This article explores the consequences of sabotage for the design of incentive contracts. The possibility of sabotage gives rise to a dynamic concern, similar to the Ratchet effect, which distorts the agents’ incentives. We first show that the mere possibility of sabotage may make it impossible to implement the first-best effort, and we then offer two distinct incentive schemes, fast track and late selection, to circumvent this problem. The present model offers a mechanism through which these two schemes arise in a unified framework.

This article considers a model where the agent is uncertain about his innate ability and instead makes an inference from others’ (namely, the principal’s) perception, as often emphasized in the psychology literature. When the principal has superior knowledge about the agent’s productivity than the agent himself, the principal has an incentive to use promotions strategically to boost the agent’s self‐confidence. Within this framework the optimal promotion policy depends not only on the agent’s current expected ability type but also on the history of his previous job assignments. We use this fact to explain why we rarely observe demotions in organizations.

This paper provides a general analysis of signaling under double‐crossing preferences with a continuum of types. There are natural economic environments where the indifference curves of two types cross twice, such that the celebrated single‐crossing property fails to hold. Equilibrium exhibits a threshold type below which types choose actions that are fully revealing and above which they pool in a pairwise fashion, with a gap separating the actions chosen by these two sets of types. The resulting signaling action is quasi‐concave in type. We also provide an algorithm to establish equilibrium existence by construction.