Abstract
The growth of knowledge literature acknowledges the importance of the institutional context of science for the discovery of knowledge. This chapter argues that in addition to the institutions that connect experts, the institutional context that exists between experts and their subjects importantly influences the relevance of the knowledge that experts discover. When experts are removed from their subject and feedback mechanisms are weak, the type of knowledge discovered by networks of experts has little policy relevance. When experts operate in a network that is closely linked to their subjects, the type of knowledge they discover will have a greater degree of policy relevance. Evidence from P. T. Bauer's work on different networks of experts in the developing world is used to illustrate this theory.