Limitations to agency control in European Union policy-making: The commission and the poverty programmes

The principal-agent model (PAM) has produced valid hypotheses for conceptualizing actor relationships, but its disadvantage - as an economic concept transferred from the field of industrial organization and the theory of the firm to that of European integration - is often overlooked. This article argues that, when applying the concept, researchers interested in the empirical analysis of the EU policy process should be aware of some sensitive points concerning both 'internal' consistency and, external' theoretical constraints. Drawing on a case study of the EU poverty programmes, three behavioural pattems - 'discourse framing', 'lobby sponsoring' and,stretching' - are identified. These all indicate how the uncritical use of PAM may lead to faulty judgements about the actual degree of Commission autonomy in EU public policy-making.