There is considerable evidence that males are more prone to take risks than females. This difference has implications for rates of promotion in hierarchies where promotion is based on random signals of ability. I explore the promotion consequences of three types of performance standards: gender-blind standards, standards designed to promote agents of equal ability on average, and standards designed to promote equal numbers of both genders. These three objectives lead to different promotion standards, which highlights among other things that the goal of affirmative action is not well defined. Lower promotion standards for females can be necessary to ensure either equal abilities or equal numbers in the promoted populations.
