"the model yields the counter-intuitive prescription that punishing non-culpable members of the team for participating in a fraudulent project is the most promising tool for increasing the fraction of research that is honest." This study is worth a read : papers.ssrn.com/sol3/papers....