The author examines the many faces of infrastructure decentralization: the costs and benefits, the government structure (constraint or variable?), the"polycentric"approach, and how to make decentralization work (for whom?). He proposes basic principles and guidelines for policy design, for both small projects and large. Broadly, these guidelines are summed up in a few propositions. In all countries, some critical infrastructure is provided through a decentralized political structure. Current trends make that likely to be more true in the future. Decentralization, however defined, in and of itself had no necessary implications for good or evil so far as infrastructure is concerned: its effects depend on the incentives various decisionmakers face. The key to ensuring that these incentives are conducive to"good"decisions (about design, siting, timing, finance, pricing, operation, maintenance, and use of infrastructure) is to ensure that those who made the decisions bear the financial (and political) consequences, as much as possible. Politically, this means that political leaders at all levels should be responsive and responsible to their constituents, and that those constituents are fully informed about the consequences of all decisions. Making politicians bear the consequences of their own mistakes is as close as one can get to a"hard"political budget constraint. Economically, it must be difficult for local residents to shift cost to nonresidents who do not receive benefits and to make local decisionmakers fully responsible to their citizens for the use they make of revenues collected from them (through local taxes), to users of infrastructure (local or otherwise) for the use made of the revenues they contribute (through user charges of various sorts), and to taxpayers in general for the use made of any transfers (or subsidized loans) they receive. Administratively, what such a system requires is a clear set of"framework"laws (on local budgeting, financial reporting, taxation, contracting, dispute settlement, rules to be followed in designing user charges and so on), as well as adequate institutional support for localities to operate in this environment. To the extent that these conditions are not met, the perverse incentives that too often exist because of the structure and finance of the public sector in many countries will probably be exacerbated by the current tendency to decentralize more and more decisions in the public sector.
