1. Background

Public participation and stakeholder involvement have long been considered central in policy and planning processes (Reed 2008, Lamers et al. 2010, Akhmouch and Clavreul 2016). Simpson and Löe (2020) argue that for complex environmental problems, such as groundwater protection, involvement of expert scientist are not enough, but that affected communities are essential for enabling lasting solution that also pay respect to local knowledge, beliefs and values. Ideally, by involving a broad set of stakeholders one enlarges the knowledge base of the processes, increasing the ownership to and legitimacy of the outcomes (Lang et al. 2012).

Participation has for these reasons increasingly also become a prerequisite for decision-making processes, and a requirement of integrated and adaptive governance arrangements (Reed 2008, Lamers et al. 2010, Akhmouch and Clavreul 2016). The European Water Framework Directive is a case in point, where the inclusion of interested parties in decision-making processes is a central tenet of river basin planning (WFD 2006), although the exact form of participation required is not given (Newig et al. 2018). The promotion of stakeholder engagement is also at the core of the OECD water governance principles (OECD 2015). Despite the general agreement on the importance of participation, it is still disputed to what extent participation is a necessary requirement to solve environmental problems (Newig et al. 2018). Hence, the exact ways in which engagement processes can contribute to sound environmental outputs and water governance processes, begs further attention (Koontz and Thomas 2006, Young et al. 2013, Drazkiewicz et al. 2015, Scott 2015, Akhmouch and Clavreul 2016).

In our attempt to address the agriculture-drinking water nexus within a multi-actor context, a useful point of departure is the existing and vast literature on multi-stakeholder approaches and platforms (Steins and Edwards 1996, Warner 2006, Reed 2008, Fish et al. 2010, Heinelt 2012, Lang et al. 2012, Graversgaard et al. 2018, Kochskämper et al. 2018).This literature points out the promises of participatory approaches, but also some of the pitfalls and limitations.

While optimal representation would require every affected party to be included, this is most often not the case; either because it is practically impossible or financially not viable. There might also be other reasons for why some parties are not invited to participate, or are not able or willing to take part (Lamers et al. 2010, Warner and Verhallen 2012). Also, once an arena for engagement is formed, although the participants might take part on an equal footing, power dynamics are at play that might work to the detriment of the quality of the process (Reed et al. 2018). Varady et al. (2016) highlight stakeholder endurance as a critical factor, that it might be difficult to sustain interest and participation over time; either because it is difficult to commit timewise or financially to enduring processes, or that stakeholder fatigue might be the outcome when goals are unclear or immediate benefits hard to reach.

Building on this literature on stakeholder engagement, we also suggest a shift in perspective, conceiving of participants in engagement processes as being actors, rather than stakeholders. To look at participants as more than just “having a stake” or representing an interest, holds a potential to move beyond agency as motivated by economic interest only, and opens up to enable more complex understandings of interests and standpoints, that might also include values and perceptions (cf. Braun 2009). Such a shift also entails a move away from focusing on finding the appropriate tool to facilitate engagement processes, but instead emphasise participation as a process (Reed 2008), and acknowledging that the process can be as important as the outcome (Young et al. 2013, Graversgaard 2018).

Involvement of multiple actors in participatory processes can be justified with reference to two different categories of benefits (Reed 2008, Graversgaard 2018). Firstly, there are functional or pragmatic reasons for participation, where the aim is to improve decisions and environmental performance, in which participation is a means to an end. Secondly, there are normative reasons, focussing on participation as a democratic right and that broad involvement ensures representation, transparency, and legitimacy of processes.

While engagement processes often concern obtaining social licence for particular interventions, there are also important social learning-aspects of engagement processes. Social learning can here be understood as the ways stakeholders “acquire (rather than just convey) knowledge and collective skills through better understanding of their situation as well as the perceptions, concerns and interests of [others]” (Wehn et al. 2018:37). Lacroix and Megdal argue that “while engagement as a learning process may not solve [conflicts] per se, it is considered very important to overcome persistent norms and ‘difficult-to-change socio-technical systems’ ”(2016:2). This also further emphasizes the process dimension of engagement. Hence, actors have an interest, but also a potential and an ability to take part in the co-production of knowledge and co-creation of solutions to complex dilemmas of governance and management in these sectors (Graversgaard et al. 2017).

In a project setting, a multi-actor approach is devised to ensure meaningful involvement, with real impacts on the research process and outcomes through co-creation of knowledge and solutions (cf. Ostrom 2010). Such engagement should take place as early as possible in the project cycle; from the planning of work and experiments, their execution and implementation up until the dissemination of results, and evaluation (Reed 2008). This will facilitate joint knowledge production and interactions between a range of actors, including end-users, in ways that will lead to shared ownership to both process and results (Levidow and Neubauer 2014, Belmans et al. 2018, Graversgaard et al. 2018).

Steins and Edwards consider an engagement platform to be “a negotiating and/or decision-making body (voluntary or statutory), comprising different stakeholders who perceive the same resource management problem, realize their interdependence in solving it, and come together to agree on action strategies for solving the problem” (1996:244). In line with the discussion of multi-actor approaches, we find that a more non-linear understanding of these platforms is necessary, to acknowledge the dynamic character of engagement processes. Hence, we will in our analyses of multi-actor platforms relate to Acquaye-Baddoo et al. (2010:4) as they define such platforms as “a more-or-less ongoing mechanism in which actors meet regularly to foster exchange and promote joint decision making and collaboration in a continuously evolving way”. We will however also engage with other frameworks and dimensions for multi-actor engagement.

2. FAIRWAY’S multi-actor platforms

The FAIRWAY project works through 13 cases in 11 EU member States. These cases are included in the project as multi-actor platforms (MAPs). These MAPs are either engagement platforms that have a longer history and have then been brought in under the project to contribute to FAIRWAY’s aims, or they have been set up within the project start. See table below for an overview of the cases and their history.

Table 1. Overview of multi-actor platforms in the FAIRWAY project, indicating history of engagement

Whether the MAPs of the FAIRWAY project are new or have a longer history is only one feature of difference between the cases of the project. While some address quality of drinking water as surface water, others concern groundwater. While some MAPs address issues pertaining to nitrates and/or phosphorus, others deal with pesticides; while yet others engage with all these issues. In some cases, there is a high level of conflicts, in others the tensions are less visible, or absent. In some cases, the platform functions with an official and formal mandate, in other cases it is more of a loose association around more or less common challenges or problems. The platforms also vary according to the kind of actors they involve; for instance, on whether and how farmers are involved.

Warner and Verhallen (2012) argue that because of the complexity and the variety of social, cultural and political contexts, there are no normative definitions of multi-stakeholder or multi-actor platforms in terms of structure or methodology used. According to Reed, “different levels of engagement are likely to be appropriate in different contexts, depending on the objectives of the work and the capacity for stakeholders to influence outcomes” (Reed 2008:2419). Hence, defining a standard or ideal engagement platform is not desirable, and it has been difficult to devise one common strategy for setting up engagement platforms within the FAIRWAY project. It also means that there are challenges in comparing across the cases; as anything from the respective MAPs’ objectives, their structure, history and context varies a lot. In the following we will present a framework for analysing engagement process that we believe is helpful, regardless of the mentioned challenges and differences among them.

3. Dimensions for multi-actor engagement

The vast literature on stakeholder participation, from the seminal work of Arnstein (1969) and her “ladder of participation”, to recent work by Reed et al. (2018), indicates an array of typologies relevant for multi-actor engagement. Warner and Verhallen (2012) sketch a typology of multi-stakeholder platforms with regards to dimensions that are considered relevant for assessing strengths, weaknesses and opportunities for change with regards to engagement platforms. These dimensions relate in different ways to the process, the content and the context of engagement processes (Figure 1).

Figure 1

Dialogues Inspired by Warner and Verhallen (Warner and Verhallen 2012), we have developed a typology fit for the multi-actor platforms of the FAIRWAY project. Some adjustments have been made from this initial typology in labelling and descriptions. Moreover, Warner and Verhallen suggest that a movement of any given engagement platform from the left to the right in their typology implies a move towards a more effective multi-stakeholder dialogue. Given the dynamic character of multi-actor process, and the differences between the MAPs of the FAIRWAY project we found that this is an assumption that is difficult to set up-front. Our suggested framework is therefore more open-ended in terms of the ideal conditions for engagement. The issue will be returned to in »Analysis and discussion.

Our framework of dimensions relevant for meaningful engagement processes (Figure 2) is presented in the following:

Figure 2

• Arenas involve what kind of actors are involved from which sectors and at which levels.
• Adaptivity refers to the capacity of the platform to adapting to changing external circumstance, as well as adapting to the dynamics of the organization, and being flexible to change direction and goals depending on identified needs with the platform.
• The synergies of a platform ranges from a focus on each stakeholder’s own interests (with no synergies), to a more joint effort to find solutions and innovations that bridges these different interests, as well as bringing in additional social learning outcomes.
• Shared goals refer to whether there is a shared understanding of the urgency and the nature of the problem, and consequently whether there is a goal that is shared by all parties and that everyone can rally behind and work towards.
• Power balance is here understood as whether within the engagement platform there are one or more actors that dominate discussions, decision-making and agenda-setting, or whether there is a more level playing field.
• Decision space refers to the kinds of mandate and legitimacy a platform has, ranging from a small/narrow mandate, e.g. as a consultative body, or a broader mandate when influencing decision-making processes. This could refer to both internal mandate (constituency to representatives) and external mandate (enabling environment) (Warner and Verhallen 2012).
• Available resources refer to the extent to which the platform is seen as having resources with regards to institutional support, funding and manpower, whether the structure set up with the voluntary contribution from all actor groups, or the platforms take the form of a network with a staffed secretariat or the like. Available internal resources may refer to different kinds of public, legal or financial support, while external resources on the other hand may be outside legitimacy of the platform with regards to the problem at hand.
• Trust refers in this context to a broad understanding of trust between actors, both encompassing relational trust (between oneself and the other) and calculative trust (relating to perceptions of past behaviour of the other and/or on constraints on future behaviour) (cf. Earle 2010).
• Outcomes can be understood as both tangible outcomes as achievements in terms of environmental improvements and reaching set targets, but can also be seen as process outcomes, such as the building of legitimacy through improved relationships and increased understanding of other actors’ positions and perspectives.

Note: For full references to papers quoted in this article see

»References