In nascent policy subsystems, political actors have a hard time identifying coalition partners, given that knowledge about subsystem boundaries (both in terms of relevant issue and actors) are unknown, and the belief systems and policy preferences of potential coalitions partners (as well as opponents) are difficult to assess. Actors in such subsystems are known to rely on prior knowledge and deep core beliefs for forming coalitions. With developing coalition structures, actors are further known to rely on the devil shift as a cue to structure the actor constellations and coalitions around them. In this paper, we trace the development of actor constellations and coalitions in two nascent subsystems, that is, AMR policy in Germany and Switzerland. We re-construct actor coalitions based on actors’ shared problem perceptions, actors’ shared blame attribution to others, actors’ shared solution propositions, and actors’ shared responsibility attributions to others. We expect coalitions to start forming based on the first, and continue forming around the four dimensions over time, in the order we discuss. Methodologically, we rely on texts from public media (two newspapers per country / subsystem) that we coded and transform into networks of actors’ similar profiles with respect to problems, blame, solutions, and responsibility. We thus analyse four network layers over four periods, in two subsystems that differ on the institutional setup but share the substantive issue of AMR.
