Order with and without design in polycentric governance systems: Exploring the role of independence and multiplicity in decision-making for WFD implementation processes

In 2000, the EU Water Framework Directive (WFD) became the central regulatory instrument of the European water protection policy. However, it is doubted that Germany, much like other member states, can achieve the ambitious aims by 2027. ‘Why is the goal achievement stagnating?’ and ‘How to overcome implementation gaps?’ are, therefore, reoccurring questions. Similar obstacles to implementation can be observed across time and policy issues in policy implementation studies. Multiple answers have been proposed and governance structures are identified as one important influencing factor. One key question became whether a policy is better implemented through a single ultimate authority or any non-central approach, what the polycentricity literature terms as monocentric vs. polycentric governance. This qualitative study contributes empirically to these questions by a cross-local comparison. It asks: How do implementation arrangements (organisational structures and institutional settings), or governance structures, influence the WFD implementation, especially at the local level, and regarding the adoption of WFD prescriptions of the river basin management approach, the coordination across borders and sectors and public participation? The implementation processes of the WFD were studied in six German states focussing on issues related to the hydromorphology and connectivity of rivers. For that, the polycentricity lens was used. The implementation arrangements varied across and within federal states and levels expressing various shades between monocentric and polycentric governance. Data were collected through policy document analysis, semi-structured interviews with relevant actors and participatory observation of participatory processes.
Micro-level implementation patterns have been found regarding implementation barriers, instruments, implementation approaches, the use of discretion, and organisational structures. These patterns resulted in a mix of central and decentral approaches at the macro-level leading to weak effects of instruments, lacking influences between levels and sectors and misfits of approaches taken by the steering levels in parallel to local levels. Nevertheless, the polycentricity of the systems also allowed lighthouse projects to develop despite the significant implementation barriers.
The patterns were found to be partially caused by fortuity and partially by the design of institutions at different levels and times – partially due to the WFD, but mainly path-dependent as structures existed before the WFD. Due to the WFD, primarily new goals were postulated and, through new instruments, steering-level actors attempted to change the resource endowments of existing actors regarding WFD measures. In contrast, relating structures between actor groups, and thus power relationships, rarely had been changed. The independence of actors makes it difficult to identify an ideal implementation system because multiple actors use the same institutional frame differently. It would also be difficult to transform systems towards the ideal even if the ideal were known. Changes would need to happen at many points in parallel – orchestrated by a central designer or concerted through a joint agreement. Cumulative, more or less undirected, design, though, is dominating which fits the nature of polycentric systems. Nevertheless, governance systems can be optimised. To improve and accelerate the design efforts learning is necessary. So far, however, the incredible diversity of approaches has been rarely used to learn from regarding governance questions. Hence, this study elaborated a cyclical approach to governance. This cyclical approach is based on repeated assessments and a central provision of data. It shall simplify data access, thus, learning from each other and shall accelerate system optimisation to satisfy the multiple (new) purposes.