https://orcid.org/0000-0001-8334-4244
![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
European administrative networks (EANs) are groups of national administrative organizations which are established to improve national-level implementation of European Union (EU) law. This paper addresses a key question concerning these networks: what drives interactions within them? To this end, the paper adds a dynamic political perspective to institutional hypotheses. It tests the resulting hypotheses using a most likely case, interactions in the European Migration Network (EMN) in the wake of the 2015–16 refugee crisis. Using social network analysis, it shows that two domestic political incentives are associated with interactions: the problem pressure experienced by member states following the 2015–16 refugee crisis and post-2015 declines in popular support for immigration. Our analysis also reveals that interactions occurred among member states with similarly high levels of government effectiveness. In sum, we show that EANs can provide significant added value for their members in politicized policy areas, although their utility may vary between network members.
Acknowledgements
We thank the reviewers as well as the participants of the Copenhagen Workshop on European Administrative Networks. The research for this article was funded by the Danish Council for Independent Research [Grant no DFF-7015-00024].
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 An additional objective for the EMN is to provide the general public with information on these subjects (article 3). This objective does not receive further attention in this paper.
2 The network also has a supranational element: the Commission chairs the steering board and network meetings, and the steering board uses majority voting (Decision 2008/381 article 4(3)).
3 NCPs have been established in the EU-27 minus Denmark, which is an observing member. Norway is an associate member, although it cannot apply for EU funding within the framework of the EMN (Ernst & Young, Citation2015: 103).
4 We only included publicly available ad-hoc queries, although a small proportion of queries are not made public. It should be noted that our preparatory interviews with EMN members did not indicate these non-public ad-hoc queries were more politically sensitive than national ones, so that we do not expect this to bias the analysis.
5 Unfortunately, the structure of our data does not allow for a longitudinal analysis of interaction patterns. The data on interaction patterns are based on aggregated queries from one member states to an array of responding member states, providing us with sufficient variation in tie weight across queries. For this reason, we could not compare interactions pre and post 2016, leaving us too little variation for a longitudinal analysis with a temporal ERGM. Instead, we control for potential structural effects by including two key structural variables: similarity in government effectiveness (H3) and organizational positioning. The inclusion of dummy variables for the years 2016 and 2017 as well as the inclusion of endogenous network structure effects prevent that we falsely ascribe network interactions to our dynamic variables (H1 and H2).
7 See table 2 in online appendix I for the complete overview of queries.
8 For examples of an included and excluded query, see box 1 in online appendix I.
9 Our focus on queries that explicitly mention EU implementation causes a risk of falsely excluding cases that deal with EU implementation without explicitly mentioning this. Yet, we preferred this over the false inclusion of other types of queries, for instance relating to comparison of national policies or gathering inputs for EU decision making. Including such queries would have introduced bias in our analysis, as these are not covered by our theory, which focuses on EU implementation specifically. Furthermore, there is ample scope for queries dealing with national policies or their implementation, given the fact that immigration is a shared competence, which entails a large scope for national policies and practices.
10 This is the technical term for an interaction between two network members. See online appendix II for a selection of ad-hoc queries and corresponding answers.
12 Please refer to the ergm package in R (Handcock et al., Citation2014)
13 See also figure 2 in online appendix I (left panel).
14 See also figure 2 in online appendix I (right panel).
15 In social network analysis, this is commonly called a ‘core-periphery’ network structure. The terms core and periphery, however, do not denote a country’s geographical position or its centrality in the 2015/16 migration crisis.
16 Model 2 does show a slight significant effect of first-entry states interacting more than non-affected states, but this effect might be attributed to the included dummy on government effectiveness and the fact that most high capacity states are also destination states.
Additional information
Funding
Notes on contributors
Ellen Mastenbroek
Ellen Mastenbroek is Professor of European Public Policy at Radboud University.
Reini Schrama
Reini Schrama is Assistant Professor at Radboud University.
Dorte Sindbjerg Martinsen
Dorte Sindbjerg Martinsen is Professor at the University of Copenhagen.