The large variation in revenues among Norwegian local governments can partly be explained by revenues collected from hydropower production. This revenue variation, combined with good data availability, can be used to extend the literature on the resource curse in two directions. First, to ensure that there is no problem of endogeneity in the analysis we obtain a purely exogenous measure of local revenue by instrumenting the variation in hydropower revenue, and thus total revenue, by topology, average precipitation and meters of river in steep terrain. Second, using data for revenue derived from hydropower production in Norwegian local governments we test the 'Rentier State' hypothesis; that revenue derived from natural resources should harm efficiency more than revenue derived from other sources such as taxation. Although we do find that higher local government revenue reduces the efficiency in production of public goods, we do not find that this effect is stronger for natural resource revenue than for other revenue.

No abstract