Incentives for Procrastinators

People procrastinate. Psychological research and common intuition suggest that this propensity reflects a time inconsistency: People often put off unpleasant tasks because they pursue immediategratification in a way that their long-run selves do not appreciate. This paper examines the implications of procrastination for the design of temporal incentive schemes, where an agent is rewarded according to when he completes some task. A risk-neutral agent is hired by a risk-neutral principal to complete some task. Delay in completion of the task is costly to the principal, but the agent faces a stochastic cost of completing the task, so that it is efficient to wait when the task cost is high. We assume the principal designs temporal incentive schemes to avoid inefficient delay. We mainlyconsider an environment where the incremental cost of delay is constant over time. When the principal knows the distribution of task costs, she can always design a stationary incentive scheme that achieves first-best efficiency. However, while for time-consistent agents this scheme will reflect the true delay cost, for procrastinators this scheme must punish delay more severely to counteract procrastination. When the agent is privately informed about the task-cost distribution, again a stationary scheme can induce efficiency for time-consistent agents. But for procrastinators, different task-cost distributions imply different propensities for procrastination, so the first-best may be no longer achievable. Furthermore, second-best optimal incentive schemes will typically not be a stationary scheme, but rather a sort of generalized deadline contract: Delay will be punished moderately at first, but after some time it will be punished severely. Such schemes encourage those with little propensity to procrastinate to wait until the cost of completing the task is low, while deadlines assure that those with a severe propensity to procrastinate don't delay too long.