Who learns what? Policy learning and the open method of coordination

This paper provides a theoretical and empirical assessment of the claim that the open method of coordination is a learning-based mode of governance. The paper presents four arguments. Firstly, learning in a political context is not a truth-seeking exercise. It is a political exercise. Secondly, the OMC may well have potential in terms of new governance. However, even when it is examined in its pure, ideal-typical form, open coordination has contradictory aims. It seeks to mute politics and to encourage high-level political coordination, to facilitate bottom-up learning and to steer learning processes from above, to encourage cooperative learning and to spawn dynamics of competitive learning. This makes learning via open coordination more difficult. Thirdly, real-world open coordination provides empirical evidence of learning at the top (or ‘EU-level learning’), embryonic evidence of cognitive convergence from the top (or ‘hierarchical learning’), and almost no evidence of learning from below (‘bottom-up learning’ from regions and local conditions, or ‘social learning’). There are several reasons for this rather disappointing track record, most pertinently perhaps, poor participation, a partially wrong choice of instruments for learning, and lack of attention to the peculiarities of learning in politics. Fourthly, the pre-conditions for learning differ across the policies in which the OMC is currently employed. The structural elements of public policies define the scope for learning.