Monitoring and Evaluation of Decentralization Reforms in Developing Country Health Sectors

PDF document icon tr-04-29.pdf — PDF document, 688 kB (705,531 bytes)

Author(s): Hutchinson PL, LaFond AK

Year: 2004


The decentralization of responsibilities from the central government to lower levels of government or semi-autonomous institutions has become an increasingly common strategy for improving the performance of health systems in developing countries and ultimately for improving the health status of developing country populations. However, efforts to monitor and evaluate the implementation and functioning of decentralization programs are often hampered by poor or incompatible data, by the absence of sound research designs, and by the sheer scope of the reform, which can encompass all aspects of health system functioning. This work presents a conceptual framework for identifying key areas for evaluation of decentralization programs and the pathways – and potential barriers – by which decentralization can affect health systems. It also identifies ways to evaluate the impact of decentralization in achieving key objectives – improved efficiency, accessibility, equity, community participation and health status. The work outlines the types of data that can be collected and a detailed set of indicators in several broad areas – political, administrative, and fiscal – that can be useful for monitoring and evaluation purposes. Most of the indicators described can be collected from existing data collection techniques, although to date, many of these data are not part of routine data collection in many countries. Numerous data collection tools are also described, as are the types of analyses – including impact evaluations of decentralization – that can be undertaken with that data. A spanish-language version of this technical report, "Monitoreo y Evaluacion de las Reformas de Descentralization en los Sectores de Salud de los Paises en Desarrollo," is also available.

Filed under:
Search Publications
Match against Title and Abstract.
Match Author Names.