scrubber_stacksUnder what conditions would local or central governments deliver public goods most effectively?  Andrew Chupp (Illinois State University) and H. Spencer Banzhaf (Georgia State University) find that provision costs for pollution abatement, spillover effects, and local conditions determine the degree of policy decentralization.  Their study appears in the June 2012 edition of the Journal of Public Economics.

“The economics of fiscal federalism has identified two book-end departures from first-best provision of a public good. Local governments may respond to local conditions, but ignore inter-jurisdictional spillovers. Alternatively, central governments may internalize spillovers, but impose uniform incentives ignoring local heterogeneity. We provide a simple model that demonstrates that the choice of pricing policy also depends crucially on a third factor, the shape of marginal costs of providing the public good. If marginal costs are convex, then marginal abatement cost elasticities will be higher around the local policies. This increases the deadweight loss of those policies relative to the centralized policy, ceteris paribus. If they are concave, then the opposite is true.

Using a detailed simulation model of the US electricity sector, we then empirically explore these tradeoffs for US air pollution. We find that US states acting in their own interest lose about 31.5% of the potential first-best benefits, whereas the second-best uniform policy loses only 0.2% of benefits. The centralized policy outperforms the state policy for two reasons. First, inter-state spillovers are simply more important that inter-state heterogeneity in this application. Second, because of the convexity of the marginal cost functions, elasticities are much lower over the range relevant to the centralized policy, dampening the distortions.”