Researchers Call for Larger Voucher Experiments

Published July 1, 2002

The lead researcher in the nation’s largest academic study of the effects of vouchers on student achievement, Harvard University professor Paul E. Peterson, argued before a recent Brookings Institution gathering in Washington, DC that there needs to be more “evidence-based reform” of U.S. K-12 education.

Randomized field trials, which permit valid comparisons between virtually identical groups as is done in medical research, could end such long-running education wars as phonics-versus-“whole language,” he said.

Peterson said it was a “good sign this message is finally getting through to the feds.” The federal No Child Left Behind Act calls for the use of educational approaches backed by scientifically valid research. (See “Federal Law Will Require Research-Based Programs,” School Reform News, April 2002.)

A voucher critic at the Brookings forum, Princeton University professor Alan Krueger, agreed with Peterson.

“The Department of Education ought to be more like the FDA (Food and Drug Administration),” he said.

NYC Vouchers Benefit Black Students

The question of the day was what to make of results of the voucher study presented by Peterson and his colleagues in a new book, The Education Gap: Vouchers and Urban Schools (Brookings Press, 2002).

A three-year evaluation showed the test scores of inner-city black students in New York City who received privately funded vouchers to attend private schools were substantially higher than those of comparable students who remained in the public schools. (See “Vouchers Hike Black Student Test Scores,” School Reform News, June 2002.)

The 9 percentile point gain for voucher students amounted to a 0.45 standard deviation, or almost one-half the difference (one full standard deviation) between test scores of blacks and whites nationwide. Thus, the Peterson study raises the possibility that vouchers could completely eliminate the white-black achievement gap if continued over a six-year period.

Peterson and his coauthor, University of Wisconsin-Madison professor William G. Howell, argued these and other results achieved in a randomized field trial—considered the “gold standard” of social research—justify the launching of even more extensive, and better-funded, voucher experiments in major cities with large concentrations of African-Americans.

Krueger did not object to the suggestion of further research but argued the mixed evidence presented so far makes him highly skeptical of the potential for vouchers to “save the children.”

One anomalous finding in the Peterson research is that test scores of voucher students from other ethnic backgrounds, mainly Hispanic, did not differ significantly from their public-school peers.

DC Charters May Affect Vouchers

Meanwhile, in Washington, DC, test scores of voucher-receiving African-American students were 9 percentile points higher than the comparison group of public-school students after two years, but after a third year, there were no significant differences.

Peterson said the Washington results might be due to the influence of public charter schools in the District of Columbia. While the impact of charter schools in New York City is minimal, approximately 16 percent of D.C. students attended charter schools in 2000-01, the highest per-capita charter enrollment in the nation.

In the District, more than one-fifth of the students in the voucher control group wound up in charter schools after one year, thereby possibly confounding the effort to determine the impact of vouchers. Also, only 37 percent of those offered a voucher in Washington, DC made use of it for three years, while in New York City 70 percent did.

Evidently, the competitive jolt vouchers can give a school system is already coming from charter schools in the nation’s capital—but that conclusion is “speculation,” Peterson emphasized.

Milwaukee Voucher Model “Much Better”

Kaleem Caire, former president and CEO of the Black Alliance for Educational Options and now an official of the American Education Reform Council, argued private scholarships such as those in New York and Washington are not the equivalent of public vouchers covering a much broader variety of educational costs. He said he was much more interested in testing a broader impact of publicly funded vouchers like those pioneered in Milwaukee for more than a decade.

“If we have three, four, five, or more Milwaukee-type programs, what effect might they have?” he asked.

Peterson enthusiastically agreed the Milwaukee model would be “much better” for a broadened research study, and added he would lift the 15 percent limit currently imposed on the number of voucher participants. His own vision entails creation of fully integrated voucher models, wherein there would be “alternatives for people of the full income range.” It’s important, he stressed. not to use vouchers simply to perpetuate segregation “along income lines.”

At another point in the discussion, Peterson suggested that by “splitting where you go to school from where you live,” vouchers could be a powerful tool for a more integrated society.

Another panelist, New York University professor Joseph Viteritti, argued the positive results for black students in inner-city New York—coupled with limited funding—made a stronger case for need-based, targeted vouchers than for instituting universal vouchers.

The data in both New York and Washington consistently showed private-school parents were far more satisfied with their children’s schools than were public-school parents. Safety evidently played a part in that: Almost two-thirds of public-school parents said fighting was a major problem in their schools, as compared to 32 percent of private-school parents.


Robert Holland is a senior fellow at the Lexington Institute, a public policy think tank in Arlington, Virginia. His email address is [email protected].