Fiscal Decentralization and Government Size

A Survey of Empirical Analyses

Abstract

There are a lot of literature which have analyzed whether governments are either competitive or monopolistic in the federal system. In other words, the hypothesis that the extent of fiscal decentralization and thus the competitive environment surrounding governments could affect a performance of each government and thus a government size has been presented and empirically tested.
We can see that this hypothesis which focuses on the relationship between fiscal decentralization and a government size, which is often called“the fiscal decentralization hypothesis”or“Leviathan hypothesis”, contains the four effects or aspects in it. Here we call those effects as (1) the centralism effect, (2) the fragmentation effect, (3) the consolidation effect, and (4) the generalizatino effect. The centralism effect results from the vertical decentralization between central and local governments. On the other hand, the fragmentation effect is due to the horizontal decentralization among local governments. We could say that any decentralization effect is made up of both effects. The consolidation effect focuses the difference before and after two or more governments are consolidated. The generalization effect means the difference by generalizing a single-purpose government to a multi-purpose government. Both two effects grasp the alteration of a government scale or government scope.
In this paper, we survey the empirical analyses concerning this fiscal decentralization hypothesis and imply an outlook for future analyses. Examining those literature, it could be shown that those have not always obtained common results but rather have presented a lot of points to be discussed. We typically mention the following four points as those, (1) in what government level, from a local to central level, the competitive mechanism among governments actually occurs, (2) how a government size should be defined, (3) how federal grants to state or local governments should be considered in this framework, and (4) how a government form, namely a generalpurpose government or a single-purpose government, could affect the performance of those local governments? Through examining these points, we clarify the common conclusions obtained and the problems left from the analyses.