Funding government services under incomplete information

This thesis studies the optimal funding of government services with distortionary taxes under incomplete information. The first essay “Funding Local Government Services and Interregional Redistribution in a Federation under Incomplete Information” examines the implications of federal government's lack of complete information about local tax bases for the design of optimal interregional redistribution schemes and its impact on the allocation and composition of local expenditures.; In the second essay “Optimal Funding of Public Service Agencies under Incomplete Information”, I adopt the techniques developed in the regulation literature to investigate how the decision rules for the optimal provision of public services, the levels of agency's effort, and public services provided are affected by the presence of asymmetric information, cost observability, the distribution of agency's types, agency's objectives, and the endogeneity of the marginal cost of public funds.; The third essay “Imperfect Tagging and the Design of Conditional Grants for Public Service Agencies” explores the usefulness of acquiring information for the purpose of tagging public service agencies in a model of conditional grants design. Acquiring information for tagging may be interpreted as a preliminary screening process. Groups with different tags are then given distinctly different incentives to control costs of operating the agencies. If there are non-negligible costs associated with the tagging process, it may pay to acquire information only if the acquired signal is informative enough. When the observed signal is positively correlated with the true cost conditions of agencies, the government should optimally be more aggressive (resp. conservative) in distorting the high-cost type's effort decision by giving less (resp. more) incentive for the group of agencies tagged as the low-cost (resp. high-cost) type than in the case without tagging.