Managing Stakeholder Involvement in Decision Making: A Comparative Analysis of Six Interactive Processes in the Netherlands

Abstract

Initiatives to encourage and stimulate the involvement of citizens but also various societal organizations in decision making can be seen in a wide variety of European countries. Citizen panels, citizen charters, new types of participation, and other forms are being used to increase the influence of citizens on decision making and to improve the relation between citizens and elected politicians. In the Netherlands a lot of local governments have experimented with interactive decision making that is enhancing the influence of citizens and interest groups on public policy making. The main motives to involve stakeholders in interactive decision making are to diminish the veto power of various societal actors by involving them in decision making, improve the quality of decision making by using the information and solutions of various actors, and bridge the perceived growing cleavage between citizens and elected politicians. In this article six cases are evaluated. The cases are compared on three dimensions: the nature and organization of participation, the way the process is managed (process management), and the relation with formal democratic institutions. These organizational features (in terms of both formal organization and actual performance) are compared with the results of the decision-making processes in the six cases. The article shows that the high expectations of interactive decision making are not always met. It also shows that managing the interactions—called process management in network theory—is very important for achieving satisfactory outcomes.

All over the world, governments are exploring different types of decision making that considers the increased interdependency of public actors on private, semiprivate, and other public actors. This also enhances the opportunity for citizen involvement in decision making. This trend—in which public actors increasingly use old and new types of citizen involvement in decision making—can be seen in all Western democracies. It occurs under labels such as citizen panels but also under labels such as community governance, open planning procedures, and others (see, e.g., Denters, van Geffen, Huisman, and Klok 2003; Lowndes, Pratchet, and Stoker 2001; McLaverty 2002).

INTERACTIVE DECISION MAKING

In the past few years there has been substantial experimentation with interactive decision making in the Netherlands. Interactive governance is described in this article as a way of conducting policies whereby a government involves its citizens, social organizations, enterprises, and other stakeholders in the early stages of the policy-making process (Edelenbos 1999). The difference with more traditional public policy procedures is that parties are truly involved in the development of policy proposals, whereas in classic opportunities of public comment, citizen and interest group involvement only occurred once the policy proposal had been developed. Interactive decision making is a policy practice. It is an experimental form of decision-making practices mainly at the local level but also in some cases at the central level (Edelenbos 2000; Klijn 2003). As such it is interesting to evaluate this new practice, as is done in this article. We see interactive decision making as a new form of network governance, which we try to evaluate empirically.

Interactive decision making is not without problems. Often, it does not fit the “normal” decision-making procedures, so separate organizational provisions have to be developed in order to conform to these “new” decision-making procedures. Evaluating the connection of this new policy practice with existing decision making and the guidance of this new practice (we call this process management) thus seems important. In this article we evaluate the outcomes and backgrounds of six interactive decision-making processes and their organizational arrangements in the Netherlands. The most important question we want to address is, “What is the influence of organizational arrangements on the outcomes of interactive policy processes?”

OUTLINE

Before we discuss the outcomes of these six processes, we first discuss some of the background of interactive decision making. We also sketch briefly our theoretical framework and network theory and also pay attention to the question of the tension between new governance forms (of which interactive decision making is one) and existing democratic institutions, which can be found in the governance literature. In the sections below, we discuss and assess the impact of three factors that are considered to influence the outcomes of interactive decision making: process design and management of the interactive decision-making process, the degree of participation, and the relation with existing political institutions. Finally, we compare the cases to discover correlations between organizational arrangements and the outcomes of interactive decision-making processes. We end the article with conclusions.

INTERACTIVE DECISION MAKING: AN OVERVIEW

For some time now, interactive decision making has been used in the Netherlands as a new type of horizontal steering for solving problems (Edelenbos 1999; Koppenjan and Klijn 2004). Interactive decision making is regarded as a way of increasing citizen involvement in government, thereby decreasing the perceived cleavage between government and citizen (Nelissen, Godfroij, and de Goede 1996; Tops et al. 1999), but also as a way to cope with interdependencies in complex processes.

Network Theory as Theoretical Framework

Governance and network theories have strongly focused on the changing nature in modern decision making (see Hanf and Scharpf 1978; Kickert, Klijn, and Koppenjan 1997; Marsh and Rhodes 1992; Rhodes 1997; Scharpf 1997). They have stressed that many actors are involved in decision making and that these actors not only posses vital resources to realize policy goals and outcomes but also have different perceptions on the problem definition and have different information and ideas on solutions. So stakeholders' interests often collide in complex decision making; there is much danger that stakeholders block decision making, because decisions are not in line with their interests. Achieving interesting outcomes often depends on finding attractive solutions, which encourage actors to activate their resources and knowledge for the problem and/or policy process at stake. So decision making is also finding ways to manage the complexity of the process, combining necessary actors and decision-making arenas, and creating interesting solutions.

One specific branch of the governance literature is network theory. Basically the network perspective on public policy sees policy as being formed in interactions between actors with their owns perceptions and strategies. These actors are tied to each other by dependency relations (Kickert, Klijn, and Koppenjan 1997; Mandell 2001; Rhodes 1997; Scharpf 1997). So policy formation and outcomes are realized through complex interaction games between actors, which have to be managed to achieve interesting outcomes. These management activities are covered by the concept of network management (Kickert, Klijn, and Koppenjan 1997; Koppenjan and Klijn 2004; Meier and O'Toole 2001). In the literature a wide variety of strategies is mentioned as well as the importance of a process design as a starting point in complex interaction processes (de Bruijn, ten Heuvelhof, and in 't Veld 1998). We take this network perspective as a theoretical framework to direct our questions (the importance of process management and process design) and evaluate outcomes. Rather than dealing extensively with the whole theoretical framework of the network perspective (which has already been done elsewhere; see, e.g., Kickert, Klijn, and Koppenjan 1997), we elaborate some of the assumptions we have derived from network theory in the sections to come.

Thus, we view interactive decision making mainly as a network process, although we are aware that this process can also be positioned in the literature on participation and democracy (Arnstein 1971; Berry, Portney, and Thomson 1993; Hirst 1997; MacPherson 1979; McLaverty 2002; Sorenson and Torfing 2003). We touch on this literature when we come to speak about the relation between citizens and elected officials. However, we keep a more network perspective; we are interested in what roles elected officials play in complex interactive processes, in which citizens, societal groups, and private companies also are actively involved. We do not question the effectiveness of representational democracy as such (see Edelenbos 2005; McLaverty 2002). Moreover, we do not want to go into the institutional tensions among various traditions of democracy (see Edelenbos 2000; Klijn and Koppenjan 2000; MacPherson 1979; Sorenson and Torfing 2003). We are mainly interested in the growing complexity of policy processes, because of the growing number of actors and their interdependencies, and the functioning of the interactive network related to the more traditional representational form in terms of satisfactory outcomes and smooth-running processes.

The ideas on which this article is built heavily rest on the earlier work (and empirical research) of authors on governance and network theory. Before we present the empirical material we first discuss how interactive decision making is supposed to be a solution for some of the problems observed in modern complex decision making.

Interactive Decision Making as Real-Life Solution

With interactive decision making, public actors attempt an alternative way of decision making that should provide a way out of perceived problems encountered in the usual type of decision making. Some problems that are perceived in policy practices are the fact that decision making takes a long time due to the resistance of various involved actors, that solutions are often not inventive enough, or that there is a large gap between politicians and civil servants and citizens. These problems have been discussed extensively in practical discussions and in the literature on governance (see, for instance, Kickert, Klijn, and Koppenjan 1997; Marin and Mayntz 1991; Pollitt 2003; Rhodes 1997; Schön and Rein 1994). Interactive decision making is different from more traditional decision-making procedures. The actual form that the process takes differs basically in the sense that it explicitly tries to involve a wide variety of actors. Interactive decision making is an open decision procedure; it tries to incorporate the values and wishes of various involved actors in the solutions that are developed during the interactive process.

With this new form, interactive decision making tries to provide a solution for a number of existing problems in complex decision-making processes, which are as follows:In time, interactive decision making is expected to result in richer policy proposals that can be implemented more efficiently and thus raise the democratic legitimacy of the decisions.

• The use of veto power. There is substantial veto power in decision-making processes because of the involvement of many actors who typically have the means to influence the outcome of decision making. By involving these actors at an early stage, it is hoped that the use of veto power by the involved actors will decrease and support for decisions will increase. This would accelerate decision-making processes. At any rate, the extra (time) investment necessary for interactive decision making can be “profitable” because it will avert lengthy legal procedures.

• Constantly changing problem formulations. Since problems are the constructions of actors, they have a tendency to change over the course of time as a result of new information, interactions between actors, and external developments. Complex problems are characterized by lengthy decision making. Fixation on a problem formulation early on might mean that a solution is pursued for a problem that appears to be something quite different at the end of the process. By involving more actors in the decision-making process, more and various aspects of the problem can be included in the search for solutions, and problem formulation becomes more flexible. The same argument applies for a premature fixation on solutions.

• Creating “poor solutions.” Go-alone strategies and hierarchical policy processes often lead to poor and one-dimensional solutions, because one rationality or perception dominates in the formulation of the solution, and other perceptions are excluded (Koppenjan and Klijn 2004). Since with interactive decision making not only different perspectives and ideas about problems and solutions are brought in the process but also multiple types of knowledge, information, skill, and experience are employed, a better analysis of the problem area is possible, and better solutions can be created. Thus the overall quality of the final policy is enhanced. Interactive decision making offers the potential to utilize the creativity, experience, and expertise of those involved in order to address issues in a broader, and possibly more innovative, way (Edelenbos 2000, 87).

• Lack of democratic legitimacy. When citizens cannot identify with the policy products of government, the expectation is that they will turn away from government and politics. Numerous problems confronting society, such as indifference to rule enforcement, abuse of collective service, overriding norms, and political nonparticipation, are ascribed to this gap (Klijn and Koppenjan 2000). By involving more actors (and certainly citizens), decision making acquires a less closed character and more democratic legitimacy.

Interactive Decision Making as Organizational Arrangement

Interactive decision making has to be given organizational shape in practice. The form it takes is greatly dependent on the specific situation and context in which these interactive processes are initiated. In this article, we evaluate the influence of some of these organizational arrangements for interactive policy processes. We reflect on the following arrangements for interactive processes:One could argue, however, that not only the arrangements of interactive decision making but also the substance of the process, particularly the degree of value conflict on the substance, matter. Effective interactive decision making depends on how different values and interests are incorporated in decisions. We did not neglect this feature but, rather, took it implicitly into account through the aspects of process design and management (the way the process manager responded to changing situations) and stakeholder participation (the degree to which the various conflicting values and interests are assimilated in a good manner in the selection process).

• the degree of formalization of the interactive process through process design and process management;

• stakeholder participation, especially how the “depth” and “width” are organizationally shaped; and

• the shaping of the relation between the interactive process and the formal position of the municipal council.

The Cases: Six Instances of Interactive Decision Making

While interactive processes are organized for decisions at the national level (Edelenbos and Monnikhof 2001; Klijn 2003), most of the cases can be found at the local level. In this article, we analyze six local interactive policy processes that all concern planning and zoning decisions. Hence, they occurred in more or less the same sectoral regimes. All these cases were studied extensively (sometimes on different occasions and in different research projects), with emphasis on rich description. This article is an attempt to generalize the findings of these cases by focusing on a limited number of variables.

The number of inhabitants varies per city/municipality. The six cases are exemplary for other Dutch interactive processes. Table 1 provides an overview of the cases that were studied for this article.

As mentioned, each of these cases was studied extensively. We closely monitored the behavior and opinions of all participants in the interactive processes. We held semistructured interviews with major stakeholders, civil servants, politicians, and process managers at the start and end of the interactive decision-making process. In these interviews we reconstructed the perceptions of the stakeholders on the interactive process, their views on the outcomes, and how they tried to influence the process. All the way through the interactive process, we also held additional “update” interviews with key persons, such as process managers and civil servants, and examined the course of the process through observation and document analysis. Next we reconstructed the decision-making process and the main issues. All relevant documents in the process (on the organization as well as documents that presented ideas, solutions, or plans) were studied. Subsequently we reconstructed the ideas that were being brought in the process. The data were collected qualitatively.

We first made a reconstruction of the phases of the interactive decision-making process and the important issues and events in the process. Then we made an in-depth analysis of these issues and events and their outcomes on the interactive process. Because we analyze six cases it is difficult to present very detailed case information. It would take simply too much space in this article.¹ We therefore present the case information at a certain aggregation level in various tables.

We use the following five-point scale to score the six cases on the three independent variables, that is, the organizational arrangements:This five-point scale is used for all the indicators designed for the three independent variables. We translated the scoring on the different indicators per variable in a ranking (1 to 6). The various indicators for the three independent variables will be presented in the subsequent sections. In the next section we score the six cases on their outcomes.²

• −−(double minus) very low

• −(minus) low

• +/−(plus/minus) average

• +(plus) high

• ++(double plus) very high

THE OUTCOMES OF INTERACTIVE DECISION MAKING: AN EVALUATION

Evaluating the effects of interactive decision-making processes is not easy. Network theory stresses first that many actors are involved, so the first question that arises is, “Whose objectives will be taken as starting points for the evaluation?” This means that a classic goal evaluation, working with the objectives of a single actor, is not sufficient. Second, interactive decision making involves dynamic processes wherein learning processes occur and objectives change as a consequence of interaction and the exchange of information (see, e.g., Edelenbos 2000; Klijn and Koppenjan 2000). At the very least, an evaluation should attempt to provide an understanding of these dynamics.

Hence, it is more useful to evaluate the six cases in such a manner that adequately considers the multiactor nature of the process and the dynamics of the interactive policy processes. Thus we include the following elements in our evaluation:

• Actor contentment. This criterion concerns whether the parties involved were content with the results of the processes. The advantage is that it involves a weighing of outcomes among different actors and takes the dynamics into account. After all, actors judge whether the outcome meets the objectives developed during the process (Klijn and Teisman 1997; Teisman 1992). The degree to which the outcome of interactive processes is regarded as positive, then, depends on how satisfied the actors are.³

• Enrichment. This criterion explicitly concerns the substance of the process. When we accept the starting point of network theory (Kickert, Klijn, and Koppenjan 1997; Koppenjan and Klijn 2004; Mandell 2001), that is, that information for achieving good policy proposals and policy products is dispersed across many actors and that good policy products are characterized by helping to solve the perceived problems of various actors, the enrichment of variety is an important criterion for the substantive enrichment of the solution (see also Edelenbos and Monnikhof 2001; Teisman 1997). In addition to this variety criterion, we also examine whether the variety of ideas actually emerges in the outcomes (decisions, plans, intentions, etc.). We call this the “impact” criterion (Edelenbos 2000; Edelenbos and Monnikhof 2001).

We speak of “good outcomes” when actors are satisfied and when there is an enrichment of ideas. To assess the last criterion, enrichment, we first looked at the actual outcome. We then traced ideas, solutions, and proposals that had come up in the process and compared them with the initial ideas that were present (mainly formulated in starting documents). The enrichment was large if many different ideas were generated that were not available at the start (variety of ideas) and if we could find many of these proposals in the outcomes of the process (mostly in an end document or explicitly formulated statements and decisions at the end). Actor satisfaction was simply measured by looking at how many of the actors were satisfied at the end of the interactive process.

Table 2 contains the most important conclusions about the outcomes of interactive decision making in the six cases. More detailed information on the coding and scoring can be found in appendix A. On the basis of individual scores, the cases have been ranked in the last column according to their degree of effectiveness. Looking at this table, a few things are striking:Now that we have described the outcome of the six interactive decision-making processes, in the following three sections we consider the organizational arrangement, that is, the process design and management, stakeholder participation, and the relations with democratic institutions.

• There are few cases where the outcomes are unambiguously positive. Apparently, it is not easy to transform the theoretically defined advantages of interactive decision making into real and achieved advantages.

• Leerdam and Doetinchem emerge as the most positive. However, Leerdam is the case where the scope of the interactive process was the smallest. It appears that tight conditions and modest ambitions lead sooner to satisfactory outcomes but also to less substantive innovation and enrichment. This is related to the first conclusion. There is hardly a case where we find a high variety of ideas and a high degree of influence. The Doetinchem case comes closest.

• Most problems are in the impact criterion. This is negative in two cases and average in two others.

PROCESS DESIGN AND MANAGEMENT

In this section we address the role of the process design and process management in the arrangement of local interactive policy processes. Interactive processes are not “self-executive”; a separate person (or group of people) is usually assigned to manage the interactive process. It is emphasized in the network literature that such complex processes can only lead to good and satisfying outcomes when they are intensively supported by process management (Klijn and Koppenjan 2000; Mandell 2001). This should also be based on well-designed organizational arrangements (a process design) for interactions (de Bruijn, ten Heuvelhof, and in 't Veld 1998; Edelenbos 1999). In the Netherlands there are no laws that prescribe certain a priori rules and norms before conducting interactive decision-making processes.

In practice, interactive processes often evolve according to agreements about substance, participation, and rules of the game for the interactive process. These are known as the process design in network theory. Since the process design supports the interaction of the parties, it is of great importance that the participants accept it. Hence, there is no standard design or blueprint for an interactive process. The actual design of the interactive process depends on the specific situational features in which the interactive process has to be carried out. Moreover, the process design is not “self-executive.” It must be developed during the interaction process, applied, and, if necessary, corrected. Together with other activities, this is part of process management (de Bruijn, ten Heuvelhof, and in 't Veld 1998; Edelenbos 2000). In other words, there is constant interplay between process design and process management, all the more so since the environment in which the process unfolds is continuously in flux. Hence, the design is not fixed; rather, it evolves with the process (Koppenjan 2001). Process management fulfils a crucial role in this. On the basis of theoretical insights, we may expect that interactive processes will yield the best results when the design is well organized (hence, a number of rules of the game for time organization, conflict management, responsibility, roles, etc.) and when there is active process management during which the process design is flexibly used and focused on the specific interaction situation.

In order to get an idea of the meaning of process design and process management for the outcome of interactive processes, we examine two elements:

• Formalization of the interactive process: Is the interactive process fixed in a formal document (process design)? What is regulated in it, including time phases, determination of budget, role allocation, manner of conflict resolution, accountability, substantive frameworks, auxiliary conditions, and so on? When the process is fixed in a formal document and many different aspects are regulated, we speak of very high formalization.

• Process management: Did the process manager follow the interactive process strictly according to the agreements and rules of the game in the process design, or did he or she adapt these when necessary to secure a smooth unfolding of the process? How active was the process manager?

Comparing the Interactive Processes

In table 3, we compare the six interactive processes with regard to the elements of process design and process management. More detailed information on the coding and scoring can be found in appendix B. An evaluation of two different scenarios is informative:

In the case of De Bilt the process started with a very detailed process design of the interactive process made by the external process manager. The process design contained elements such as time phasing, descriptions of the roles of participants in the process, policy conditions, participation methods, rules to handle conflict, and so on. The process design had a very detailed character. In the execution of the process the process manager wanted to hold firmly to this design. He did not tolerate any deviations. A striking illustration of this rigid attitude is the reluctance of the mayor to perform a referendum in order to determine how far the people of De Bilt supported the outcomes of the interactive process. The determination of both the process manager and the mayor resulted in a political fight, which had negative effects on the course of the interactive process. This process was delayed for several months.

In the case of Bijlmer the process started with the creation of a project group in the middle of 1995. The project leaders were two people from the project bureau of the central city who had experience with this kind of project. In a way, given the fact that the decision to install the project group was made by the submunicipal county, they were outsiders. Apart from the official decision to start and to redefine the content of the process (which was derived from earlier documents on Bijlmer as a whole), only some ideas were formulated on how to involve tenants (especially the ones who were normally absent in these processes, like the many immigrants who inhabit Bijlmer and the neighborhood of the case, the K-neighborhood, in particular). For this element a participation plan was drafted at the beginning (explicitly accepted by the council at the end of 1995). The submunicipal council was identified as the organization that would assign and control the project leader. But apart from a formal decision to start and the participation plan, no other aspects of a process design were agreed upon, and no documents exist in which these aspects were regulated. In terms of formalization, this clearly made this case a low formalization one (there was only an official starting decision, rough decisions on project leaders, and a participation plan but no decisions on all the other aspects).

This comparison demonstrates that in all six cases there was a formalization of the interactive process through a process design, and only the degree of formalization varied greatly. The interactive process in De Bilt operated on a very detailed process design, where many issues were formally fixed, such as role allocation, final responsibility, time phases, work forms, process organization, rules of interaction, and so forth. On the other hand, the interactive processes in Enschede, Leimuiden, and Bijlmer worked with a rudimentary process design that only regulated issues at a very general level. The Leerdam and Doetinchem cases occupy the middle ground.

We see variation in the implementation of the process design. Although all the process managers in the cases were very active, which seems logical given the experimental nature of the decision-making processes, the way they operated was not the same. Thus, in the case of De Bilt, the process manager rigidly held on to the process design, even when circumstances in the interactive process called for an adaptation of it. This style can be characterized as “blueprint process management.” In the cases of Enschede and Bijlmer, we see that a rudimentary process design was “compensated” with a more flexible and active implementation, resulting in a style we call “improvised process management,” since deviations from the design often occurred during implementation because of intermediate developments in the interactive process. Leimuiden, like Enschede and Bijlmer, had a rudimentary process design but also an active process manager rigidly holding on to the main outlines of the design (time phasing, role allocation, etc.).

We qualify “adaptive process management” (Leerdam and Doetinchem) as good management, because there is a reasonably detailed process design that evolves with the developments in the interactive process. Improvised process management (Enschede and Bijlmer) is qualified as reasonable; although there is a rudimental process design before the start of the interactive process, this is compensated through adequate and creative actions from the process manager. We qualify blueprint process management as moderate; there is a thought-out process design, but the process manager follows this design too rigidly during the execution of the interaction process. The process manager ignores meaningful new developments in the interactive process, which has negative effects on the course of the interactive process (see the De Bilt example above for illustration). Process management on main outlines (Leimuiden) is qualified as bad process management; this is when process design is rudimental and process management style is inflexible.

STAKEHOLDER PARTICIPATION

In this section, we discuss stakeholder participation in the interactive process. Stakeholders include societal organizations, private parties, and organized and nonorganized citizens.

The Depth and Width of Participation

In order to assess whether the participation structure of an interactive policy process results in more meaningful participation, we consider two dimensions of participation. Inspired by Dahl's “preconditions for a polyarchy,” Berry, Portney, and Thomson (1993, 55) have formulated two dimensions of participation that are important for a system of strong participation. These are the width and depth of participation, which together determine the strength of participation in the policy process (see also Wille 2001). The width of participation is the degree to which each member of a community is offered the chance to participate in each phase of the interactive process. The depth of participation is determined by the degree to which citizens have the opportunity to determine the final outcome of the interactive process. In the analysis of participation width and depth, it is important to distinguish the process, on the one hand, and the final outcomes of that process, on the other. In this section, we only consider the process itself.

Citizens usually become active when invited to participate: hence it is largely mobilized behavior. This is also the starting point of various types of interactive policy development. In the analysis of width of participation, we consider how municipalities have shaped this “invitation” policy. In short, what opportunities for participation have been made available? Did citizens frequently receive information about how they could participate? Was participation accessible to all?

An evaluation of the width of participation during the interactive process is focused on the articulation of interests. The analysis of the depth of participation in the outcome is focused on the degree and type of influence citizens have had in shaping opinions and the realization of outcomes. In order to map the influence of participation, participation ladders are frequently used (e.g., Arnstein 1971, 71–78). To determine the depth of participation, the participation ladder outline below is used (Edelenbos 2000, 43–44):

• Informing: To a large degree, politicians and administration determine the agenda for decision making and inform those involved. They will not use the opportunity to invite interested actors to have input in policy development.

• Consulting: To a large degree, politicians and administration determine the agenda but regard those involved as a useful discussion partner in the development of policy. Politicians do not, however, commit to the results of these discussions.

• Advising: In principle politicians and administration determine the agenda but give those involved the opportunity to raise problems and formulate solutions. These involved actors play a full-fledged role in the development of policy. Politicians are committed to the results in principle but may deviate (if accounted for) from them in the final decision making.

• Coproducing: Together politicians, administration, and those involved determine a problem-solving agenda in which they search for solutions together. Politicians are committed to these solutions with regard to the final decision making, after having tested this outcome in terms of a priori conditions.

• Co-deciding: Politicians and administration leave the development and decision making of policy to those involved, and the civil service provides an advising role. Politicians simply accept the outcomes. The results of the process have an immediate binding force.

These levels are organized in such a way that when the input and involvement of citizens increase, the influence and role of government decrease. At lower levels (consulting and advising), the citizen is regarded as a supplier of ideas, mobilized by local government, which wants ideas about specific policies. A higher degree of interaction occurs when citizens help determine the agenda in a particular policy area and cooperate in producing problem definitions and solutions, though the final decision rests with local government (coproduction). Finally, together with the government, citizens can decide about plans made in cooperation (co-deciding). The different modes of participation in width and depth lead to different types of interorganizational structures (see Mandell and Steelman 2003).

From the motives for interactive decision making, as discussed above, it is expected that more intensive involvement of participants, in terms of both width and depth, must lead to substantively richer policy proposals. Logically, these are linked to a larger degree of satisfaction among actors with the outcomes. Probably the width of participation is strongly linked to the variety of the outcomes, while the depth of participation is linked more to the satisfaction of the outcomes and (logically) to the influence.

Comparing the Interactive Processes

In table 4, the six interactive policy processes are compared with regard to stakeholder participation. More detailed information on the coding and scoring can be found in appendix C. Two examples are illustrative:

In the case of Leerdam a small working group was formed in which, in total, fourteen representatives of organized interest groups participated (entrepreneurs, nongovernmental organizations, etc.). Although this is a rather small amount of participation, it made it possible to realize a reasonably “deep participation.” Civil servants and participants worked in coproduction toward alternatives for the realization of the renovation of the city square. Ideas for the renovation were developed in extensive and time-consuming design teams and working sessions.

In the case of De Bilt there was very wide participation. Every citizen had an opportunity to join the interactive process. Through open invitations and direct mailing, stakeholders were mobilized. Over 200 participants contributed actively in several interactive methods such as workshops. However, their participation was not deep. They had the opportunity to raise ideas, but the selection of these ideas was mainly done by civil servants and was communicated to the mayor and aldermen and not to the stakeholders.

When the cases are compared, we see that there was generally fairly broad participation. Only in the Leerdam case was there limited participation. As far as depth of participation is concerned, most cases involved lighter types of participation. Advising and consulting dominate (four cases), while in only two cases do we see a somewhat heavier form (coproduction). In characterizing the strength of participation (Berry, Portney, and Thomson 1993), we see that only the Doetinchem case experienced this. Weak participation was characteristic for the De Bilt, Enschede, Leimuiden, and Bijlmer cases. The Leerdam case is difficult to characterize since there was reasonably influential participation but from only a few participants.

RELATION WITH THE MUNICIPAL COUNCIL

In this section, we discuss the relation between the interactive processes and the existing democratic institutions at the local level, more particularly analyzing the relation of the cases to the municipal councils involved.

Coordination of Interactive Process with the Political Environment

Relations between interactive processes and the existing political-administrative policy world are not without problems. There is a risk that the interactive processes will become uncoupled from the “normal” decision-making procedures, as is clear from various reflections about interactive decision making (Edelenbos 2000; Koppenjan 2001). A lack of commitment from political officeholders in the normal policy arenas may lead to the emergence of parallel policy-making trajectories: the interactive and the traditional process. Thus, the first question is whether political officeholders have been informed and consulted about the initiative of starting an interactive process. Have they played a role in confirming the process design for the interactive process? These two formal indicators for political involvement are the first to be compared in the cases. We label them with the terms initiation (Who initiated the interactive process?) and confirmation (Was the initiative for an interactive process solidified in a formal decision by the municipal council?).

Next, organizing feedback to the municipal council is important. Lacking coordination and feedback between the interactive process and the normal policy- and decision-making arenas may result in “hard linkages” at the end of the interactive process: traditional decision-making processes and interactive processes bump into each other. Decision makers in the “traditional” decision-making arenas are unaware of or uninvolved in the interactive process. They are surprised by the outcomes and experience these as bothersome. Since they lack commitment to the interactive process, they do not take it into account. Political officeholders ought to be “taken along” in the interactive policy-making learning process and become familiar with the arguments and ideas. This may result in “soft linkages”: although political officeholders make their own assessments, they can use the insights from the interactive process. This requires constant feedback between the interactive process and the governing bodies involved. Here we examine whether during the interactive process, formal (through regular procedures) and/or informal (ad hoc through the interactive process) feedback to the municipal council occurred.

Roles of Politicians: True Participation of the Council

Interactive decision making is a type of direct democracy, which is applied in the game of representative democracy (Klijn and Koppenjan 2000; Lowndes, Pratchett, and Stoker 2001; McLaverty 2002). This involves a role conflict for political and administrative officeholders because decisions taken by direct participation possess a legitimacy of their own that may challenge the legitimacy of the representational decision-making channel. As a result, politicians are sometimes disinclined to participate in interactive processes because they do not want their hands to be tied at the end of the process and thereby be prevented from living up to the mandate given by the electorate. On the other hand, early involvement of these actors may “kill” the process: there must be something that other parties can bring forward. Keeping political officeholders out of the interactive process raises the chances of a hard linkage at the end. One must search for a kind of coordination between political officeholders and the interactive process that gives proper consideration to the position of both (Klijn and Koppenjan 2000).

To assess this, we look at the frequency with which council members participated in the interactive process. The idea is that the more they participate in interactive sessions, the better able they are to assess the outcomes of the interaction. We use a threefold division in determining the participation of council members: always to often present, present now and then, and once to never present. Next, we consider the role that council members played if they participated in the interactive process. We distinguish among three types of roles, going from passive to active participation: passive auditing/information collection, questioning participants/providing information, and active participation.

On the one end of the spectrum is the role of auditor. During the interactive process, these council members do not actively engage in discussion and negotiation with each other or with other participants but, rather, observe these processes. They do not participate in the discussion and in designing policy, even when participants explicitly request their opinion or perspective.

In the middle of the spectrum is the role of information provider, which includes both passive and active aspects. In the passive element, for instance, prior to the process these council members and civil servants provide information in the form of auxiliary conditions, data from reports, memos, and results from research. The active part involves providing information during the process, through presentations and/or brief (informative) answers to questions from participants.

At the other end of the spectrum, we find the role of participant, the most active role. These council members participate in the process in order to provide substantive input from their own perspective, interest, and value. They actively engage with other participants in the interactive policy process through discussion and negotiation in order to arrive at informed opinions about problems and solutions.

Comparing the Interactive Processes

In table 5, the six interactive policy processes are compared with regard to the relation between interactive process and city council. More detailed information on the coding and scoring can be found in appendix D. For examples, note the following:

In the Doetinchem case, the relation between the city council and the interactive process was very tight. Councilors were actively involved in the start of the interactive process and in the determination of their role during the process. During the process they were kept up to date extensively, formally in the meetings of the council and informally through attendance of the interactive workshop sessions. The councilors also took the role of debater in the interactive process; they not only listened carefully to the debates between citizens but were not afraid to join the debate. This attitude lead to very active council involvement in the interactive process.

Table 5

Overview of the Relation between Interactive Process and Council in the Six Cases

Case .	Role Before the Start of the Process .	Role during the Process .	Characterization .
De Bilt	no involvement of the council at the start of the process (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−−)
Enschede	council confirmed the start of the interactive process (+/−)	on occasion informal involvement of the council (+/−)	rather limited council involvement (+/−)
Leerdam	council was not involved at the start of the process; it ritually approved it, after it had already started (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−)
Leimuiden	council was informed after the idea of starting an interactive process and approved the idea (+)	no involvement during the process (−−)	limited council involvement (−)
Doetinchem	council was informed after the idea of starting an interactive process and approved the idea (+)	active involvement of councilors during the process through feedback and in their role as debaters (++)	very active council involvement (++)
Bijlmer	council initiated the idea of the interactive process (++)	no involvement during the process (−−)	rather limited council involvement (+/−)

Case .	Role Before the Start of the Process .	Role during the Process .	Characterization .
De Bilt	no involvement of the council at the start of the process (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−−)
Enschede	council confirmed the start of the interactive process (+/−)	on occasion informal involvement of the council (+/−)	rather limited council involvement (+/−)
Leerdam	council was not involved at the start of the process; it ritually approved it, after it had already started (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−)
Leimuiden	council was informed after the idea of starting an interactive process and approved the idea (+)	no involvement during the process (−−)	limited council involvement (−)
Doetinchem	council was informed after the idea of starting an interactive process and approved the idea (+)	active involvement of councilors during the process through feedback and in their role as debaters (++)	very active council involvement (++)
Bijlmer	council initiated the idea of the interactive process (++)	no involvement during the process (−−)	rather limited council involvement (+/−)

Open in new tab

Table 5

Overview of the Relation between Interactive Process and Council in the Six Cases

Case .	Role Before the Start of the Process .	Role during the Process .	Characterization .
De Bilt	no involvement of the council at the start of the process (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−−)
Enschede	council confirmed the start of the interactive process (+/−)	on occasion informal involvement of the council (+/−)	rather limited council involvement (+/−)
Leerdam	council was not involved at the start of the process; it ritually approved it, after it had already started (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−)
Leimuiden	council was informed after the idea of starting an interactive process and approved the idea (+)	no involvement during the process (−−)	limited council involvement (−)
Doetinchem	council was informed after the idea of starting an interactive process and approved the idea (+)	active involvement of councilors during the process through feedback and in their role as debaters (++)	very active council involvement (++)
Bijlmer	council initiated the idea of the interactive process (++)	no involvement during the process (−−)	rather limited council involvement (+/−)

Case .	Role Before the Start of the Process .	Role during the Process .	Characterization .
De Bilt	no involvement of the council at the start of the process (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−−)
Enschede	council confirmed the start of the interactive process (+/−)	on occasion informal involvement of the council (+/−)	rather limited council involvement (+/−)
Leerdam	council was not involved at the start of the process; it ritually approved it, after it had already started (−−)	sporadic involvement of the council during the process (−)	very limited council involvement (−)
Leimuiden	council was informed after the idea of starting an interactive process and approved the idea (+)	no involvement during the process (−−)	limited council involvement (−)
Doetinchem	council was informed after the idea of starting an interactive process and approved the idea (+)	active involvement of councilors during the process through feedback and in their role as debaters (++)	very active council involvement (++)
Bijlmer	council initiated the idea of the interactive process (++)	no involvement during the process (−−)	rather limited council involvement (+/−)

Open in new tab

In the case of De Bilt the councilors were not informed at the beginning of the interactive process. The process was not even politically approved by the council. The involvement of the council was also very limited during the interactive process. Councilors sporadically joined the interactive process, and when they did, they took a very passive role as auditors and information collectors.

In comparing these six interactive processes on this aspect, it becomes clear that one case jumps out positively: only the interactive process in the municipality of Doetinchem had both formal (initiation and confirmation) and actual (feedback and council member participation) close involvement of the municipal council in the interactive process. The Bijlmer case shows a situation where the (neighborhood) council was formally involved but hardly at all in practical terms. The other four cases display limited to very limited involvement of the municipal council with the interactive process. In the cases of De Bilt and Leerdam, the limited involvement of the municipal council is, of course, striking. After all, they did involve experiments that explicitly aimed at strengthening the relation between citizens and politics.

ORGANIZATIONAL ARRANGEMENTS AND OUTCOMES

Table 6 presents a comparison of the analyses of the previous sections to each other. We sum up the scores for the three organizational characteristics, process management, stakeholder participation, and relation to municipal council. This also holds for the scores for the outcomes of the six cases. The last column also provides the ranking of the six cases. When two cases have (almost) the same score, they have, in principle, been given the same ranking.

In some cases, the large difference between cases is also taken into account. Thus, for all cases there was medium to weak involvement of the council. The only exception is the Doetinchem case, and this is expressed by giving it the ranking 1 and giving the two following cases, which have a much lower score, a ranking of 3. In the ranking for stakeholder participation, equal scores for the cases resulted in emphasis upon depth of participation to determine the ranking.

Process Management and Outcomes: Adaptive Process Management Enhances Good Outcomes

Looking at table 6, we can see a clear link between a positive score for the process management aspect and the score for outcome. The two cases where process management was assessed positively, and where it may be labeled as adaptive process management (Doetinchem and Leerdam), also score the best when outcomes are compared. Interestingly, the distance between the cases of Doetinchem and Leerdam, on the one the hand, and the other cases, on the other, is large when considering process management, and this is reflected in the outcomes. In other words, cases with adaptive process management have good outcomes, while the other cases display a weak or even negative score for both process management and outcomes. Adaptive process management leads to outcomes that are supported and enriched by stakeholders. Hence, there is a strong correlation between the scores for process management and those for outcomes. This is probably the most interesting finding of this research on the six cases.

Interactive Decision Making and Stakeholder Participation

The good position of Doetinchem is again striking when looking at the relation between stakeholder participation and outcomes. Also striking is that De Bilt occupies a relatively high position, while in terms of outcomes it is much lower. This is because the variety of ideas had limited influence on the end results. In the case of broad stakeholder participation (Doetinchem, De Bilt, and Bijlmer), there was substantial variety if actors were brought into the interactive process. The assumption in the literature that an increase in participation results in more variety and, in principle, richer plans appears to be supported. It is crucial, however, that this variety is also assimilated in a good manner in the selection process. Thus, the variety in Leerdam was not so great, but an outcome that was good for all parties was achieved since the ideas put forward also influenced decision making and end results. This requires adaptive process management. It appears that broad stakeholder participation is an important but not necessary condition for a good outcome. The depth of participation is more important for a positive assessment about the outcome of interactive decision making.

Interactive Decision Making and the Municipal Council: A Problematic Item

What is most striking about table 6 is the involvement of the municipal council in interactive processes in almost all of the cases. This indicator is only strong in the Doetinchem case, where, from the start, there was substantial attention by the process manager to involving council members in the interactive process. There was also a positive attitude among most council members about involvement in the interactive process. However, this also makes it more difficult to draw clear conclusions about the relation between outcomes and the degree to which interactive decision making was embedded in the normal political decision making. The low score for Leerdam (last) is striking, while its score for outcomes is good. Apparently, it is possible to compensate for a limited relationship with the municipal council with good process management. We also need to take into account the fact that the council in the one municipality is more prominently and forcefully involved in local politics than that in another municipality. Good organizational structuring of the relation between the interactive process and the municipal council is important when the council has a strong influence upon municipal affairs. When the council is less strong (it can be that the mayor and aldermen overrule the council), organizational structuring may be less important. In the short run, not involving the council in interactive processes may have limited consequence since an alderman can carry the outcomes of the interactive process through the council. In the longer run, however, there is potential danger. The council may become irritated and may decide to block the outcomes of the interactive process that once appeared set in stone. Nonetheless, the conclusion that the relation to the council is less important than previously thought is striking, and this conclusion also contradicts findings about interactive processes at the national level. One explanation could be that the relation between politics and the interactive process is of greater importance at the national level in the Netherlands, since national political officeholders can develop more counterweight to administrators than their counterparts can at the local level.

Another explanation can be that one indicator is more relevant than the other. When we look at the indicators (see table 5) we can see that the “feedback” indicator scores positively for the cases of Doetinchem and Leerdam. These are exactly the cases that show good outcomes. This finding corresponds with earlier research on this topic (Edelenbos 2005) but still needs further attention in future research.

Compound Lenses: The Importance of Process Management

When we consider all three dimensions of the organizational arrangement of interactive decision making, process management comes across as the most important condition. This score is most similar to the scores for outcomes. Furthermore, there are no deviations (such as high scores for process management and low scores for outcomes or vice versa) that sometimes occur with other organizational characteristics of interactive processes. In short, low performance on one of the other organizational features can be compensated for (as, for instance, in the Leerdam case), but a low score for process management cannot be made up. This confirms the opinion often stated in network literature that process management is of paramount importance to complex interactions.

CONCLUSION: THE IMPORTANCE OF GOOD PROCESS MANAGEMENT IN INTERACTIVE DECISION MAKING

In this article we considered the organizational arrangements of interactive decision-making processes. We focused on three characteristics: the formal organization of process management and the practical use of it, the degree of involvement of societal actors, and the relation of the process to normal political decision making (i.e., the relation to the municipal council).

The most important conclusions are as follows:Placed in the discussion on participation and governance these findings make an interesting contribution. First, the outcomes seem to stress that participation is strongly appreciated by stakeholders if they see real outcomes from this participation. On the basis of our material we are even inclined to say that one can better afford no participation at all than bad participation that is not well managed and in which voiced preferences are neglected. This may be reason to draw the research, even more than is already the case, out of the normative discussion that participation is good in itself and focus on the way it is achieved in a really satisfactory and efficient way.

• Greater input from a variety of parties generates a variety of ideas and potentially enriches process substance.

• Greater input does not guarantee good outcomes. The Leerdam case demonstrates that good outcomes can be realized with less variety, and the Bijlmer and De Bilt cases demonstrate that large variety does not guarantee good outcomes. In Leerdam, the variety was not great, but this was compensated for with good influence and process management.

• Process management emerges as the most important condition for good and satisfactory outcomes. There is a high correlation in the six cases between good process management and good outcomes.

• It is difficult to find a link between outcomes and the degree to which the municipal council was involved in the interactive process since in most cases that involvement was not substantial. The Leerdam case, which combines low council involvement with good outcome, leads us to conclude that council involvement is not unimportant and can, in fact, be an obstacle (see Edelenbos 2000; Klijn and Koppenjan 2000), but it is not a decisive factor for a negative outcome.

The relation between these new forms of decision making and the elected officials in city councils still remains ambiguous and is certainly something that requires more research. Our findings on these six case studies do, however, provide a good impression of the importance of good process management for the success of interactive processes. Management matters in the successful evolution of interactive decision-making processes. This is in general also stressed in the literature on governance and network management (Agranov and Mcguire 2001, 2003; Gage and Mandell 1990; Mandell 2001; O'Toole 1988). Our addition to the existing literature on network management is that we have distinguished different styles of network management and assessed which styles are more appropriate for using in the guidance of complex interactive decision-making processes. Initiators of interactive decision-making processes must adopt an adaptive style of network management in order to be successful in the end. If initiators of these forms of governance lack the organizational flexibility and creativity to manage these and there are no other actors who are prepared and willing to fulfill the role, maybe they should simply refrain from action (see also Koppenjan and Klijn 2004, 252). This is of course a controversial statement, because what should we do if there is an urgent problem that needs solving? We think that realism is still needed, and in such a case we should work on preconditions before acting. In the long run, badly managed projects and disappointed stakeholders are worse than rhetorical actions.

APPENDIX A: OUTCOMES

We describe the outcomes that were realized at the end of the interactive process and the actors who were satisfied and dissatisfied with the outcomes in table A1. We present the enrichment of the outcomes in tables A2–A3. We conceptualized enrichment as the variety of ideas and the influence of ideas. We used “variety of ideas on problems” and “variety of ideas on solutions” as indicators. In order to determine the influence of ideas on decision making we used the indicators “influence during the development of the plans” and “influence recognizable in the final documents of the interactive process.” The two variables “actor satisfaction” and “enrichment” determine the quality of the outcome. We speak of “good outcomes” when actors were satisfied and when there was an enrichment of ideas.

APPENDIX B: PROCESS MANAGEMENT

Table B1 indicates the presence of a process design and the detail level of this design. These two indicators determine the formalization level of the process design. Table B2 indicates the dominance of the process manager in the interactive process through his or her activities and the flexibility of the process manager in executing the process according to the process design. These variables determine the style of process management in the six cases. In the last case these indicators are used to create a typology of management (active very rigid [−−], passive rigid [−], passive flexible [+−], active flexible [+], active very flexible [++]). The degree of flexibility thus determines the positive or negative nature of the score—see the case of De Bilt: very rigid and active process management has a score of −−, which is composed of active (++) and hardly any flexibility (−−). The scores of the two tables together determine the characterization in table 3 in the main text.

APPENDIX C: STAKEHOLDER PARTICIPATION

Table C1 states who were invited to participate and what the actual participation was in the interactive processes. These two indicators determine the width of participation. Table C2 indicates who set the agenda, brought in ideas during the process, and made the final decision. (Note, this is not the same as influence of ideas [see indicators for outcomes], but one would expect a relation, although with a lot of conflicting actors who can all set the agenda and so forth, the influence can still be minor.) These three indicators determine the depth of participation in the six cases.

APPENDIX D: RELATION WITH MUNICIPAL COUNCIL

In table D1 we used four indicators: (1) who initiated the interactive process, (2) was the process confirmed by the council before it started, (3) was the council kept up to date of the progress of the process, and (4) did the council members participate in the interactive process? Together these variables determine the way the council was related to the interactive process.

References