We have delayed examination of the effects of social science on public policy long enough. This, and the succeeding chapter by Wittrock, now address ways in which the social sciences influence the development of policies in the modern state. Where Majone (chapter 13) drew ideas from the philosophy of science, I adopt an idea from the legal system, the idea of argumentation.

In the first half of the chapter, I examine the influence on policy of three types of research products: data and findings, ideas and criticism, and arguments or briefs for policy action. Whereas the traditional output of a policy study is a report of the first kind, heavy on data, conclusions, statistics, and findings, a review of the sketchy evidence available suggests that in some settings research has greater impact when it becomes part of advocacy for a preferred position.

The second part of the chapter then wrestles with the normative question: what stance should researchers adopt? It confronts the question of whether advocacy has a place in the policy researcher's kit. Policy research is a close relative of social science, and even though it has put on its working clothes and gone out to labour in the offices and chambers of government, it has not relinquished the ‘science’ label: thus, policy sciences. There is something uncomfortable in the thought of abandoning the norms of objectivity that characterize a science and embracing a notion of advocacy more suitable to an interest group or lobbyist.