The cost-effectiveness of five policies for improving student achievement

Comparisons of student achievement effect sizes suggest that systems in which student performance in math and reading is rapidly assessed between 2 and 5 times per week are 4 times as effective as a 10% increase in per pupil expenditure, 6 times as effective as voucher programs, 64 times as effective as charter schools, and 6 times as effective as increased accountability. Achievement gains per dollar from rapid assessment are even greater-193 times the gains that accrue from increasing preexisting patterns of educational expenditures, 2,424 times the gains from vouchers, 23,166 times the gains from charter schools, and 57 times the gains from increased accountability. Two sensitivity analyses suggest that the relative advantage for rapid assessment is not sensitive to the particular parameter estimates.