This book combines political-economic, sociological and historical approaches to provide a coherent framework for analysing the changing relationship between politics and science in the United States. Fundamental to this relationship are problems of delegation, especially the integrity and productivity of sponsored research: politicians must see that research is conducted with integrity and productivity, and scientists must be able to show it. A science policy regime changes when solutions to these problems change. After World War II, the 'social contract for science' assumed that the integrity and productivity of research were automatic and, despite many challenges, that contract endured for four decades. However in the 1980s, as rich empirical studies show, cases of misconduct in science and flagging economic performance broke the trust between politics and science. New 'boundary organizations', in which scientists and nonscientists collaborate to assure the integrity and productivity of research, were created to mend the relationship.