Risky research projects are, other things being equal, intrinsically harder to monitor than projects that are less risky. It is shown using agency theory that a standard cost benefit analysis, which ignores the agency problem, will introduce a bias towards excessively risky projects, and it will...