School expenditures and student achievement: evidence for the United States

Using state-level panel data, this study estimates a simple achievement function in the fixed-effects format to explore further the nexus between school expenditure and student achievement in the United States. Five main points are noted. First, the effect of per-pupil expenditure is positive and carries high statistical significance in some reasonable models. Second, however, the positive estimate is quantitatively modest. Third, the estimates suggest a structural dissimilarity between the models for verbal and mathematics scores on the scholastic assessment test, and the effect of expenditure seems stronger for the latter. Fourth, introduction of state-specific fixed-effects dummies leads to some changes in the pattern of estimates. Fifth, methodologically, despite only minor variations in the variable values over time, the fixed-effects format generates highly significant parameter estimates in most cases. A secondary exploration does not indicate that the effect of expenditures is stronger in low-achievement contexts than in the high-achievement cases.