This paper analyses how the interaction between social institutions on the one hand and actors' strategies and interests on the other hand is shaping European integration in research policy. We specifically focus on the implications of the existence of different conceptions of European integration (logics) on the emerging landscape of research funding programs jointly managed by the European Union (EU) and National States (joint programs). Our results display the central role of the introduction of a logic of coordination by the EU; it created a breeding ground for a new generation of programs and, at the same time, allowed to make the integration model more flexible and acceptable to National States (as funding became only virtually integrated). Most newly created programs were characterized by largely symbolic commitments and very small budgets, while stronger commitments had to be constructed through successive steps of integration. This process was highly selective and dependent on the presence of strong interests from the research community; additional funding from the EU was critical to ensure stability of national commitments. Further, National States by large delegated to independent funding agencies the management of national participations: delegation allowed to achieve greater homogeneity among national participants, but also to decouple decisions to participate (driven by compliance to institutional pressures) from the level of resources to be committed (driven by national interests). While in the year 2000, the European Research Area (ERA) strategy of coordinating national research policies was largely an empty concept, our case study shows how in the following decade, it was filled in with specific experiences and practices, led to the redefinition of actors' understanding of European integration and roles in European research policies and, ultimately, to the emergence of original models of integration.