Decentralization Task Force Meeting, New York City 2006

205kb pdf
Maria Caridad Araujo,
Peter Frederick Lanjouw,
Berk Ozler,
Francisco H.G. Ferreira

Despite numerous evaluations of Social Investment Funds in various countries around the world, quantitative evidence on the local determinants of project choice remains scant. In this paper, exploiting a unique combination of data sets from Ecuador, we test a simple model of project choice in the presence of local inequalities in influence. Consistent with the predictions of our model, we find that communities with higher levels of estimated (consumption) inequality are less likely to receive any projects from Ecuador’s Social Investment Funds (FISE) and, conditional on receiving a project, less likely to have projects that provide excludable goods for the poor, such as latrines. Controlling for inequality, poorer villages are more likely to receive excludable pro-poor projects and projects that require no counterpart funding. The results are sensitive to the measure of inequality used in the empirical analysis, and are strongest for the income share of the elite. Finally, we find that the poverty headcount, population size, and the level of political support for the incumbent party at the local level are also important determinants of project choice. The findings indicate that information on local inequality can be of potential value to the design of decentralized mechanisms of project delivery.

57kb pdf
Dilip Mookherjee,
Pranab K. Bardhan

147kb pdf
Daniel Treisman,
Hongbin Cai

Since Justice Brandeis’ 1932 remark that in a federal system states can serve as “laboratories” of democracy, a widespread belief has held that political decentralization stimulates policy experimentation. We reexamine the political economy underlying this claim, using a model of retrospective voting and contrasting centralized and decentralized institutions. Although under some conditions (when experiments are politically promising) decentralization can result in greater experimentation, several effects combine to lead, under most conditions, to both more frequent and better organized policy experimentation under centralized government.

287kb pdf
Ali Cheema,
Shandana Khan Mohmand

This paper uses household survey data from Punjab Pakistan to empirically analyze who is gaining and who is losing as a result of recent reforms that have instituted electoral decentralization and expenditure devolution. The identification strategy relies on exogenous kinship or caste identities and historic settlement patterns of households to estimate the change in the post-reform bias of service provision outcomes between different household types. We find that electoral and expenditure decentralization in Pakistan has had the positive effect that households in previously underprovided villages have seen a substantial increase in the provision of essential services. However, the results also show that post-reform targeting of provision between households is conditional upon political geography and membership to dominant village level patron-client factions. The results show that being a demographic majority does not always count in a decentralized structure and that there are incentives in the system that may lead politicians to discount groups that are in a demographic majority.

290kb pdf
Fabio Sanchez Torres,
Jean-Paul Faguet

The effects of decentralization on public sector outputs is much debated but little agreed upon. This paper compares the remarkable case of Bolivia with the more complex case of Colombia to explore decentralization’s effects on public education outcomes. In Colombia, decentralization of education finance improved enrollment rates in public schools. In Bolivia, decentralization made government more responsive by re-directing public investment to areas of greatest need. In both countries, investment shifted from infrastructure to primary social services. In both, the behavior of the smallest, poorest, most rural municipalities was what drove these changes.

181kb pdf
Stuti Khemani

Theoretical and empirical analysis suggests that federations are prone to fiscal indiscipline, because of a classic “common-pool” problem of distributive politics—when decision-making power over government spending is distributed across multiple agents but financed out of generalized taxation. An institutional solution typically adopted by many countries is the delegation of sub-national debt oversight to the national political executive, which is assumed to have stronger incentives for fiscal discipline. Can such delegation successfully address political distortions to fiscal discipline? This paper provides evidence on this question from India, arguing that such delegation has limited success because the political incentives of the national government are such that it is unwilling or unable to enforce hard budget constraints on sub-nationals. In light of this evidence, the paper explores the role of an alternate form of delegation, to an independent fiscal agency, of monitoring and oversight of consolidated government debt, as a solution to the common- pool problem, again providing evidence from India to support its arguments.

255kb pdf