A new wave of development programs that explicitly use incentives to achieve their aims is under way.They are part of a trend, accelerating in recent years, to disburse development assistance against specific and measurable outputs or outcomes. With a proliferation of new ideas under names such as “payments for performance,” “output-based aid,” and “results based financing,” it is easy to lose sight of basic underlying similarities in these approaches and to miss some significant differences.
This paper proposes a way of classifying and distinguishing the range of incentive programs being debated today, emphasizing two particular dimensions: the agent whose behavior the incentive seeksto change and the specificity of the output or outcome measure. It begins by characterizing a basic incentive arrangement, discussing the range of available contracts and how they appear in development programs, presents a classification of existing incentive programs and illustrates the scheme with examples. The paper concludes by identifying four broad categories that address different problemsand offers some cautionary notes.