PRIME governance has been organised as a demonstrator of a lasting potential quasi-organisation serving as a base to the collective needs of the SIPS community. It was thus important that we tested the full set of organisational arrangements to address this objective.
We identified 4 difficulties which we have addressed and on which we intend to summarise the lessons we derive at the end of the network :

Members’ involvement: Consortium Agreement and Governing board
The first difficulty deals with representation and internal democracy. Should we mimick either inter-governmental arrangements (who give one voice to any actor, independant of its ‘weight’) or full representative demoncracy (with one voice to all academic staff, independant of the structures in which they work). We decided to choose an alternative way whereby research groups (one institution may be composed of more than one research group, as is the case of CSIC or CNRS) are given voting rights depending on three criteria: recognised excellence of its members, investments in preparing the next generation (in particular doctoral and post-doctoral candidates) and involvement in PRIME activities. Of course the latter was not present at the start of the NoE, but became more important with the deployment of the NoE. To insure the working of such a system, an “independent body” is needed to propose votes and a process has to be put in place to take into account potential disagreements. The former was delegated to the ‘characterisation group’ (see point 4 below) while the latter was taken care of by a well-carved consortium agreement, which we consider as a resource for other networks willing to adopt such an approach (click here to access the PRIME consortium agreement).
This enabled to have an evolutive Governing Board in charge of strategic orientations and of nominating members of the Executive Committee, the main operational structure of the NoE (see point 2). Again the approach developed was made to insure a wide representation of the numerous smaller groups without making the role of large ones disappear. The main principle carved in the consortium agreement is that each set of 10 votes (there are 120 votes, and groups choose how to associate) has the right to propose one member to the Governing Board (if the latter rejects the proposal, the gathering has to propose another candidate). We take as one proof of success of the approach the renewal of 40% of members of the Executive Committee at mid-term.

Operational management: the Executive Committee supported by a Scientific Committee and a professional management team
The second difficulty deals with organising the operational management. We chose to delegate all operational matters, from the preparation of the strategy to allocations of funds, to an Executive Committee (ExC) composed of 12 persons and nominated by the Governing Board (see point 1). They are supposed to decide collectively and the rule is that most decisions require a two-thirds majority. De facto nearly all decisions have been taken unanimously, but often after very lively discussions, which the minutes keep track of (they are all available on the management platform, see below).
The Executive Committee was asked to nominate a Scientific Committee to help in the design of priorities and in the selection of projects. We also thought that such a committee would help us review our achievements and identify our limitations. This has proven more challenging than expected especially in a small community with limited differentiation in status between members of both committees.
The ExC was supported by a management team dealing with financial, administrative and legal aspects. Armines who was the recipient of the contract acted as the management team. The very professional team gathered by Armines enabled the Committee to focus on scientific and strategic matters and we consider that it was an important aspect of our ability to function smoothly. This is in part due to their ability to simultaneously manage other NoEs and European projects, which enabled to have access to accumulated resources, to shares resources, problems and lessons with other, and to avoid a painful learning process.

Allocation of funds: 2 types of activities and 3 types of projects
The third difficulty lies in money allocation. Officially NoEs were created on the basis of a yearly allocation associated with researchers’ involvement (here it was far from the offical rule, being a mere 5000 euro per active researcher per year!). We thus refused ex-ante allocations associated to a formal sharing of work via involvement in work packages. We decided that all funds would be allocated on a project / activity base. We recognised that there were both ‘structuration activities’ and ‘research activities’. For both the idea is that money is allocated on the basis of specific time-bounded projects, with a maximum of 18 months for any fund allocation.
Structuration activities concern mostly our shared infrastructures, training and indicators (see respective themes). It also concerns all the integration and dissemination activities. For instance each summer school is a project per se that has to be accepted by the ExC. The same applies for the indicators conferences (Lugano and Olso) or for each annual conference of the network.
We decided from the start that the NoE would support research activities, but this had to be different from ‘business as usual’ (the latter meaning projects that can be supported by one research council or that could find room in EC programmes). Very rapidly we identified three complementary roles: help shaping the long-term research agenda, explore new heterodox ideas, and, in a few cases, considered critical, develop comparative aggregative approaches to demonstrate the value of one new approach. This drove us to implement three types of research projects.
a) Initiation and review actions aim at producing a state of the art and define future research directions on a given topic / issue where there has been difficulty in initiating new research projects in recent years (Innomil and the work on defence is a good example). Most of these go through workshops, seminars, conferences, meetings and end up with a journal special issue or a book, and quite a few with new projects (now mostly outside of Prime).
b) Exploratory projects. They represent a first support to a research proposed by at least 3 members, judged original but very risky by the SC or/and the ExC. Aquameth on universities started like this before developing in a research project. But one must accept the counterpart of it, that is that a number of projects do not succeed as such (our experience has been that the learning acquired has enabled new succesful projects to develop later).
c) Comparative and aggregative projects. The latter correspond to more classical projects but with 3 considerations: originality compared to mainstream lines (on the topic at large), larger geographical inclusion (for comparative or aggregative purposes), importance of investing in the future generation. They need to be in line with our challenges and our anticipation is that we only co-support them until they can go into the normal funding channels (especially FP7 as has been the case for quite a few of the projects, like Venture Fun, Rebaspinoff or national funding as in the case of Nanodistrict).

Learning about implementation and effects: the Characterisation Group
The fourth difficulty for us was that PRIME was a double experimentation: as a support to the development of our community, as a learning process about a new policy instrument. We thus from the start decided to nominate an independent group with its own financial means to oversee the activities of the NoE. The Characterisation Group has followed in detail the development of activities, developed a conceptual framework and conducted two surveys, at the end of the first and fourth years of activity. Their final report as well as their presentations in numerous conferences (Toulouse 2008, Mexico 2008) demonstrate that this was an interesting choice for a network delaing with research and inovation policies. There is a lot to learn both for NoEs in general, ans for us a community about future collective actions.