On the care and feeding of a gift horse: The recurrent cost problem and optimal reduction of recurrent inputs

In contrast to the existing literature, which explains the recurrent cost problem as largely the result of institutional impediments to LDC welfare maximization, this paper analyzes the problem by constructing a series of scenarios in which the less developed country would be led to reduce the flow of recurrent inputs to a development project in order to maximize its own welfare. A distinction is drawn between situations in which the project design is correct ex ante, and those in which it is wrong ex ante. If the project design is correct ex ante, then the recipient country which deviates from the project design does so at a cost to its own welfare unless one or more of the assumptions embodied in the project design have turned out to be false. This paper develops a typology of the possible "surprises" that could lead a host country to reduce recurrent input to a project in order to maximize social welfare. It is proposed that donors use such a typology to help determine the true cause of recurrent input reduction in any given instance. An alternative possible reason for the host country to reduce recurrent inputs to a project is that the project was incorrectly designed in the first place: i.e. the project design could be wrong ex ante. The paper uses a simple model of donor and recipient nation objectives to describe the contractarian relationship between the two nations with respect to the project design and implementation process. According to the model, the donor and recipient countries have a common interest in the output of the development project, but their interests are not identical. In the situation described by the model, both nations have an incentive to agree to a project design that is wrong ex ante. Subsequently the LDC’s reduction of recurrent input can be viewed as its attempt to do the best it can given the inappropriate project with which it is saddled. The analysis is illustrated with a modified Edgeworth-Bowley box diagram. Finally, the paper demonstrates that, like the free-rider problem, the recurrent cost problem can be formulated as a variety of the “prisoner’s dilemma” game form. This observation leads to several policy recommendations for the resolution of the recurrent cost problem which are analogous to the solutions that have been developed for the general free-rider problem as it is characterized by the prisoner’s dilemma. The paper concludes by summarizing the major policy recommendations of the analysis and by discussing some of the possible difficulties that would arise in attempting to implement the recommended policies.