QUANTIFYING SPATIAL MISALLOCATION IN CENTRALLY PROVIDED PUBLIC GOODS

We show how an optimization algorithm can be used to approximately quan-tify the costs to users of spatial misallocation in centrally provided public goods. This method can be employed to evaluate the large programs of public good construction that have been central features of economic plans in many developing countries. We apply these methods to the allocation of post-offices in an administrative block of South India between 1981-1991 and find that more appropriate choices for post office locations could have reduced aggregate costs of travel to citizens in this area by at least 20%.