Reviewing Oates' insights in a repeated game setting

This paper analyzes the optimal assignment of public good policies to layers of a federal system in a repeated game setting. Under a centralized regime, public goods are financed jointly across regions, and a federal legislature decides on the regional quantities. Under a decentralized regime, public goods are financed locally, and governments play a non-cooperative provision game. We find that a centralized (decentralized) regime is more likely to provide the efficient public good policies in case spillovers are small (large). Received: September 2003, Accepted: October 2004 JEL Classification: H11, H41 I wish to thank Clemens Fuest, Anke Kessler, Christoph Lülfesmann, and an anonymous referee for valuable comments on this paper. Financial support by the DFG (SPP 1142) is gratefully acknowledged.