Task-specific effort costs and the trade-off between risk and efficiency

This paper considers a principal-agent model in which an agent chooses the level of effort on two tasks that determine two separate outputs. The level of effort is private information to the agent. We examine the relationship between risk and incentives when the agent has a preference towards one of the two tasks. This is modelled by the assumption that the cost of effort is task specific. We show that when outputs are not perfectly measurable, even if the measurement errors have the same variance, the principal has to set different marginal incentives. This is because the agent tries to spread the risk, created by imperfect output measurement, by distributing more evenly his effort between the two tasks, thereby moving away from an efficient allocation of effort. If the two outputs are measured with different precision, it can be optimal for the principal to set higher marginal incentives on the output measured with less precision. Hence the standard result of the negative correlation between risk and incentives does not necessarily hold.