DHB: Decentralization and Local Public Goods: How Does Allocation of Decision-Making Authority Affect Provision?


Malgosia Madajewicz
Associate Research Scientist in the Center for Climate Systems Research in the Earth Institute
Regina Dolgoarshinnykh
Associate Research Scientist in the Department of Statistics; Adjunct Assistant Professor of Statistics
Alexander Pfaff
Professor of Public Policy, Economics and Environment, Duke University


Access to services such as sanitation, health care and education remains inadequate for much of the world's population. As much as 20 percent of the world's population lack safe drinking water and sanitation. The objective of this study is to determine under what conditions decentralization of decision-making authority improves access to services. The project focuses on safe drinking water, though the analysis applies to all services and more broadly to other types of organizations. Decentralization and its benefits have been popular topics in academic and policy literatures. The World Bank advises that communities should be involved in all projects designed to assist them, including the provision of services. Evidence regarding the effect of community participation, however, is mainly anecdotal and the record is mixed. Existing literature does not provide a categorization of conditions under which decentralization improves outcomes and data documenting the impact of decentralization in practice is scant.
This project identifies which decisions should be delegated to the community under what conditions in order to maximize social welfare. The study is based on the idea that social dynamics determine when delegation is desirable. The theoretical analysis models the effect of community characteristics, such as the distribution of wealth, social networks and community size, on decisions made by individuals who consume and provide services, i.e. community members and service agencies. These decisions in turn determine access to services and social welfare. The effect of social dynamics on organizational structure has not been modeled formally in the decentralization literature and the study advances modeling techniques.