... standardized math test scores rose significantly more rapidly for students who used vouchers to attend private schools than for their counterparts in public schools.
Milwaukee's experiment with school vouchers has been a source of heated debate among parents, politicians, and educators, and also among social scientists trying to measure its effects. The program, started in 1990, gives selected children from low-income families taxpayer-funded vouchers to allow them to attend private schools. The structure of the voucher program, and Wisconsin's insistence that statistics on the performance of students in the program be collected annually, have provided rich data with which to help resolve important public policy debates on the effectiveness of vouchers and the quality of private versus public schools.
But attempts to evaluate the data have run up against difficulties in defining an appropriate control group against which to measure the accomplishments of students in the voucher program. A series of studies commissioned by the state of Wisconsin measured the voucher students against Milwaukee public school students generally, and found no significant difference in academic performance. Another academic study, on the other hand, measured the voucher students against public school students who applied for the voucher program but lost out in the admission lottery, and found that the voucher students outperformed their peers.
In Private School Vouchers and Student Achievement: An Evaluation of the Milwaukee Parental Choice Program (NBER Working Paper No. 5964), NBER Faculty Research Fellow Cecilia Rouse attempts to resolve this dispute by comparing the voucher students with both control groups and using various statistical methods to remove biases in the data. Her conclusion: In standardized reading tests, there appears to be no difference between voucher students and their public school peers. But standardized math test scores rose significantly more rapidly for students who used vouchers to attend private schools than for their counterparts in public schools.
This suggests, Rouse concludes, that "providing vouchers to low-income students to attend private schools could help increase the mathematical achievement of those students who participate." But the Milwaukee data does not, she cautions, answer broader questions about whether vouchers improve the quality of education for all low income children.