Unveiling the anatomy of autonomy: dissecting actor-level independence in the European External Action Service

ABSTRACT

Formulating and implementing public policy in Europe has historically been a core task of national administrations. This study suggests how this role has become challenged in a least likely policy field – foreign affairs. The ambition is to reassess the autonomy of the European External Action Service (EEAS) by examining actor-level autonomy of EEAS staff, while also suggesting key determinants thereof. Two conditions of actor-level autonomy are empirically illuminated: bureaucratic structure and the geographical location of the EEAS. Benefiting from two novel data sets which include a survey and élite interviews of EEAS officials, two empirical observations are highlighted. First, despite being an embryonic organization embedded in a field of core state powers, EEAS officials demonstrate substantial actor-level autonomy. Second, the behavioural autonomy of EEAS staff reflects primarily the supply of organizational capacities inside the EEAS, but much less the geographical location of staff. Actor-level autonomy is thus not only profound, but largely supplied by in-house organizational factors.

KEY WORDS:

• Actor-level autonomy
• bureaucratic structure
• organizational affiliation
• organizational approach
• geographical location

ACKNOWLEDGEMENTS

This study is financially supported by two grants: one from the Norwegian Research Council (EURODIV) and one from the University of Agder (COMPOL). Previous versions of this article were presented at the Seventh ECPR Pan-European Conference, The Hague, and the Norwegian Political Science Association in Tromsø, in 2014. The authors would like to thank the participants in the public administration and public policy group, three reviewers, and the editors for valuable comments and suggestions.

Notes

1 These proxies have earned the status as standard measurements of actor-level autonomy in the literature. Studies of both national administrations and the EU administration have shown that these parameters are closely associated and measure the same latent concept (e.g., Egeberg et al. 2015; Sowa and Selden 2003; Trondal 2010; Yesilkagit and Van Thiel 2012). Bivariate correlations (Pearson's r) also show strong and significant correlations among our proxies. Still, this study does not assume any correlation among these measurements since this is not the aim of the study, but rather using them as separate measurements that allow for discrete observations of actor-level autonomy.

2 The 21 officials indicating ‘other’ as their previous affiliation refer either to various Commission DGs, not mentioned in the questionnaire, other national authorities (ministries) or agencies (e.g., police, development agencies), as well as international organizations (IOs) or research institutions as their affiliation of origin. Four respondents did not disclose their institution of origin.