The complexities of modern post-industrial society have forced upon governments, regardless of their political structure or ideology, a large measure of planning, and of active intervention in the operations of the economy and the organization of community services. If such planning is to be effective, it requires an information base, continuing analysis of trends and relationships, and evaluation of results or ‘outcomes’. A question of growing importance is what forms of research organization can best meet these needs. At issue are the relation between basic or theoretical and applied or task-oriented research, the links between data collection, analysis and research, the modes of interaction between research and policy-making, the balancing of immediate relevance and the long-term view, and the safeguarding of objectivity and independence in matters of vital importance to the political process.