Did the ''tax revolt'' reduce school performance?

This paper uses detailed school-level data from 49 states to analyze the effects of tax-revolt era property tax limitations on school services. I find that limitations are associated with larger student-teacher ratios and lower cost-of-living adjusted starting teacher salaries, all else equal. These results are robust to modelling the results as endogenous and using a variety of measures for whether the limitation binds at particular schools. However, I find no such evidence that schools subject to limitations have reduced their administrative costs. Furthermore, I find that limitations are associated with lower student performance on mathematics, science, social studies and reading examinations, all else equal. (C) 1997 Elsevier Science S.A.