Despite important policy implications associated with the allocation of education resources, evidence on the effectiveness of school inputs remains inconclusive. In part, this is due to endogenous allocation; families sort themselves non-randomly into school districts and school districts allocate money based in order to compensate (or reinforce) differences in child abilities, which leaves estimates of school input effects likely to be biased. Using variation in education expenditures induced by the location of natural resources in Norway we examine the effect of school resources on pupil outcomes. We find that higher school expenditures, triggered by higher revenues from local taxes on hydropower plants, have a significantly positive effect on pupil performance at age 16. The positive IV estimates contrast with the standard cross-sectional estimates that reveal no effects of extra resources.