Collaboration to mitigate climate change – Does the institutional context matter? A comparative study of Finland, South Korea, Switzerland, and the United States

Policy actors often try to influence decisions and outputs by joining forces and coordinating their action. Understanding patterns of coordination is important for the field of environmental policy and governance, as environmental policies typically target a wide range of stakeholders from different sectors. The public policy literature has advanced and tested several hypotheses concerning the factors that promote or hinder collaboration at the national or subnational level. However, an important caveat of the literature is that only a handful of studies compare these factors across different institutional contexts. In this paper, we draw on the existing literature to formulate hypotheses concerning the role of beliefs, actor types, and resources in shaping collaboration patterns. We test these hypotheses using data on the climate change policy networks in four countries with very different institutional contexts: Finland, Switzerland, South Korea, and the US. The crucial institutional axes along which the cases differ from each other are (a) federal (Switzerland, US) vs. unitary (Finland, Korea), (b) corporatist (Finland, Switzerland) vs. pluralist (US, Korea), and (c) majoritarian (USA, South Korea) vs. consensual (Finland, Switzerland). Estimating Exponential Random Graph Models for each country, we find no systematic variation across the different polity contexts, which is good news for important public policy theories such as the Advocacy Coalition Framework, policy networks, ecological modernisation theory, or resource dependence theory. However, we do find significant variation between different national policy contexts. Firstly, actors collaborate with others sharing beliefs or the same type of actor in particular during conflictive phases in the policy process, but not when major decisions are already made. In particular, the Swiss case illustrates that when important directions of a policy process are already laid out, actors of different beliefs or actor types start to collaborate across conflict lines. In a similar vein, public authorities are also more important targets in phases were important decisions are still to be made. However, one finding that holds for all countries across all polity and policy contexts is that resourceful and influential actors are popular collaboration targets.Many scholars would agree that political decisions are the result of collective action. By joining forces and coordinating their behaviour, actors try to influence decisions and outputs. Understanding these patterns better is highly relevant for the field of environmental policy and governance, as policies target a wide range of stakeholders from different sectors. Increased collaboration efforts are needed to ensure that environmental concerns are considered particularly in sectors with traditionally contradicting interests, such as economic, energy, agricultural, or traffic policy. The public policy literature already offers an in-depth knowledge of factors that promote or hinder collaboration at the national or subnational level. However, only a small number of studies also compare these factors across different institutional contexts. This is an important caveat of the public policy literature, as a consistent theory that explains collaboration across countries and policy domains is still absent and more comparative studies are needed. Hence, this study promotes a systematic comparison of the climate policy subsystem in four countries with different institutional settings to disentangle important characteristics that might influence collaboration. Specifically, we ask whether the institutional set-up (polity) in our case countries affects drivers of collaboration (politics). We assess three crucial institutional dimensions to systematically assess differences in collaboration structures, i.e. federalism (Switzerland, US) vs. centralism (Finland, Korea), corporatism (Finland, Switzerland) vs. pluralism (US, Korea), and representative democracies (USA, South Korea) vs. consensual democracies (Finland, Switzerland). We base our theoretical argument in the ACF literature and expect that corporatist and consensual countries exhibit less clearly demarcated advocacy coalitions based on collaboration than in representative or pluralist countries. Also, it is likely that all kinds of actors collaborate in corporatist and consensual countries, whereas vertical institutional levels structure collaborative patterns. Finally, in federal countries the additional level of governance raises the number of different interests that must be coordinated in the policy subsystem. We assume that in federal systems collaboration patterns are characterised by a higher level of complexity and level of conflict that bonds actors by beliefs and interests. As collaboration is necessarily interdependent, we use policy network data for our analysis, which was collected in connection to the international research project Comparing Climate Change Policy Networks (COMPON). To date, in 19 countries, elite surveys were conducted to gather information on the most important actors, with respect to their collaboration patterns, their sources of scientific information, their resources, interests, political activities, and their beliefs about climate policy. We started our analysis by running Exponential Random Graph Models (ERGM) for all four countries. Our preliminary results underline the assumption that collaboration obeys other rules in different institutional settings and gave a first impression of possible correlations between institutional characteristics and collaboration. Refining the analysis with a systematic, qualitative comparative of the four cases is the next step. The aim of this endeavour is to develop a blueprint of the relationship between institutional factors and collaborative actor behaviour to be further developed by future research and to draw some first conclusions about the implications for policy output