This paper analyzes how evaluators' private information and subjective biases affect evaluations in academia. We use evidence from centralized selection exams in Spain, where evaluators are randomly assigned to promotion committees. Candidates are significantly more likely to be promoted when...