Decentralization and local government performance: A comparative approach with application to social policy reform in Vietnam
Few works within the burgeoning literature on decentralization in developing countries systematically compare and explain differential local government performance. This thesis attempts to do so for a specific administrative reform: decentralization of the National Program of Action for Children (NPA) to all local governments in Vietnam over the 1990's.
Decentralization of the NPA evoked diverse responses from local governments. Those with competent administrators committed to social policy reform, in parts of the country with high social cohesion, successfully implemented most reform components. Yet the reform failed completely in the poorest localities, due to their passive fiscal and political position vis-à-vis the center. A reform aimed at promoting dynamic local governments in effect widened the gap between high- and low-performing localities. Low central-level capacity to promote a technically complex reform, combined with political constraints on the reform of intergovernmental fiscal relations, were heavily implicated in this failure.
The thesis holds several implications for decentralization studies. The capacity and proclivity of local governments to perform well under conditions of administrative decentralization can vary dramatically within a country. The pattern of intergovernmental relations and a weak legal framework undermine many administrative reform efforts because they provide disincentives to dynamic local performance and disable local capacities. ‘Demand-driven capacity development’—attempting to restructure the incentives for local governments to perform well—is a promising, though politically difficult and technically complex, solution. Since local government requirements and positions within the regional political economy vary dramatically, decentralization may be most effective if undertaken asymmetrically, to different degrees for different regions.
The methodological approach may be applicable to other decentralization studies. Key characteristics include: (i) systematic modeling of local government performance, along multiple dimensions of reform; (ii) a focus on diverse causal combinations underlying similar reform outcomes; (iii) integration of case-based and quantitative methods; and (iv) interpretation of findings within the context of center-local political economy.