Is there an optimal structure for decentralized provision of roads?

The authors empirically investigate how decentralization affects the efficiency of road provision from the viewpoint of the local goods provider and the road user. The theoretical model: a double cost hidden level effort. They include both user and provider concerns in determining the optimal decentralization level. They find that 100 percent maintenance decentralization produces the most efficiency gains, as quality roads are provided at lower unit costs. There is little justification for central government to be involved in road maintenance. Uniform standards combined with decentralized maintenance remove the incentive to reduce costs and erode most local maintenance efficiency gains. Maintenance is a local activity and should reflect user preferences. Central governments should regulate safety and other network externalities by having a stake in road administration financing, planning, policy setting, safety regulation and other network externalities. Central governments should finance no more than 10 percent of administrative costs. Construction depends on the country. Ensure that contracting procedures are efficient before suggesting decentralization. It is easier for local governments to incorporate user preferences in spending decisions. Similarly, determining where to make investments, deciding how to procure works, and monitoring the quality of construction and maintenance is often done more efficiently locally. The results point to the benefits of decentralized provisions of roads, but many countries contract out maintenance and provision. In that case, it may not matter whether local competitive bidding is carried out by a central or local agency.