Coercive Capacity and the Politics of Implementation: Deportation

Abstract. Why are some bureaucracies in highly coercive policy fields able successfully to implement controversial policies while others bow to political opposition? This article challenges the common argument, based on a principal-agent model, that bureaucratic non-implementation is the result of the absence of effective legislative oversight. Instead, the article argues that, in coercive policy fields where the state imposes significant costs upon its targets, non-implementation can in fact be understood as the result of control efforts by elected officials. The article empirically tests this argument by comparatively examining the politics of implementation in the policy field of migration control. Drawing on interview data from Germany and the United States, the article identifies significant cross- and subnational variation in the capacity of bureaucrats to implement contested deportation orders. The article argues that this variation can be accounted for primarily by institutionally-determined differences in the degree of political insulation of bureaucratic agencies. The author gratefully acknowledges the financial support of the Social Science Research Council's International Migration Program which funded the fieldwork this article draws on. During the writing stage, a Dissertation Year Fellowship from Brandeis University provided much-appreciated financial support. For their immensely constructive comments on various drafts of this article, I thank Fiona Barker, Tim Büthe, Virginie Guiraudon, Alan Jacobs, Steven Teles, and two anonymous reviewers. I also greatly benefited from comments on the presentation of previous versions of this material at Boston College and the University of British Columbia, and at the conferences of the Council for Europea