Introduction

In the modern so-called ‘knowledge society’, the state is expected to act rationally in the pursuit of its tasks (Voßkuhle, 2008). This implies, among other things, that public policy decisions must be based on the best available knowledge. What is needed is both descriptive knowledge of given situations, including statistical data, and theoretical knowledge of causal relations – knowledge needed to identify the causes of a problem and to guide the choice of policy measures by pointing out their likely effects. The historical process of rationalisation undergone by Western civilisation has included both the ‘scientification’ of knowledge and the ascent of bureaucratic organisation. The highly personalised, experience-based knowledge of the practitioner is no longer deemed sufficient. Whether in agriculture, education, medicine or governing, knowledge produced according to the standards of scientific methodology and meeting the criteria of reliability, reproducibility and validity is what is now asked for. The tendency towards formal organisation, expressed most clearly in Max Weber's concept of bureaucracy, has not only affected the management of public affairs, but also the production of scientific knowledge and the way it feeds into the policy process. In a classical public bureaucracy, bureaucrats are expected to possess the knowledge required for the effective performance of their duties. There has therefore been a tendency to provide for scientific counsel within the political-administrative system, whether in the form of research institutes attached to one of the federal ministries, or in the form of the Wissenschaftlicher Dienst (scientific service) assisting individual members of parliament. But in developed societies the in-house provision of scientific knowledge is not sufficient to make rational policy decisions, and must be supplemented by external inputs.

Policy analysis

The function of policy analysis for today's policy-making and the way policy analysis is produced and delivered reflects the tendency towards ‘scientification’ of knowledge as well as towards formal organisation. Quasi by definition, policy analysis is a scientific activity; it is produced by and delivered to policy-makers by organisations and in an organised fashion. The formalisation of providing knowledge to political decision-makers is intended to safeguard the reliable, ‘scientific’ character of the knowledge brought to bear on policy, and to prevent the collusion with particularistic interests that easily takes place where elite interaction and informal advice shape policy decisions.