Peter Zweifel, Hansjörg Lehmann, and Lukas Steinmann
The health care sector is characterized by a great deal of delegation of authority. In particular, the patient as the ultimate consumer of health care services frequently delegates decision making with respect to several dimensions to the physician in charge, such as the timing of the treatment, the procedures to be followed, the provider of some components of care, and the location of treatment. Delegation of authority occurs because the physician has superior knowledge of the health effects these choices have. However, the question then arises of how the physician can be made to choose in the best interest of his or her patient. Clearly, the same problem exists whenever asymmetry of information forces delegation of authority. The generic solution to the problem is principal-agent modelling, whose main result states that the agent may be controlled through the judicious choice of a payment scheme (Holmström, 1979).
Devising an optimal payment scheme typically requires information about the relationship between the likelihood of observing some outcome and the agent’s unobserved effort. In the case of health, the patient as the principal may often rely on subjective measures, while the physician as the agent typically describes outcome in clinical terms. Once there was disagreement about the outcome, however, the likelihood of observing some outcome is not defined. This failure creates demand for what shall be called a complementary agent, whose task could be to provide the necessary information. 1 This line of thought points to models in which a monitoring agent indicates to the principal the type of agent he is dealing with, thus rendering the information interpretable (see e.g. Tirole, 1986). However, upon reflection one notices that in health care, complementary agents that merely provide information hardly exist. For example, governments (which might provide the lacking information) typically establish the payment scheme as well when stepping in as a complementary agent. Accordingly, this chapter revolves around the choice of a complementary agent in role of designer of incentives rather than provider of information in health care.