Final release of the network database and associated documentation: Deliverable D 2.2, Work Package 2 Stocktaking of the decision making context

Task 2.2 “Clustering and network analysis” of the PLACARD project aimed at supporting the mapping of interactions within and between CCA and DRR communities and assessing quantitatively the roles that different actors have in them, using a Social Network Analysis (SNA). SNA techniques are used to answer questions about how actors are connected to each other, how strong their relationships are and which actors are best positioned to connect other actors in a network through the calculation of indicators such as degree, closeness, betweenness centrality and clustering coefficient. In the context of the PLACARD project, two SNA exercises were carried out to investigate the intensity of actors’ interactions (i.e. on a scale from 1 to 5 representing lack of interaction, weak and strong communication, and weak and strong collaboration), as well as the type of interactions (i.e., whether an interaction is related to CCA, DRR, or both fields). Social network metrics were calculated to quantitatively assess the roles of different actors in the network and their interrelationships. In particular, we focused on centrality measures – degree, in‐ degree, betweenness and eigenvector – that are considered good indicators of an actor’s power position, meaning the strength of the role played by an actor in influencing interactions in a network. The first exercise took place in summer 2016 and focused on the interactions between CCA and DRR actors operating at the European and International level. Data was collected from the responses of 32 out of the 35 actors that were invited to participate in an online SNA survey. The European Climate Adaptation Platform (Climate‐ADAPT) emerged as the actor with the highest degree, eigenvector and betweenness centrality values. Besides Climate-ADAPT, the Directorate‐General for Research and Innovation (DG R&I), the European Environment Agency (EEA) and the Directorate‐ General for Climate Action (DG CLIMA) were identified as actors with high centrality values. The actor with the highest in‐degree centrality, an indicator that considers the number of ties that a certain actor has as specified by other actors in the network, was DG CLIMA, followed by EEA and IPCC. When it comes to the analysis of whether interactions are related to CCA and/or DRR issues, collaboration appeared to be most often related to both CCA and DRR, while communication mostly related to one of the two areas when these were considered separately. Cluster analysis was applied to explore if actors could be grouped on the basis of SNA metrics. The application of Clauset‐ Newman‐Moore algorithm revealed two large groups, which clearly depicted the two main communities: DRR and CCA, and one significantly smaller cluster. A detailed description of the first SNA can be found in the PLACARD Milestone 10 report (Bojovic et al., 2017). The second round of the SNA initiated in autumn 2017. This aimed at exploring the two-way interactions between national level actors in four European countries: Germany, Italy, Switzerland and the United Kingdom, as well as the one-way interactions between national level actors and a small group of international actors whose role was identified as important based on the output of the first SNA. Data collection took place in parallel in all four countries, while data was analysed separately for each one of them. The output of the analyses along with a description of the methodological approach employed in this study and a review of CCA and DRR studies that have used SNA during the last two decades are presented in detail in this report (PLACARD Deliverable 2.2 report) (Karali et al., 2018). Selected observations and general remarks from this second exercise are presented below: Governmental actors and research institutes were well represented in all four countries. Output of the SNA analyses confirmed our hypothesis that these actors have a central role in national CCA / DRR networks. • Knowledge platforms and businesses were less equally represented across the four countries, yet they had an important role in the networks in which they were present. Especially for the case of business / companies, relevant actors appeared to be very active in networking, reaching out to a large number of actors (see the out-degree centrality metric). The importance of these actors’ role in the CCA / DRR networks is expected to increase further in the near future. • In many cases the same actors occupied the top places in the ranking of more than one of the calculated SNA metrics. This observation reflects the key role of certain actors in interacting with other actors in their networks as well as in setting up interactions with actors that are less well-connected. • The one-way interactions of national level actors in the four countries with the six international actors (as indicated by the former) suggest that the former interact more with international actors with whom share a similar field of expertise. • Differences exist in the type and intensity of interactions among national level actors in the four countries and the ranking of the different types of actors considered in each network. While some general remarks can be made based on the output of the four SNAs, a full crosscountry comparison is not considered useful in this study due to the differences in the composition of the four networks. • Both CCA and DRR are dynamic and highly evolving fields. This implies that the roles and the importance of actors may change over time as a result of the changing needs in the different networks. Already in this study, the important role of actors such as networks, nongovernmental organisations or non-governmental advisory boards emerged. • Ensuring the diversity of voices and the involvement of actors that are often neglected in relevant exercises is expected to benefit the network in terms of its potential to expand and increase its strength and relevance in the long-term. • Several challenges emerged at the different stages of the SNA exercise. Setting the network and identifying suitable survey participants were key challenges during the design of the SNA. High workload, fatigue, scepticism about the way that data will be used and the repetitive pattern of the SNA questionnaire were identified as barriers during the data collection process that had to be overcome in order to minimize their impact on the response rates of the surveys and the completeness of the collected responses. Finally, during the phase of the data analysis and the interpretation of the results, we were confronted with two ‘weak’ points of the SNA method: (i) the fact that the method is a rather static, as it can reflect actors’ interactions only at a specific point in time, and (ii) that it cannot capture the reasons why certain patterns emerge. These challenges reinforce the idea that the applications of SNA need to be complemented by other methodological tools (i.e. interviews, focus groups) to support a more insightful interpretation of their results.