Decentralizing the provision of health services : an incomplete contracts approach

Abstract

The author studies the
\n allocation-between a central government and a local
\n authority--of responsibility for planning, financing, and
\n operations for the delivery of health services, in the
\n context of an incomplete contracts model. In this model,
\n inputs are required of both the central government and local
\n authorities but they are unable to write down, and commit
\n to, a complete and binding contract describing the actions
\n both should take. The model is meant to capture the tradeoff
\n between central and local authority in decisions about both
\n financing and the provision of services. Each party provides
\n a specific input--for example, the central government
\n establishes a drug procurement system while the local
\n authority designs and implements an incentive scheme to get
\n doctors to carry out their responsibilities appropriately.
\n The responsibility for delivery of services is identified
\n with the ownership of essential infrastructure, such as the
\n clinic or hospital. The author finds that to maximize the
\n joint surplus of the two public bodies: Ownership of the
\n facility should be given to the party that most values the
\n well-being of local residents. (This way, if ex post
\n bargaining breaks down, each still enjoys some benefits from
\n the other's actions.) Financing authority and
\n responsibility for delivering services should be negatively
\n correlated. Generally it is optimal to allocate tax
\n authority to the party that values the residents'
\n well-being less--in other words, separate spending
\n responsibility (ownership) from financing authority. A
\n heavier financing burden (access to a small and inefficient
\n tax base) has the same incentive effect as asset ownership:
\n It increases the return to effort. If transferring ownership
\n of the physical asset is costly (because the party that
\n builds the asset has an inherent advantage in operating
\n it-that is, there is some human capital embodiment), it may
\n be optimal for the party with the higher construction costs
\n to have planning authority. Somewhat paradoxically, the
\n greater the costs of transferring assets from one party to
\n the other, the more likely that ownership of the facilities
\n and their provision should be separated.