The evolution of policy beliefs in times of crises: a network approach

Since its birth, the Advocacy Coalition Framework (ACF) has become one of the most employed frameworks in order to analyse public policy, focusing on policy subsystems, advocacy coalitions, belief systems, policy change, and policy learning. Methodologically, Sabatier and Jenkins-Smith proposed content analysis as the most suitable method, not only to identify the multiple belief systems, but also to uncover changes in them (Sabatier & Jenkins-Smith, 1993). However, especially in the last decade, the ACF has been supported by the study of policy networks. Therefore, the basic premises of the framework, namely belief systems, coalitions, and policy learning processes have been merged with the material coordination of networks among actors (Leifeld, 2020). This paper aims at deepening the interconnection between the core elements of the ACF and network approaches. In accordance with the basic assumptions of the theoretical and analytical framework, the paper focuses on a specific policy subsystem, namely EU asylum policy. In effect, throughout the last decade, the second phase of the Common European Asylum System (CEAS) was finalised, two proposals for amendment of the system have been put forward, as well as the system has been put under extreme pressure due to the multiple crises impacting the overall EU polity, and more specifically, EU asylum policy. However, even though this policy subsystem is considered one of the most dynamic policy areas of EU integration, it appears to be in complete deadlock. Acknowledging that the ACF assumes that policy change may occur – among others – due to policy-oriented learning and shocks, the present paper investigates both these routes. On the one hand, it focuses on potential policy-oriented learning processes through the change of policy core beliefs within the aforementioned policy area. On the other hand, it looks at the possible effects of internal, external, and/or creeping crises on EU asylum policy subsystem. In order to do so, the timespan explored is of ten years (2012-2022), and the data analysed consist of newspaper articles from two European outlets (Politico.eu and Euobserver.com). From a methodological perspective, the tools of Discourse Network Analysis (DNA) are employed. DNA is a combination of qualitative content analysis and social network analysis, allowing for the depiction of policy actors, coalitions, and policy beliefs, as well as their evolution over time. The innovation of this paper stands not only on the development of some core ACF analytical concepts, but mostly on the empirical testing of these concepts through DNA.