I propose that some major academic organization such as the American Economic Association randomly and secretly choose a few members and request that they attempt to get fraudulent work accepted into the highest ranked journals they can. They reveal the fraud as soon as an article is accepted. This procedure would give us some idea how of easy it is to engage in fraud, and give journals additional incentives to search for it. For some academic disciplines the incentives to engage in fraud seem similar to that with illegal performance enhancing drugs and professional sports, and I wonder if the outcomes are similar.