A r c h i v e d  I n f o r m a t i o n

Impact of Vouchers on Schooling

• Private schools have little capacity to absorb a substantial number of additional students. According to a California study, less than 1 percent of public school students in that state could expect to find spaces in existing private schools. [Corwin and Dianda, "What Can We Really Expect from Large-Scale Voucher Programs," Phi Delta Kappan, 1993] The study found that most private schools that were interested in participating in a voucher program were already operating at 85 percent or more of their capacity. Although some private schools might be created or expanded, a voucher system would do little or nothing to address the needs for educating the 89 percent of students in public schools (46.5 million students).

• Most private schools are religious in nature, and few are likely to give up their religious mission in order to overcome constitutional barriers to receiving public funds. Religious schools account for 79 percent of all private schools and 85 percent of private school students. A recent survey found that most associations of religious schools believe their schools would not participate in a voucher program if they were required to permit exemptions from religious instruction for students transferring from public schools [Muraskin et al., Barriers, Benefits, and Costs of Using Private Schools to Alleviate Overcrowding in Public Schools: Preliminary Report, 1997]. In addition, government officials should not be expected to choose which private and religious schools merit taxpayer support. For a religiously diverse country like the U.S., this is a road to onerous problems.

• Private schools could select the best and the brightest students. As Representative Bill Goodling (Pennsylvania), then ranking minority member of the Education and Labor Committee, said in 1991, "If you have 500 students in a school and 250 of them are the 'thousand points of light' and decide to go to a school of choice, that leaves 250 fallen angels behind." [Congressional Quarterly, 1991, p. 379] Research shows that private school vouchers do skim more advantaged students -- those whose parents have more education, higher expectations for their children, and higher incomes. This finding holds even for programs where vouchers are restricted to low-income families. For example:
  • In San Antonio, mothers of low-income voucher students were three times more likely to have had some college education than mothers of comparable public school students (55 percent vs. 19 percent). [Martinez, V., et al. (1995). "The Consequences of School Choice: Who Leaves and Who Stays in the Inner City," Social Science Quarterly. September.]

  • In England, researchers found that parents who expressed preferences for schools outside their immediate neighborhood tended to be more highly educated and have more prestigious occupations than those who expressed a preference for their neighborhood school. Moreover, in the most competitive areas, "schools are more likely to choose students than students are to choose schools." [Stearns, School Reform: Lessons from England, 1996]

When a school is failing, providing an escape hatch for a few students will do nothing to improve the quality of education for the majority of students who remain in that school. Indeed, vouchers could hasten the deterioration of the public school system by creating a two-tier educational system in which the motivated, pro-active families and students -- those who have the potential to make the school system better -- would attend private schools and the less involved families and students would attend the public schools. Instead of giving a few students a way out, we need to give all students a way up by improving the quality of all schools.

Impact of Choice And School Type on Student Achievement

• Evaluations of existing voucher programs provide no conclusive evidence of the benefit of these programs for student achievement, while public school choice programs show promise for raising academic proficiency; and

• Differences in student achievement can often be explained by course-taking and high standards.

Descriptions of studies that support these themes follow.

Impact of Private and Public School Choice Programs on Student Achievement

Three separate evaluations of the longest-running publicly-funded voucher program (in Milwaukee) found vastly different results. The first evaluation found voucher students' achievement did not improve significantly from their previous achievement in public schools [Witte et al., Fourth-Year Report: Milwaukee Parental Choice Program, 1994], but a second evaluation of the same data did find evidence that the Milwaukee voucher program had a substantial positive impact on the achievement of students who remained in the program for 3-4 years [Greene, Peterson, and Du, "The Effectiveness of School Choice in Milwaukee: A Secondary Analysis of Data from the Program's Evaluation," 1996]. This second analysis, however, has serious methodological flaws, including the attrition of 85-95 percent of voucher students in the two years in which significant results were found, and a failure to account for student family background and prior achievement. Yet a third analysis [Rouse, "Private School Vouchers and Student Achievement: An Evaluation of the Milwaukee Parental Choice Program," 1997] repeated Greene, et al.'s analysis, and found that voucher recipients did significantly outperform non-enrolled voucher program applicants and a random sample of Milwaukee public school students in math, but did not in reading. These conflicting results provide more of a lesson in the art of statistics than in the effectiveness of voucher programs. The one clear implication of these studies is that the impact of voucher programs on student achievement remains unproven.

In Cleveland, the voucher program has only been in existence since the Fall of 1996, and no systematic research on the impact of the program on student achievement is yet available. The Ohio Department of Education is currently sponsoring an Indiana University evaluation of third-graders in the Cleveland voucher program, and a first-year report is expected in late September of this year. However, a recent, well-publicized Harvard press release on the achievement gains of Cleveland voucher students has been touted as evidence of the benefits of voucher programs. [Harvard University, "Gains in Test Scores in the Cleveland Voucher Program Found," 1997] This analysis, conducted by Paul Peterson and funded by the Olin Foundation, claims that students enrolled in two schools in the Cleveland program have realized "moderately large" gains in reading and "even more substantial" gains in math. However, even this analysis concedes that these academic results are mixed, with language scores declining by 5 percentage points overall, and declining by 19 percent among first graders. More importantly, this analysis is in no way a representation of the effectiveness of the Cleveland voucher program, as it is based on only 15 percent of participating students who are enrolled in uniquely operated, resource-intensive schools, and it suffers from severe methodological problems.⁽³⁾

Lastly, a Hudson Institute study [Weinschrott and Kilgore, "Educational Choice Charitable Trust: An Experiment in School Choice," 1996] of the privately-funded voucher program in Indianapolis found that the voucher students performed "as well as [public school] students in the earlier grades and seem to be doing better in the middle-school grades." However, this appears to be an overly generous characterization of their data, which showed that the voucher students' average test scores were below those of public school students in grade 2, about the same in grades 3 and 6, and higher only in grade 8.

An analysis of 1991 International Assessment of Educational Progress data in 16 industrialized nations provides international evidence showing that private schools do not have significantly higher student achievement than public schools after controlling for student background. [Bishop, "Nerd Harassment, Incentives, School Priorities and Learning," 1996] This study found that, after controlling for family background, independent schools did not have higher math and science achievement (for 13-year-olds) than public schools and religiously-controlled schools had significantly lower achievement levels. In Canada, where there is no constitutional prohibition against public subsidies of religious schools and 20 percent of the schools are religiously controlled, students at religious schools scored 3.9 to 4.5 points lower on science and math tests than public school students.

Some research indicates that public schools of choice show as large a benefit (if not larger) than private schools in producing better student achievement. For example, a recent analysis of data from the National Educational Longitudinal Survey [Gamoran, "Student Achievement in Public Magnet, Public Comprehensive, and Private City High Schools," Educational Evaluation and Policy Analysis, 1996] found that magnet schools were more effective than comprehensive public high schools at raising the proficiency of students in science, reading, and social studies, while secular private schools did not offer any advantage, after controlling for pre-existing differences among students.

Impact of Course-taking and Standards on Student Achievement

A growing body of evidence demonstrates that public school reform efforts like challenging standards and rigorous course-taking can improve achievement for the majority of students who are in the public schools. States and local communities that have set more challenging standards are seeing substantial gains in student achievement. For example:

• Kentucky's comprehensive school reforms continue to result in substantial improvement in school performance. More than 92 percent of Kentucky's schools posted achievement gains in 1995-96. Fifty percent of schools in the state met or exceeded their performance goals. These schools are distributed across all grade levels, throughout every geographic region in the state, and in poor as well as wealthy communities. Elementary schools performed well above expectations, as they exceeded their statewide performance goal by 27 percent. [Kentucky Department of Education Press Release, October 1996]

• In New York City, tougher graduation requirements are spurring thousands more high school students to take and pass college-preparatory mathematics and science courses. The number of Hispanic and black students who passed the science test more than doubled from the previous year [New York Times, 5/9/95]. Entering freshmen at the City University of New York are the best prepared academically in two decades [New York Times, 12/10/95]. Grade schools in the city continue a four-year rise in test scores [New York Times, 6/21/96].

-###-

[ Part 1 ] [ Part 3 ]