings of
the various groups.

3.2.2. Models and instruments to generate information and

knowledge
In order to identify and utilize stakeholder values (knowledge)
for scenario development, an approach of covaluation
(collaborative valuation) was developed during the process
(Van Schie and Bouma, 2008; Van Schie, 2010). Stakeholder
values and perspectives on the relevant area were identified in
interviews. The scenario development was started based on
this information. The stakeholder perspectives were characterized by statements on what was important to them,
usually expressed in qualitative terms. Experts (both external

environmental science & policy 14 (2011) 675684

and governmental) were interviewed as well, revealing their

main points of concern for the redevelopment of the area.
These concerns were mainly to do with the technical and
financial aspects of the plans.
In order to integrate the perspectives and achieve coproduction, the process aimed to develop a broad analysis
combining the financial measures and perspectives provided
by the experts as well as the non-financial perspectives and
values provided by the stakeholders. This proved difficult as
the stakeholders stated they preferred to involve no financial
information in such an analysis, while the bureaucrats
pressed the need for including only financial information.

3.2.3. Use of information in the project and decision-making

stages
The project was aimed at providing advice that was supported
by stakeholders as well as bureaucrats and experts. The
project took the perspectives of stakeholders as a starting
point: the end result was predominantly a product of the
stakeholders involved. Still, the bureaucrats and experts
actively supported the stakeholders. The stakeholders
implemented most of the changes proposed by the bureaucrats and experts and generally respected their expertise on
professional subjects, which resulted in more or less
coproduced results. The bureaucrats and experts themselves, however, saw no use in the inputs of the stakeholders
in the process as, they argued, they lacked scientific grounds
and expertise.
The decision-makers and politicians approached the
advice with reservation and a severe dose of scepticism.
The municipality was used to base decisions on short-term
financial effects and knowledge provided by (bureaucratic)
experts only, and as a result bureaucrats felt that there was
inadequate information for decision-making. Experts felt they
had not been able to contribute their specialized knowledge.
Since they were unable to process the interactive results
within the existing procedures of decision-making, politicians
and municipal civil servants postponed making a decision on
the advice.

4.

681

Case comparison and analysis

This section compares and analyzes the two cases on the way
in which knowledge is (co)produced and implemented in the
process. First, an overview table for the two cases is provided.
Second, the cases are discussed in more detail and their main
differences and similarities are highlighted.

4.1.

Overview of knowledge production in the two cases

Based on the analytical framework presented in Section 2,

Table 2 provides an overview of the knowledge production
processes in the two cases described above.

4.2.
Coproduction of stakeholder, expert and bureaucratic
knowledge
In both cases, experts as well as civil servants and stakeholders provided knowledge that was relevant to the decisionmaking process. The interactions between these different
sources of knowledge differed and were problematic for a
number of reasons. These are discussed in detail below. Table
3 provides a classification of the levels of interaction between
the different sources of knowledge following the analytical
framework.

4.2.1.

Coproduction among experts and bureaucrats

This kind of interaction is well established in existing

institutions and working procedures of knowledge generation
and decision-making. It was observed in the cases that experts
tend to listen to their principals from within the government.
They want boundaries and conditions to be established for the
knowledge that they must produce. Moreover, experts focused
on issues within their field of expertise and did not easily
transcend the boundaries of their field of expertise (in the case
of the area around Arnemuiden).
In the case of Gouwe Wiericke, in later phases of the
project, civil servants remained at a distance from the experts

Table 2 Overview of knowledge production in the two cases.

Theme

Case 1: Gouwe Wiericke

Case 2: Area around Arnemuiden

Role of stakeholders,
bureaucrats and experts

Stakeholders were first neglected but,

after fierce resistance, they were involved
Experts were dominant at first. Later they
functioned to fine-tune the stakeholder proposal
Bureaucrats played a subordinate role,
especially in the second round
The EIA was opened up to stakeholders: they
could bring in knowledge, but the scope could
not be altered
This was followed by more participative methods
and a joint MCA
Stakeholder knowledge was used for the development
of the alternative plan. Expert knowledge was the basis
for authorities to abandon the alternative of retention
areas. However, their alternative plan was refused
by bureaucrats and decision-makers because they
distrusted its factual underpinning

Stakeholders received a prominent position

Experts played a more reactive role
Bureaucrats were involved from the start in
the process, in the set up of the process and
in monitoring its course

The use of instruments

and models to coproduce
knowledge

Use of knowledge in project

and in decision-making

Both stakeholders and experts were interviewed.

Stakeholders developed scenarios in workshops
A broad analysis involved both qualitative and
quantitative information
Plans were developed based on stakeholder
knowledge. Both expert and stakeholder
knowledge was used to make the plans
more feasible. Bureaucratic knowledge
stressed short term financial feasibility.
Decision-makers preferred expert and
bureaucratic knowledge.

682

environmental science & policy 14 (2011) 675684

Table 3 Assessment of the coproduction of different sources of knowledge in the two cases.
Cases

Coproduction of

Gouwe Wiericke
Around Arnemuiden

(1) Experts and civil servants

(2) Experts and stakeholders

(3) Civil servants and stakeholders

Medium (+/ )
Medium (+/ )

Medium (+/ )
Minor ( )

Minor ( )
Minor ( )

in order to prevent commitment to the process and the

outcomes. In the case around Arnemuiden, a similar attitude
was seen on the part of civil servants; they were reluctant to
start interactions with the experts in the expert group. The
experts regarded their role in the interactive process as being
very different compared to their common tasks in the internal
organization. In reaction to this, the civil servants started their
own internal working routines and developed project groups
inside the municipal bureaucracy, which were kept at a
distance from the project.

4.2.2.

Coproduction among experts and stakeholders

This kind of interaction was problematic in both cases.

Stakeholders got involved too late in the water management
process (first phase in the case of Gouwe Wiericke). In working
on the EIA, experts had a predefined research question.
Classical instruments were used to facilitate the interactions
between experts and stakeholders: experts presented their
research approach and stakeholders could reflect upon this.
However, after the cancellation of the EIA and the retentionalternative, the level of interaction between experts and
stakeholders changed drastically. Stakeholders and experts
collaborated to maximize the quality of the proposal and
found methods (MCA, joint field visits, etc.) to realize
synchronization. The interactive stakeholder approach in
the case around Arnemuiden did not allow for direct
interaction between experts and stakeholders, as the process
facilitators feared that the experts would dominate such an
interaction through the use of difficult technocratic speech.
The experts also had the opportunity to critically reflect on the
drafts of the scenarios. However, they were more comfortable
questioning the scenarios than providing concrete suggestions for improvement. Even though this prevented a clash
between different knowledge bases from occurring, the
experts did not take the stakeholder input as being very
useful to the process of scenario development.
The tools, instruments and methods for knowledge
generation that are commonly used by experts and decision-makers can hinder the process of joint fact-finding and
coproduction of knowledge (both cases). Such models or
instruments can be strongly expert-driven. The application of
these methods is often rigid, as shown in the fruitless
attempts in the case of the area around Arnemuiden to
combine stakeholder information with bureaucratically required financial information. Also, the procedure of the EIA in
the Gouwe Wiericke case could not be adjusted to the wishes,
questions and input of the stakeholders. Such methods thus
reinforce the distance between expert and stakeholder
knowledge. Moreover, the use of different methods to involve
interests and knowledge did not stimulate the coproduction
between stakeholder and expert input (in the case around
Arnemuiden). The different backgrounds and orientations of
the experts and stakeholders were not reconciled in the

project. More informal and qualitative methods are required to

bridge the gap between stakeholders and experts (as observed
in the second round in Gouwe Wiericke).

4.2.3.

Coproduction among civil servants and stakeholders

This kind of interaction was troublesome due to the existing

rigid (technocratic) forms of knowledge production and
decision-making, which exclude stakeholders from the
process.
In both cases, civil servants were not receptive or
responsive to the knowledge provided by stakeholders. They
showed an attitude similar to that of the experts in both cases.
This can be explained by the fact that the civil servants in the
field of water management and urban and regional development are often experts in these domains themselves, at least
in The Netherlands.
Besides, newly developed models or approaches (in the
case around Arnemuiden) based on coproduced knowledge
between experts and stakeholders may not be accepted by
civil servants and policy-makers, because they do not fit into
the regular working methods and procedures within bureaucracy. Institutionalized procedures and methods that focus on
expert-knowledge (perspectives and values) complicate the
development, implementation and legitimization of new
approaches that integrate stakeholder knowledge. In the
decision-making stage, it turned out that decision-makers
and politicians were unable to incorporate the results of the
interactive process in the existing institutions and decisionmaking procedures (in the case around Arnemuiden).

5.

Conclusion

This article has reported on research on two cases in order to

gain an understanding about the nature and level of
coproduction between experts, bureaucrats and stakeholders
in generating knowledge for water-related decision-making. It
is acknowledged that the categories of experts, bureaucrats
and stakeholders are generalizations that neglect the (sometimes huge) diversity within these categories. The selection of
experts and stakeholders in the cases described here is just as
much a political and biased process as is the process of
knowledge coproduction. That means that the starting
conditions of a process of knowledge coproduction are already
(implicitly) politicized.
Besides this, this comparative case study research showed
that the three-fold interaction between these three categories
of actors is problematic and that many forms of politics distort
attempts for coproduction. The two cases showed that, in
general, both experts and bureaucrats are not willing to
acknowledge that stakeholder knowledge has the potential to
improve the identification of problems and the search for
feasible solutions that address the circumstances of the direct

environmental science & policy 14 (2011) 675684

environment. Stakeholders, on the other hand, do not have

much eye for the scientific soundness of knowledge development and the political and strategic relevance of knowledge.
They focus on what is relevant or appropriate for them, and
often undervalue and distrust the input of bureaucrats and
experts. In both cases it was observed that stakeholders
questioned the expert input and did not take their input for
granted.
The level of interaction between experts and civil servants
is at least compared to the other two strands of coproduction
better developed. The reason for this is that experts often
receive research assignments from civil servants. Their
interests are interconnected. Moreover, in the field of water
management and urban and regional development, civil
servants and (external) experts often come from similar
disciplinary backgrounds, which make it easier for them to
communicate and to interact. However, when experts get in
touch more with stakeholders, as in the case of Gouwe
Wiericke, it can be seen that civil servants move strategically
away in order to maintain some room to maneuver and dodge
commitment.
Some methods used for knowledge production are better
applicable to realize coproduction compared to other methods. In our cases the multi criteria analysis (Gouwe Wiericke)
and the method of covaluation (Arnemuiden) enable the
various groups to bring in their knowledge and to integrate
this knowledge. In addition specific techniques of knowledge
mobilization and exchange can be helpful to realize coproduction as we saw with regard to for example field visits and
scenario workshops. However, it turned out to be that the way
in which methods of knowledge production are used and the
intention of the involved actors to combine and harmonize
knowledge is more decisive for realizing coproduced knowledge, then the methods themselves. The level of interaction in
the method used is important for realizing coproduced
knowledge.

6.

Discussion

This research provides a strong indication that the fragmentation of perspectives and values between experts, civil
servants and stakeholders is a strong force and incentive
for a modular and fragmented knowledge process. Visions and
viewpoints of experts, civil servants and stakeholders reflect
their deeply rooted belief systems, ambitions and values (cf.
Rinaudo and Garin, 2005: 287). All the actors involved (experts,
stakeholders, civil servants, and also decision-makers) have
different perspectives on the relevant issues, and different and
opposing interpretations and assumptions are involved. It is
difficult to achieve a body of information that is recognized
and accepted as authoritative and useful for all parties
involved. In both cases, it was shown that experts, bureaucrats
and stakeholders use different norms and criteria for
knowledge production, ranging from scientific validity
(experts), policy usefulness (bureaucrats) and social validity
(stakeholders). These different orientations lead to different
valuing of the relevance of knowledge to be used for
assessment and decision-making. Therefore, it can be
concluded that the three-fold coproduction between experts,

683

bureaucrats and stakeholders is characterized by politics of

coproduction in which certain values and perspectives in
knowledge production are biased out. Participation is certainly
not a panacea for realizing coproduced knowledge.
The differences in appreciation of different knowledge
sources do have consequences for their impact on decisionmaking. Decision-makers traditionally rely on expert and
bureaucratic knowledge, which they value more as compared
to stakeholder knowledge. This kind of knowledge therefore
has an impact on decision-making. However, stakeholders do
not always support decisions that are based on expert and
bureaucratic knowledge, as has been shown in the two cases
presented in this study. As their knowledge was not included
in a joint knowledge base for decision-making, they did not
consider the decision to be legitimate. This includes the risk of
stakeholder strategies to postpone or stop decision-making,
resulting in a low impact of knowledge on decision-making.
Concluding, it is important that stakeholders as well as
experts and civil servants are involved at an early stage and
that their input is jointly combined in assessment methods,
models or instruments which enable coproduction. These
methods should be anchored to the actual decision-making
procedures, in order to be considered legitimate by decisionmakers. Only when the knowledge of bureaucrats, experts and
stakeholders is properly synchronized, will this result
in scientifically valid, policy relevant and socially robust
knowledge that has the potential to impact decision-making,
because experts, bureaucrats and stakeholders not only can
recognize their own insights, but they also agree about the
result of the coproduction: knowledge which is harmonized
and combined between different sources, and forms a new,
overarching and integrative body of knowledge which is
collectively perceived authoritative to underpin decisions.

references

Arcadis, 2002. Verkennende studie duurzaam waterbeheer

Gouwe Wiericke. Rapport 110402/WA2/3J7/000535/001, 14
Augustus 2002.
Backstrand, K., 2003. Civic science for sustainability: reframing
the role of experts, policy-makers and citizens in
environmental governance. Global Environmental Politics 3
(4), 2441.
Beinat, E., Nijkamp, P., 1998. Multi-criteria Analysis for Land-use
Management. Kluwer Academic Publishers, Dordrecht.
Bernstein, R.J., 1991. The New Constellation: The Ethicalpolitical Horizons of Modernity/Postmodernity. MIT Press,
Cambridge Massachusetts.
de Bruijn, J.A., Heuvelhof, E.F.ten, 1999. Scientific expertise in
complex decision-making processes. Science and Public
Policy 26 (3), 179184.
Busenberg, G.J., 1999. Collaborative and adversarial analysis in
environmental policy. Policy Sciences 32, 111.
Caswill, C., Shove, E., 2000. Introducing interactive social
science. Science and Public Policy 27 (3), 154157.
Edelenbos, J., 2005. Institutional implications of interactive
governance: insights from Dutch practice. Governance 18 (1),
111134.
Edelenbos, J., van Schie, N., Gerrits, L., 2010. Organizing
interfaces between government institutions and interactive
governance. Policy Sciences 43 (1), 7394.

684

environmental science & policy 14 (2011) 675684

Edelenbos, J., 2010. Water as Connective Current. Boom/Lemma

Publishers, The Hague.
Ehrmann, J.R., Stinson, B.L., 1999. Joint fact-finding and the use of
technical experts. In: Susskind, L., McKearnan, S., ThomasLarmer, J. (Eds.), The Consensus Building Handbook: A
Comprehensive Guide to Reaching Agreement. Sage, London,
pp. 375400.
Eshuis, J., Stuiver, M., 2005. Learning in context through conflict
and alignment: farmers and scientists in search of
sustainable agriculture. Agriculture and Human Values 22
(2), 137148.
Gallopn, G.C., Funtowicz, S., OConnor, M., Ravetz, J., 2001.
Science for the twenty-first century: from social contract to
the scientific core. International Journal of Social Science 168
(2), 219229.
Hajer, M., Wagenaar, H., 2003. Deliberative Policy Analysis:
Understanding Governance in the Network Society.
Cambridge University Press, Cambridge.
Hoogheemraadschap van Rijnland, 2000. Waterbeheersplan
20002004, September 2000.
Hoppe, R., 1999. Policy analysis, science and politics: from
speaking truth to power to making sense together. Science
and Public Policy 26 (3), 201210.
Hunt, J., Shackley, S., 1999. Reconceiving science and policy:
academic, fiducial and bureaucratic knowledge. Minerva 37,
141164.
Innes, J.E., Booher, D.E., 1999. Consensus building and complex
adaptive systems: a framework for evaluating collaborative
planning. Journal of the American Planning Association 65,
412423.
Irwin, A., Simmons, P., Walker, G., 1999. Faulty environments
and risk reasoning: the local understanding of industrial
hazards. Environment and Planning A 31, 13111326.
Jasanoff, S., 1990. The Fifth Branch: Advisers as Policy Makers.
Harvard University Press, Harvard.
Kingdon, J.W., 1984. Agendas, Alternatives and Public Policy.
Little, Brown and Company, Boston.
Knorr-Cetina, K., 1999. Epistemic Cultures: How the Sciences
Make Knowledge. Harvard University Press, Cambridge.
Latour, B., 1999. Pandoras Hope: Essays on the Reality of
Science Studies. Harvard University Press, Cambridge.
Lindblom, C.E., Cohen, D.K., 1979. Usable Knowledge: Social
Science and Social Problem Solving. Yale University Press,
New Haven, CT.
Lintsen, H., 2002. Two centuries of central water management
in the Netherlands. Technology and Culture 43, 549568.
McClean, S., Shaw, A., 2005. From schism to continuum? The
problematic relationship between expert and lay knowledge:
an exploratory conceptual synthesis of two qualitative
studies. Qualitative Health Research 15 (6), 729749.
Nowotny, H., Scott, P., Gibbons, M., 2002. Re-thinking Science:
Knowledge and the Public in an Age of Uncertainty. Polity
Press, Cambridge.

Petts, J., 1997. The public-expert interface in local waste

management decisions: expertise, credibility, and process.
Public Understanding of Science 6, 359382.
Petts, J., Brooks, C., 2006. Expert conceptualizations of the role of
lay knowledge in environmental decisionmaking: challenges
for deliberative democracy. Environment and Planning A 38,
10451059.
Pielke, R.S., 2007. The Honest Broker: Making Sense of Science in
Policy and Politics. Cambridge University Press, Cambridge.
Provincie Zuid-Holland, 2006. Brief Lid Gedeputeerde Staten:
Voortgang Gouwe Wiericke West. 1 februari 2006.
Raad Landelijk Gebied, 2007. Achter open deuren. Over de
provinciale regierol bij gebiedsontwikkeling. RLG, Utrecht.
Ravetz, J.R., 1999. What is post-normal science. Futures 31 (4),
647653.
Rinaudo, J.D., Garin, P., 2005. The benefits of combining lay and
expert input for water-management planning at the
watershed level. Water Policy 7, 279293.
Royal Haskoning, 2005. MER Integrale inrichting Gouwe
Wiericke West. Concept rapport. 10 januari 2005.
TNO-NITG, 2000. Uitwerking duurzaam waterbeheer in het
herinrichtingsgebied Reeuwijk. Bruisend water als leidraad
voor waterberging en dynamisch peilbeheer. TNO-Rapport
NITG-00-128-B. Delft.
Van Buuren, A., Edelenbos, J., 2004. Conflicting knowledge: why
is knowledge production such a problem. Science & Public
Policy 31 (4), 289299.
Van Buuren, A., 2006. Competent Decision-making. Ph.D.
Thesis, Lemma, Utrecht (in Dutch).
Van Buuren, A., 2008. Vermorste kennis en de kwaliteit van
publieke wilsvorming. Bestuurskunde 2008 (2), 3140.
Van Buuren, A., 2009. Knowledge for governance, governance of
knowledge. Inclusive knowledge management in
collaborative governance processes. International Public
Management Journal 12 (2), 208235.
Van Buuren, A., Edelenbos, J., Klijn, E.H., 2010. In woelig water.
Water governance op het snijvlak van adapatief waterbeheer
en ruimtelijke ontwikkeling. Boom/Lemma Publishers, Den
Haag (in Dutch).
Van Schie, N., Bouma, J.J., 2008. The concept of covaluation:
institutionalizing the involvement of local (public) values in
regional planning on water. Competition and Regulation in
Network Industries 9 (4), 363396.
Van Schie, N., 2010. Covaluation in Spatial Water Management,
PhD thesis, Rotterdam.
Woolgar, S., 2000. Social basis of interactive social science.
Science and Public Policy 27 (3), 165173.
Wynne, B., 1991. Knowledge in context: science. Technology,
and Human Values 16, 111121.
Yearley, S., 2000. Making systematic sense of public discontents
with expert knowledge: two analytical approaches and a
case study. Public Understanding of Science 9,
105122.