Assessing participatory process-system linkages in polycentric water governance: Insights from WFD implementation in Germany

INTRODUCTION

A common, but so far insufficiently answered, research question concerns how participatory approaches for governance improve policy implementation. “(P)articipation can have a twofold impact on effective policy delivery. The first is that it influences the decision itself (otherwise it would not be participation). (…) The second is that participation can lead to a more complete implementation of decisions” (Fritsch & Newig, 2007, p. 4). This definition reflects a common, but narrow, understanding of participation. One criticism of the literature is that accounts of public participation are predominantly normative and virtue-laden instead of assessing the actual mechanics of participation as they are practised (Heijden & Heuvelhof, 2012). Such mechanics are, in part, related to the governance systems in which participatory processes are embedded in. While participation is expected to fit well with the philosophy of polycentricity, its actual implementation is assumed to be easier in monocentric systems because a single actor could interact with and provide feedback to the public (Huitema et al., 2009).

However, at present, little is known about the interactions among implementation structures, the governance system, and different instrument types (Steinebach, 2022) both generally and specifically with regards to participatory processes. Over time, governance systems tend to become increasingly complex because a multiplicity of diverse actors pursue an increasing number of aims (Adam et al., 2019) as a consequence of changes to environmental regulations and economic development. All of this complexity and change results in various place- and context-specific decision-making structures in terms of the actor types and decisions to be taken. Significant system-related research gaps persist concerning participation and scale dependency (Stringer et al., 2006), scale and level of collaboration (Margerum & Robinson, 2016), multi-level governance (Newig & Fritsch, 2009; Niles & Lubell, 2012), decision-making context (Speer, 2012), the nature of linkages among processes in polycentric governance systems (Pattberg et al., 2018), the role of network variables (Niles & Lubell, 2012), and the nesting of collaborative efforts at different scales (Margerum & Robinson, 2016). Acknowledging this, this paper offers some fresh insights and knowledge by assessing the embeddedness of participation and responding to the question of: What mechanisms provide linkage and enable influence between participatory processes and wider governance systems? This paper examines this question using the example of WFD implementation in Germany where the federal states' decision-making arrangements include polycentric governance structures.

The European Union (EU) Water Framework Directive (WFD) has led to a tremendous increase in participatory processes in various decision-making settings, pursuing, in general, similar ecological outcomes—the protection of all EU waters. The WFD (Article 14) prescribes the encouragement of public participation with the expectation that this would improve the overall achievement of its key goals (Preamble 14). The CIS guidance document No. 8 (European Communities, 2003, p. iv) defines public participation generally “as allowing people to influence the outcome of plans and working processes” and as “a means of improving decision-making, to create awareness of environmental issues and to help increase acceptance and commitment towards intended plans. Public participation for the implementation of the Directive is recommended at any stage in the planning process”. However, apart from public consultations for river basin management plans (RBMPs) and associated programs of measures (PoMs), the WFD leaves open the matter of how exactly participation should be encouraged. This situation has resulted in a huge diversity of participatory processes among and within the member states (e.g., Liefferink et al., 2011)—including Germany with its federal state structure.

By recognizing institutional complexity, the idea of polycentric governance has increasingly attracted attention from researchers (Jordan et al., 2018; Thiel et al., 2019; Van Zeben & Bobić, 2019). The term ““Polycentric” connotes many centers of decision-making which are formally independent of each other. Whether they actually function independently, or instead constitute an interdependent system of relations, is an empirical question in particular cases” (Ostrom et al., 1961, p. 831). Polycentricity as a lens (Blomquist & Schröder, 2019) allows an open analysis of any kind of decision-making structure without restricting the view to federal structures, market structures, or networks or the presence of multiple levels. Thus, this lens helps us shed light on Germany's high degree of diversity. While being characterized by a similar cultural and regulatory (transposition into the National Water Law) background, the federal states vary in their decision-making structures—that is, levels, types, and multiplicity of actors and decision-making power—as well as in their approaches toward public participation. This diversity allows us to compare the processes' and systems' characteristics individually as well as combinations of these characteristics regarding processes and actors.

Contrasting intended process purposes and characteristics of the decision-making systems with theoretical considerations on polycentricity allowed us to identify mechanisms linking participatory processes and governance systems. In doing so, we applied an explorative approach which looked for commonalities and differences in data. Data were collected through semi-structured interviews with a cross-level selection of actors for each governance system, combined with participatory observation of processes and document analysis.

PARTICIPATORY GOVERNANCE IN POLYCENTRIC SYSTEMS

The performance of polycentric governance systems is widely debated (Aligică & Tarko, 2012; Huitema et al., 2009; McGinnis, 1999; Morrison et al., 2023; Pahl-Wostl et al., 2012; Schlüter et al., 2010). Questions regarding performance relate to the debate whether resources are better governed by a central government or decentralized and self-organized by rather local actors (Andersson & Ostrom, 2008). Compared to monocentric governance systems, polycentric systems are expected to better adapt to problems of different scales and local needs but also to miss scale-effects and to face difficulties in achieving and sustaining agreements (Huitema et al., 2009; Ostrom & Parks, 1999).

Similarly, it is often asked whether participation is worth the effort? Are there correlations between process design or process outputs and social or environmental outcomes (Ansell & Gash, 2007; Blackstock et al., 2007, 2012; Newig et al., 2018; Özerol & Newig, 2008; Rauschmayer et al., 2009; Rowe & Frewer, 2000, 2004; Ulibarri, 2015)?: Does participation contribute to effectiveness, beyond emancipation and legitimacy (Albrecht, 2016; Newig, 2007), or to a substantive or instrumental effect beyond normative or legalistic rationales (Blackstock et al., 2007; Wesselink et al., 2011)? Or might participation be even disadvantageous when leading to a non-adoption of environmentally favorable plans (Heijden & Heuvelhof, 2012)? As such, participatory processes require a more systematic analysis of causal mechanisms (Trein et al., 2021), especially on how to leverage benefits and on understanding pathologies associated with power asymmetries (Carlisle & Gruby, 2017).

We take the view here that the phenomena of polycentricity and participation are neither necessarily good nor bad. We seek to assess how polycentricity characteristics affect the functioning of the overall governance system and how some of those characterizing variables, especially the multiplicity and independence of decision-making centers (Schröder, 2018), interact with participatory governance.

However, only a few case studies have analyzed the rationales for participation and choices regarding process design. The question of ‘why’ is ignored in most literature (Wesselink et al., 2011).

Numerous studies, in contrast, have nevertheless highlighted important variables influencing the effectiveness of participatory processes and their environmental outcomes, including process design characteristics, the previous experiences of participants and power asymmetries (Ansell & Gash, 2007). To date, research has concentrated on the participatory process itself, meaning that participatory processes were analyzed separately from the systems' decision-making structures. This process orientation manifests itself by focusing on inclusiveness/representativeness, effects on representatives and decisions taken by a process, and how decisions are characterized regarding innovation, cost-effectiveness, environmental standards, etc. (e.g. Beierle, 2000; Holley, 2010; Newig et al., 2018; Vente et al., 2016; Wright & Fritsch, 2011).

However, it is also critical to know what happens outside and after the immediate participatory process (Bull et al., 2008), particularly since the context may determine outcomes (Vente et al., 2016). The context is set, in part, by the distribution of power (Speer, 2012; Vente et al., 2016) and the institutional set-up, the degree of political decentralization, the relation between the executive and the legislative, government resource endowments, and the size of the jurisdiction (Speer, 2012). Thus, expanding the analysis to the level of the governance system and its characteristics allows to identify mechanisms which link processes to their systems. This can help understand the challenges for participation resulting from scaling up (Stringer et al., 2006).

In general, more centralized systems may achieve enhanced adaptive capacity and good institutional fit by creating sub-units, while more decentralized systems may achieve them by coordination (Ostrom et al., 1961). In particular, decision-making centers need linkages which allow for deliberation and learning (Carlisle & Gruby, 2017). Such linkages can be provided by cooperative or participatory processes which allow decision-making centers to consider interdependencies (Morrison et al., 2023) and which may facilitate the processes of convergence and divergence (Pattberg et al., 2018). These linkages may be generated through joint memberships, which bridge participatory processes and different organizations (Pattberg et al., 2018), and joint decision-making.

Joint membership reflects dimensions of polycentricity as well as participation, particularly the importance of a multiplicity of actors (Blomquist & Schröder, 2019) and the breadth of involvement (Newig et al., 2018). Nevertheless, membership can only link processes and wider governance systems if actors are involved. Hence, it is crucial to investigate participation and non-participation jointly (Angst et al., 2022). A link to those not involved may occur due to cascading events which are described in network science as knock-on effects on neighborhood nodes after an initial change in one node leading to large-scale effects (Pattberg et al., 2018). However, the effects of processes on constituencies were rarely studied (e.g., Campbell et al., 2011).

Joint decision-making is a form of power delegation and is another dimension of participation (Newig et al., 2018). Following Lubell and Robbins (2022), joint decision-making through participation might be understood as a form of decision-related centralisation, if the decisions are binding. If the decisions are binding, the independence of decision-making centers in the respective polycentric system is reduced through participation.

Both joint membership as well as joint decision-making are a matter of process design and often viewed and analyzed as involvement or representation and power delegation in process-oriented studies. However, joint membership and decision-making also deserve attention beyond individual processes because process-focused studies already pointed to the importance of independence in polycentric decision-making in the aftermath of participatory processes: for example, explanations for why public managers encourage collaborative governance (Scott & Thomas, 2017) include the importance of external decision-making for providing resources and taking actions. Furthermore, there are often uncertainties regarding whether environmental outcomes result from a sound management plan or an imperfectly implemented poor management plan (Rauschmayer et al., 2009).

Thus, in this paper, our goal is to contribute to assessing the functional quality of polycentric systems (Morrison et al., 2023) more systematically (Carlisle & Gruby, 2017) and especially regarding the role of cooperative and participatory processes (Lubell & Robbins, 2022; Morrison et al., 2023) by reflecting on the system-theoretical assumptions of polycentricity (Pattberg et al., 2018). The paper empirically analyzes the “why” question in addition to exploring the nature of joint membership and joint-decision-making. The findings and insights shed fresh light on how processes are embedded into wider governance systems.

CASES AND METHODS

The WFD aims to bring about good ecological and chemical status in all European Waters by 2027 at the very latest (Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit [BMU], 2010). This includes a requirement to establish monitoring programs and management cycles which include the preparation of river basin management plans (RBMPs) and associated programs of measures (PoMs) which are to be implemented during each 6-year cycle.

Nevertheless, it appears highly unlikely that the Member States will achieve their ambitious aims by 2027 (European Environment Agency, 2018). Only 8.2% of the German surface waters reached the stated ecological goals by 2018 (Bund/Länder-Arbeitsgemeinschaft Wasser [LAWA], 2018). Therefore, a larger in-depth comparative study of WFD implementation in the six German federal states of Saxony, Saxony-Anhalt, Hesse, North Rhine-Westphalia, Thuringia, and Lower Saxony was conducted to investigate the influences on implementation decisions in polycentric governance systems. Among these influences are participatory processes which are the focus of the research presented here. To address the complexity of polycentric governance, the study adopts an exploratory approach and focuses on decision-making related to the achievement of WFD goals regarding hydromorphology and connectivity.

This focus was selected because, additional to measures addressing nutrient pollution from agriculture and toxic substances, measures addressing hydromorphology and connectivity (e.g., re-meandering rivers and diversifying shorelines with river wood and gravel) showed the largest gap between identified but not yet implemented measures (LAWA, 2018).

The German states are characterized by complex decision-making structures of three to four (Bogumil & Jann, 2009) general purpose administrative levels (municipalities, counties, district governments/middle authorities (state-wide responsibility below ministries)/none, and ministries) and in addition various special purpose authorities as well as public and private entities. The study covers a diversity of the structures of the territorial federal states, which were expected to face cooperation and participation challenges differing from city-states due to the scales which need to be recognized. The selected states, three from the former East and three from the former West of Germany, represent the different general-purpose administrative level structures (macro-organizational structures). They are all further characterized by a multiplicity of independent decision-making centers implementing WFD goals regarding hydromorphology and connectivity, but they vary in the degree of multiplicity, the independence of actors, and actor-types (micro-organizational structures). This allows the identification of cross-cutting patterns among varying polycentric governance systems.

In order to assess the various roles and functions related to WFD implementation, semi-structured interviews were conducted between January 2017 and November 2019 with actors (see Table 1) from the water sector at all levels within the federal states and especially different types of local level actors actively realizing WFD measures (public and private actors, associations). This was complemented by non-state actors, who were in a position to give a detailed overview of the implementation situation in the states, and particularly included nature conservation associations due to their roles as critical observers and environmental advocates.

Interviewees were identified through the analysis of policy documents, participation in WFD-related exchange formats (e.g., organized by nature conservation associations or professional organizations across Germany), the observation of participatory processes, and a snowballing approach which involved asking contacts to recommend further contacts. The listed actor types in Table 1 are not all relevant as WFD decision-makers for all states (compare with Table 2) due to state-specific arrangements, and some types do not exist at all in some states (e.g., the aforementioned level structures or special-law water associations which only exist in North Rhine-Westphalia). In a few cases, no contact could be made with a few actors (especially small actor-types without WFD implementation capacities) or no interview agreement could be achieved: There was however a brief conversation with the staff of the Saxonian Ministry during a participatory observation. In North Rhine-Westphalia, water and soil associations were examined through an interview with the county's water authority, in Lower Saxony through interviews with other actors, which are specified in the interview list. Similarly, municipalities were captured. “Other” actor-types were relevant especially when tasks had been transferred, e.g., to counties, landscape planning associations, umbrella organizations, or any organizational solution which is not covered by the overview in Table 1. The applied explorative approach adopted for the study made it possible to identify such structures.

The 70 interviews each lasted approximately 2 hours and used open-ended questions to explore pertinent issues such as how WFD implementation happens in each state, the roles interviewees played in WFD implementation, their responsibilities, which participation/cooperation processes they utilize and/or participate in, their expectations for the processes, and how they perceive the overall processes and other actors.

To see how German WFD implementers understand participation, we looked for processes which the federal states had themselves named as “participatory” or “collaborative” with regards to complying with the WFD on official websites, in documents, and in statements by steering level authorities. As process names changed over time and interviewees sometimes recalled process names only partially, tracing some participatory processes proved difficult (see the Supplementary Material for a list of all processes found). Overall, the states rarely explicitly distinguished between coordination, collaboration, and participation processes or between participation of the wider or interested public or participation of the water sector and other sectors. Thus, the processes which we found are placed all along the steps of Arnstein's ladder of participation (Arnstein, 1969).

Regarding process observation, potential dates were identified through public announcements and by approaching interviewees. Whenever possible, the first author here sought permission to participate and, when granted, joined the meeting. Some process organizers were reluctant in disclosing dates and allowing research observation, while others were very open to the idea, even though participation was only formally possible by invitation. Thus, 12 processes were observed (between April 2017 and November 2019) out of 21 processes which were included in this analysis.

Additionally, we analyzed policy documents, such as interim reports, decrees, state-level contributions to RBMPs, programs for water protection and working papers, as well as recorded information from participatory processes (see the Supplementary Material for lists of documents providing information on each process). This data source reflects the overall diversity of the states: Some processes have their statutes, while others are only briefly described in WFD-related administrative documents. For some processes, meeting minutes were recorded and published on a dedicated website, while others made them only internally available to participants or not at all. For some, but not all, processes, intended or actual participants' lists could be found. Their level of detail varied for individual events associated with the same process. For example, names for representatives of particular organizations appeared in some but not all documentation, or only changes of representatives were documented. Due to the variation in data quality of the open data collection process, the study is partially constrained by a lack of data comparability. Nevertheless, this approach allows some general and interesting insights.

Data were coded in MAXQDA 2022 (VERBI Software, 2021) to identify cross-cutting patterns among the selected states, processes, settings, and rationales. First, statements or information on tasks and responsibilities, structures, and the actual decision-making behavior and in- and interdependencies were identified, as well as on process purposes, participants, non-participants, and organizers. Afterward, the data were coded in greater detail, e.g., intended vs. actual participants, and clustered: process purposes regarding their kind and participation in relation to the overall number of decision-makers. Details on the process of clustering are provided linked to the results in the respective results' sections.

In the remainder of the paper, we use the interview number ‘[Ix]’ and the process observation number ‘[Ox]’ to refer to interview statements or, respectively, particular aspects that were observed in participatory processes (see the references for a complete and numbered list of interviewed actors and observed processes).

DISENTANGLING THE SCENE: PROCESSES AND ACTORS

This part of the analysis disentangles process elements from system elements, allowing cross-case comparisons regarding the following questions: Who takes decisions in WFD implementation systems? Who organizes participatory processes and how many of systems' decision-makers are involved in the processes? How do the processes themselves and their organizers seek to influence WFD implementation?

LINKING PARTICIPATORY PROCESSES TO WIDER DECISION-MAKING SYSTEMS

This section links the analyzed processes to their wider decision-making systems by further elaborating and generalizing the process intentions. We conceptualize different influencing situations by distinguishing between whether power was transferred or not and whether decision-makers are directly involved or not.

We also considered how power transfer and involvement determine the effect of aggregation and multiplication on decision-making.

Influence within processes means that decision-makers are directly involved (Figure 2). If binding decisions are taken within a process, the decision-making rule (e.g., majority voting) defines how information is aggregated into a decision, while commitment and compulsion mechanisms define how effects are multiplied. Through collectively binding outputs, follow-up decisions along the chain from the output of the process to the (environmental) outcome become less independent, and the system less polycentric. This effect depends on the effectiveness of commitment or compulsion mechanisms (e.g., enforcement mechanisms). In contrast, if processes take no decisions or non-binding decisions (decision-makers stay as independent as without the participatory process), the process becomes one out of many factors influencing their decision-making. In such cases, it does not matter whether decisions are taken by organizers or participants.

Influence beyond a process refers to situations in which decision-makers are not directly involved (Figure 3). Decision-makers can be represented or not considered at all. This adds additional decision-making points (consecutive and requiring clearance or processing to be passed) to the linkages between processes and governance systems. For aggregation, this is a representative's aggregation rule (e.g., how an interest association collects and relays its members' interests). For multiplication, the representative as well as the ultimate decision-maker decides how to consider processes' information for their decision-making. Binding decisions, though, may allow compulsion or sanctioning in cases of non-compliance.

Overall, due to the independence of decision-makers, we expect aggregation as well as multiplication to become increasingly difficult and challenging the more decision-makers are involved in a system—in other words, the more polycentric a system is. From our findings, we expect these mechanisms also to show more imperfections (dotted lines in Figure 2 and Figure 3) in contrast to ideal conditions, as depicted in Figure 1: Some information is not aggregated into a decision, and not all information is multiplied to all decision-makers.

IMPLICATIONS FOR RESEARCH AND PRACTICE

Considering the observed embeddedness of participatory processes in polycentric governance systems, these findings imply that our current understanding of decision-making and what we may or may not expect from participation need to change.

CONCLUSIONS

By analyzing the implementation of the EU Water Framework Directive (WFD), this paper provides some fresh insights regarding the role of participatory processes in polycentric governance systems.

The decision-making power regarding measure implementation showed, even within the chosen sub-set of WFD implementers, to be widely spread across multiple actors. These multiple actors themselves had no ultimate planning power in their area of responsibility. The study revealed multiple small-scale, rather local-level, decisions regarding implementing measures, which contrasted with a few large-scale participatory processes organized by higher level authorities.

Although the processes reflected different strategies for coping with the multiplicity of decision-makers, such as plural parallel processes, representation or rather large events reduced to information giving, in most cases a large share of WFD implementers were not directly involved. The analyzed processes varied in both scope and the types of issues addressed, as well as process purposes. However, in none of the processes, decision-making power was transferred from actors to the process.

Nevertheless, the organizers of participatory processes intended to positively influence the wider governance system through influencing participating decision-makers, but also the represented constituencies. The intention analysis showed that the participatory processes and governance systems are linked through mechanisms with two directions: Through the mechanism of aggregation, the information of many actors and decision-makers is aggregated into one decision, e.g., when advice is intended. Through the mechanisms of multiplication, many decision-makers shall be influenced by one or few other decision-makers, e.g., through the provision of information.

The nature of these links makes involvement and process decisiveness (power transfer) important variables for the effectiveness of these mechanisms. Based on this distinction, participatory processes may either affect the overall governance system through the binding character of decisions (commitment and compulsion), or, without power transfer, they solely rely on process design variables to affect directly involved decision-makers and on multiplication to affect the wider governance system.

The multiplicity and independence of the decision-makers in polycentric systems hamper the achievement of ideal aggregation and multiplication through participatory processes. The larger the scope of a process, the more polycentric a system can be within that scope, and the more difficulties can be caused by the multiplicity of actors. These difficulties are, for example, transferring power to processes, and, involving all actors within the scope of a process. The resulting imperfections in aggregation and multiplication may let effects of participatory processes on the wider governance system appear rather random.

Looking to the future, even if process effects become rather random in polycentric governance systems, the added value of participation should also be analyzed in the light of alternatives to its basic functions, aggregation and multiplication, for communication among multiple actors: Participation offers alternatives to multiple single processes and efforts with many actors individually. Thus, participation might tie communication. We should further ask how participatory processes change the system's communication patterns in the long run in polycentric systems. Processes might have fewer effects on ultimate decisions implemented in the environment. However, they certainly affect planning through selecting plans or ideas out when recognizing insurmountable obstacles for implementation. Hence, we may ask whether and under what conditions participation affects policy implementation (accelerates implementation processes or hampers pathways) through identifying the difficult or non-options and those ideas which get a chance (anticipated chance for success), especially if processes cannot be proofed to improve particular decisions.

INTERVIEWS AND OBSERVATIONS

The following tables show the actors interviewed and processes observed for the analysis of each German federal state. They are numbered for referencing in the text. The time frame for interviews is indicated.

Interviews:

Saxony-Anhalt: January 2017, March–June/August 2018.

Saxony: January/April/May 2017, December 2018, January 2019.

Hesse: September, November 2018.

North Rhine Westphalia (NRW): October–December 2018, February 2019.

Thuringia: January–March 2019.

Lower Saxony: January, June, July 2017, September 2019.

Participatory observation: