The basic objective of this study has been to analyze whether the allocative and the redistributive functions of the government can be assigned to the regions of a federal state. This concluding chapter tries to give a comprehensive answer to that question and intends to derive the most important policy applications of the analysis in this book. Section 11.1 first turns to the allocative branch of the government, Section 11.2 proceeds with its redistributive function, and Section 11.3 finally draws the policy conclusions.
This section aims to reconsider the conditions ensuring that decentralized government decisions result in an efficient allocation if the individual regions of a federal state are connected by a high degree of interregional mobility. It thereby assumes that governments are not self-serving.
The first important condition is that regions must have a sufficiently flexible instrument set available to achieve an efficient allocation. Mobile firms and households cause crowding costs at their location. Regions must therefore be able to collect direct location-based taxes on mobile firms and households in order to internalize these costs. Without such location-based taxes, it is impossible to ensure the efficient locational pattern across regions, in general. Since, however, marginal crowding costs are generally lower than average costs of providing local public goods and factors, the availability of direct firm and household taxes is not sufficient to finance local public services without distorting locational choices. Regions must have an additional undistortionary revenue source (e.g., a tax on land rents) for balancing their budget-a source that may also help to make an efficient interregional transfer of resources. If this minimum requirement of a complete tax instrument set is not satisfied, then an inefficient allocation must always be expected. This holds even if regions had the correct incentives to choose an efficient allocation.