Access to scientific knowledge is not restricted by patents or industrial secrets. However, it is often difficult to utilize this knowledge in political processes and decision-making. Knowledge brokerage systems have to consider that the path from data to information and further to knowledge is based on processes of transformation, contextualization, interpretation and generalization. The importance of context for the use of knowledge and its high level of innovativeness are main reasons why knowledge brokerage between research and policy-making cannot be resolved by market mechanisms. The required translation process presupposes the selection of certain theories, as well as the concretization and application of scientific knowledge. Many scientists develop the assumption that other communities are also knowledge-driven and that knowledge is the primary factor in formulating and choosing between different policy options. Knowledge brokerage based on normative statements is to a certain extent possible within the framework of "agenda-setting games".