Multitasking, Two-part Contracts and Bunching: an Application to Doctors' Tasks and Incentive Contracts
This seminar is hosted by Nigel Rice.
It is well known theoretically that the optimal design of incentive contracts depends on the complementarity and substitutability of tasks performed by agents. However, the empirical evidence is lagging far behind and there remains very little research measuring the extent to which tasks are complements or substitutes. This is because standard methods rely on contracts differing across agents, which they rarely do in practice. We develop a novel and widely applicable test for this, that can be applied to cases where there is no contract variation across agents. The test only requires that the incentive contract is piece-wise linear and that the piece rate of some tasks change over time. It exploits the insensitivity of effort on a particular task to variations in the piece-rate of other tasks for agents who are bunched at the kinks.
We provide an application of the test to the UK Quality and Outcomes Framework, one of the largest pay-for-performance programmes for primary care services in the world, finding that some tasks are complements and that none of them are substitutes. Overall, our results indicate that pay-for-performance schemes in primary care should be successful because increasing the effort exerted on most tasks decreases the marginal cost of effort on other tasks, leading to efficiency gains. The results also have implications for the design of a health care system, suggesting that one based on family doctors rather than specialist doctors would lead to efficiency gains as it would group complementary tasks together.