African-American students in the elementary grades may have the most to gain from scholarship programs that let children switch from public schools to private ones, suggest two new studies co-written by voucher researcher Paul E. Peterson.
But the research also shows that the benefits are far greater in mathematics than in reading. In fact, black students in grades 6-8 who took part in one such program scored somewhat lower in reading than similar students who remained in the public schools.
The studies—to be presented at a school choice conference at Harvard University this week—focus on privately financed scholarship programs serving low-income students in Dayton, Ohio, and the District of Columbia.
The research is the latest in a string of studies in which the Program on Education Policy and Governance at Harvard, which Mr. Peterson directs, has examined so-called privately funded vouchers and found some positive effects. Still, Mr. Peterson said the new studies reveal fresh information.
“We learned something that I did not anticipate at all,” he said, “and that is that the benefits of vouchers are evident particularly for African-American students, and that there’s a question mark as to whether there’s a benefit to shifting to a private school in the middle grades.”
The studies examined test results from about 340 students in Dayton and 810 in the District of Columbia who took the Iowa Tests of Basic Skills both in 1998 and last year. The other authors are William G. Howell, a graduate student in political science at Stanford University, and Patrick J. Wolf, an assistant professor at the Georgetown Public Policy Institute at Georgetown University in Washington.
Who Benefits?
Because both scholarship programs use random lotteries to pick voucher recipients, the researchers say they were able to eliminate selection bias and compare two similar groups of test-takers: students who used the aid, and eligible children who sought a scholarship but didn’t win one.
Among white students, the studies showed almost no differences in test results between the two groups, but they did indicate some differences for black students. (Hispanic and other minority students accounted for very few participants.)
In the Dayton program, African-American scholarship users in grades 2-8 scored, on average, nearly 7 percentile points higher in math than those who did not receive scholarships. About the same gains in math were posted by black voucher users in grades 2-5 in the District of Columbia analysis, which included more grade-level breakdowns.
In neither analysis, however, did African-American participants demonstrate statistically significant improvements in reading. Moreover, the Washington 6th to 8th graders in the program showed insignificant gains in math and even a slight loss in reading.
The results suggest that it’s harder for middle school students to adjust when transferring schools than for younger children, Mr. Howell said. “One of the things we’re trying to get a handle on is who vouchers help most. It may not be the same for everyone,” he said.
But others argue that the findings—which came after students had participated in the programs for less than a year—don’t show that vouchers hold strong promise.
“Even if you take some of the achievement results that Peterson found, they’re very small, or one cancels out the other,” said Janet Bass, a spokeswoman for the American Federation of Teachers, which opposes vouchers. “You need far more robust results to justify turning education over to a private entity without any accountability.”