This paper investigates the developments of research councils in Austria, Norway and Switzerland, and analyses their responsiveness to government and science. By use of various data sources and indicators, our results suggest that councils are sometimes more responsive to the beneficiaries' interests (science) than to policy-makers, but are more responsive to government policies when under stronger government control. Patterns of responsiveness tend to be stable through time despite important changes in national contexts and the changing roles and organisation of research councils. These findings have interesting theoretical consequences about responsiveness and independence ESEARCH COUNCILS have been the preferred way of institutionalising 'the second stream' of research funding in most industrialised countries. Both researchers and research institutions are perceiving council funding for research projects allocated through research grants, programmes and projects as increasingly important (Geuna, 2001;Geuna and Martin, 2003). Instruments and criteria are co-decisive for the selection of research topics, for researchers' adaptive strategies and their ability to perform high quality research (Laudel, 2006), but also because general government appropriations often are tied up with the coverage of running costs, leaving the institutions with little freedom. Funding from intermediaries such as research councils might thus be the only way to make possible new research initiatives and thereby scientific development.Research councils are in many ways semiindependent agencies operating at arm's length from government and can often be closely linked to the research community. Their loyalty can thus be bidirectional, and the councils might have to balance between the interests of policy-makers and scientists. Our main concern is to analyse how research councils act as intermediaries influenced by national research policies on one hand and the research community on the other. By comparing research councils in Austria, Norway and Switzerland, we will shed light on the importance of the councils' embeddedness in national contexts and history for how they play their roles.We first introduce the research questions we are posing. Then we discuss the theoretical framework underlying our understanding of research councils' actions and responses to policy, before we lay out the chosen analytical model and methodology. We then present the councils and give empirical evidence on the developments along the selected variables. Finally, we discuss the findings.
Research questionsResearch councils play a crucial role in scientific development and in science policy in most industrial countries and are often assigned the task of allocating R Stig Slipersaeter is NIFU STEP,