Policy makers and managers often turn to experts when in need of information, but we should expect experts to be systematically biased. This is because the decision to research a question implies a belief that research will be fruitful. If priors about the impact of one’s work are correct on average, then those who choose to research a question are optimistic about the quality of their work. The bias varies predictably with attributes of the question being studied. This fact has implications for a variety of mechanism design applications and yields predictions in accordance with a large literature in psychology. This paper was accepted by Teck-Hua Ho, behavioral economics.