Shaping Incentives through Measurement and Contracts

I develop an agency model in which performance measurement and compensation contracts are jointly used to shape incentives. When an agent has extensive control over firm value, there tend to be many optimal measures — each implying a unique optimal contract with a strikingly simple closed form — that efficiently motivate the same productive action. Because incentives can be embedded in the measure or the contract, any distortions in one are offset by adjusting the other, which can lead to highly nonlinear contracts. When the agent is risk neutral and has unlimited liability, there exist optimal measures that induce linear, convex, concave, bang-bang, piece-wise linear, S-shaped, or standard bonus contracts with floors, ceilings, and hurdles. When the agent has limited liability, standard bonus contracts written on understated measures tend to be jointly optimal. When the agent is risk averse, perfect measures are uniquely optimal whereas imperfect measures tend to distort firm value.