School Vouchers: The Emerging Track Record
Proponents of private school tuition vouchers make a wide array of claims about their benefits. They claim that competition will spur public school improvement, vouchers will reduce the cost of education, students who get vouchers will show dramatic achievement gains, and vouchers are a success in most industrialized nations. None of this has happened.
Real evidence of how vouchers work now exists. Private school tuition vouchers began in Milwaukee, Wisconsin, beginning in 1990 and were followed by two other voucher plans in Cleveland, Ohio, and in the state of Florida. Private scholarship programs also have a clear track record.
Have vouchers had the impact predicted by some economists, education theorists, and others? Do the results argue for wider experiments or the adoption of broad-based vouchers? Following are some of the results.
Vouchers fail to significantly expand choices for parents.
In the places where vouchers exist, access means a chance in a lottery. One's name is thrown into the hopper. If it is pulled out, the parent gets a chit good for use in a limited number of places. One might be able to use the voucher to pay private school tuition, if the school has space available and there are no other barriers - such as exclusions or preferences based on race, gender, ability, or other factors.
Vouchers are never likely to be widely available because they lack popular and political support. The August 2001 Gallup Poll for Phi Delta Kappan magazine found that "When given the specific choice, 71 percent of the general public would improve and strengthen existing public schools while just 27 percent would opt for vouchers, the alternative most frequently mentioned by public school critics." (http://www.pdkintl.org/kappan/k0109gal.htm)
Wisconsin state law sets the cap for voucher participants at 15,000. And yet, only 10,739 students use them in 2001-02, less than 10 percent of the Milwaukee public schools enrollment. Some schools have declined to accept any voucher-bearing students; most of the rest have some exclusions or preferences based on ability, gender, religion, or race. Enrollment in public schools has increased from 78 percent to 80 percent of the school aged youth, according to a recent report published by voucher advocates.
Less than 5 percent of Cleveland students use vouchers, about 4,195 students in 2001-02. About two-thirds of the Cleveland students who use vouchers never attended public schools. Vouchers in Cleveland are mostly rebates for families who were already sending their children to private schools. According to the Akron Beacon Journal, "rather than bring about a shift from public to private schools, the voucher program merely slowed an exodus from Cleveland's Catholic schools to the city's public schools." (December 14, 1999).
In the Florida "statewide" voucher plan, about 47 students participate in two schools in Pensacola in 2001-02. At least 93 percent of the schools in the state announced they would not accept any voucher students.
Vouchers have failed to improve student achievement significantly or consistently for students who have moved from public to private schools.
Research on the impact vouchers on student achievement is surrounded by enormous controversy, with questions about the motives of those conducting the research, the methods, and the sources of funding. According to Time magazine, "A study of private programs in New York, Washington, and Dayton, Ohio . . . showed a headline-grabbing 6.3% gain in test scores by African-American students who used vouchers. However, one of the research companies [Mathematica] that gathered data for [Paul] Peterson expressed concern about how he used the information, and called his study's findings premature." (Time, 10/9/00)
"Statistically significant" achievement gains for voucher students are negligible. The gains have not been consistent, they have been far below projections, and they give no compelling evidence to justify expanding vouchers.
In 1990, Dr. John Witte of the University of Wisconsin was hired by the Wisconsin Department of Public Instruction to conduct an evaluation of all aspects of the Milwaukee Parental Choice Program. Witte and his colleagues released annual reports during the first five years, before the legislature discontinued funding for the studies. Two other research groups reanalyzed the Witte data. Jay Greene, Paul Peterson, and Jiangtao Du conducted one of the studies; Celia Rouse conducted the other.
Among Witte's conclusions:
- Student achievement in the fourth year of the program, after controlling for prior achievement and demographic variables, was not significantly different for voucher students than for other low-income Milwaukee public school students.
- While voucher and public school students tested came from demographically similar families, "the families of the voucher children were better educated and more interested in their child's education, both before and after entering the program." (John Witte, et al., Fifth Year Report: Milwaukee Parental Choice Program, University of Wisconsin-Madison, 1995; Achievement Effects of the Milwaukee Voucher Program, University of Wisconsin-Madison, 1997)
The study conducted by Jay Greene, Paul Peterson, and Jiangtao Du concluded the Milwaukee voucher students outperformed public school students in math and reading. But the press release and the report itself were not the same. The research team found no statistically significant advantage for private school students in math in the first three years and in reading in the first four years. (The Effectiveness of School Choice in Milwaukee: A Secondary Analysis of Data from the Program's Evaluation, Harvard University, 1996)
Cecilia Rouse of Princeton University conducted a third analysis in an effort to clear up the disagreement between Peterson and Witte. Rouse found a statistically significant and positive effect in math achievement, but not in reading. She concluded: "these are average effects that do not necessarily mean all of the choice schools are 'better' than the Milwaukee public schools." (Cecilia E. Rouse, Private School Vouchers and Student Achievement: An Evaluation of the Milwaukee Parental Choice Program, Princeton University, 1997)
The Ohio Department of Education commissioned an evaluation of Cleveland's voucher program beginning in April 1997. Among the findings of Dr. Kim Metcalf, et al., at Indiana University:
- Voucher students had been achieving at higher levels than their public school peers prior to receiving the vouchers.
- After accounting for prior achievement and demographics, there were no significant differences in third-grade achievement between voucher students and their public school peers at the end of the first year, but significant and positive effects in language and, less clearly, in science, at the end of two.
- Students attending the two newly established private schools were achieving at significantly lower levels by the end of the second year than either their public school or private school peers.
A research group including Jay Greene and Paul Peterson conducted two additional studies of the Cleveland vouchers, including one that found "positive school choice effects in some subject domains among third grade students." A balanced summary of all of the research on Cleveland and Milwaukee is provided in an article entitled Free Market Policies and Public Education: What Is the Cost of Choice? by Kim Metcalf and Polly Tait.
In August 2000, Paul Peterson, et al., released their conclusions from a study of private scholarship programs in New York City, Washington, D.C., and Dayton, Ohio. Among their findings:
- In the three cities taken together, the overall test-score performance of African-American students who switched from public to private schools was higher after the first and second years.
- No statistically significant effects were observed for students from other ethnic groups who switched from public to private schools.
Within a matter of weeks, the company that gathered the data in New York called Peterson's findings premature. David Myers of Mathematica Policy Research noted that “students who were offered scholarships to attend private schools as part of one of the nation’s largest private voucher programs performed about the same on standardized reading and mathematics tests as students who were not offered scholarships.”
Further, the Mathematica analysis showed “no gains for Latino students in any grade, large and statistically significant gains for African American students who are now in 6th grade, and no impacts for African American students in the 3rd, 4th, or 5th grades,” and concluded that “Because gains are so concentrated within this single group, one needs to be very cautious in setting policy based on the overall modest impacts on test scores.” Voucher Claims are Premature in New York City, David Myers.
Three other significant analyses were concluded in 2001.
Kim Metcalf, director of the Indiana Center for Evaluation, conducted an analysis of student achievement in the Cleveland voucher plan covering the years 1998-2000. According to Metcalf, “Students’ academic performance in each of the four areas [reading, language, mathematics, and total achievement] improved significantly from beginning first grade…This was true for all students, independent of their scholarship status and suggests that students benefited significantly from their schooling, whether in private or public schools and whether they were scholarship, applicant/non-recipient, or non-applicant students.”
In other words, students learn in public and private schools. Contrast those findings with the statements by critics, such as Clint Bolick of the Institute for Justice, who describes the Cleveland voucher as “a program designed to rescue economically disadvantaged children from a failing public school system.”
The United States Government Accounting Office released a study in August 2001, conducted for voucher supporter Senator Judd Gregg (R-NH) that looked at both publicly funded vouchers and private scholarship programs. According to the study, “The contract researcher teams for Cleveland and Milwaukee found little or no statistical significant differences in voucher students’ achievement test scores compared to public school students, but other investigators found that voucher students did better in some subject areas tests.”
Virtually all of the glowing reports published by Paul Peterson and others in praise of the student achievement benefits of vouchers have been funded by pro-voucher individuals and organizations, such as the Walton Foundation and Rose and Milton Friedman Foundation. Most of these studies did not fit GAO’s standards for objectivity, including peer review.
In December 2001, The RAND research organization released a report entitled “Rhetoric Versus Reality: What We Know and What We Need to Know About Vouchers and Charter Schools.” The study found that “Long-term effects on academic skills and attainment in both voucher and charter programs are as yet unexamined. Moreover, there is little information that would permit the effectiveness of vouchers and charters to be compared with other, more conventional reforms, such as class-size reduction, professional development, high-stakes accountability, and district-level interventions.”
To some extent, their conclusions beg the question about the impact of vouchers. Pro-voucher advocates take credit for improvements in Milwaukee public schools, including class size reduction and enhance professional development and Florida voucher advocates take credit for improvements in school rated “F” for two consecutive years – with vouchers as a consequence. But, in fact, other school districts and states have had class size reductions and professional development enhancements with good effect – without the threat of vouchers, and school districts and states have had similar positive effects by strong accountability measures. In fact, the accountability legislation in Florida, pre-cursor to the voucher scheme, had the same impact. Schools designated as low-performing soon moved off the list with the right combination of attention and resources.
States with private school vouchers have not provided safeguards to protect against fiscal irregularities or educational deficiencies.
- Milwaukee voucher schools do not have to administer or report test results. The state eliminated all funding for student achievement data collection in 1995.
- About 40 percent of the costs of the Milwaukee vouchers ($11.5 million) represent payments to schools above the tuition costs charged other students. For 21 schools, about one-third of the total, the state pays between 200 and 400 percent of the tuition and fees charged other students.
- In 1999, the Metropolitan Milwaukee Fair Housing Council found that voucher schools violated state laws regarding open admissions and opt-outs for religious practices. At the beginning of the 2000-01 school year, all of those schools remained on the state's approved list.
- In the 1995-96 school year, four of the 18 voucher schools were shut down because of fraud, mismanagement, or negligence.
- Nine Milwaukee voucher schools have no accreditation, were not seeking accreditation, and administered no standardized tests, according to a recent state audit.
- In September 2000, one individual was just days away from receiving about $560,000 from the state to educate 130 students at an "Institute for Holistic Learning." The school director claimed he had to sign the application for some of the parents because they could not themselves, and he had told parents their children could stay at home and not be marked 'absent' because "technically they were still 'present'."
- An independent auditor hired by Ohio found almost $2 million in questionable expenses in the Cleveland voucher program in the first year. Of that, $1.4 million was spent for taxis to transport students - including $419,000 in over-billing by taxi companies charging for absent students.
- Five schools were able to collect about $1 million in vouchers prior to completing their applications process, resulting in schools with serious fire code violations, health hazards, inadequate curricula, and unqualified teachers. Three of the five schools remained in the program as of December 1999.
- The Islamic Academy of Arts and Sciences was allowed to operate for two years in a 110-year old building with no fire alarm or sprinkler system and lead-based paint eight times greater than the level considered safe. Eight of the 12 instructors did not have teaching licenses, and one had been convicted of first-degree murder. In 1999, more than one-half of the students for whom the school received voucher payments did not attend the school or did so for only part of the year.
- The Golden Christian Academy was found to be a "parent-run video school." Students watched recorded lessons given by an on-screen teacher. The video lessons and workbooks are provided by the Pensacola Christian Academy, which features a "faculty of master [video] teachers." The school remained eligible for vouchers, even after discovery of numerous violations of safety codes by the Cleveland Plain Dealer (7/10/99).
- At least 93 percent of the private schools in Florida refuse to accept any voucher students.
- The state assigns public schools a letter grade, A through F, based on students' test scores on the state standardized test. But private school students are not required to take the test, and therefore private schools are not rated. In short, there is no comparable way to discover if a private school is, in fact, better than the public school a student with a voucher leaves.
- AJC Management schools are projected to receive $5 million voucher money under the “McKay Scholarships” for students with disabilities. Most of the schools in the system are located in churches and pay their teachers as little as $10.50 an hour. In their first seven months of operation, the AJC schools have faced allegations of abusing students physically, of providing students with no textbooks or outdated ones, of failing to provide required therapy and counseling, of falsifying applications for state money. Some students have endured seven different teachers in seven months.
- “But that, apparently, is how Florida's new educational marketplace is supposed to work. Invite students to leave their public schools, point them to places that in some cases have been created for the sole purpose of attracting tax dollars, and then let the families deal with the consequences,” St. Petersburg Times, 4/14/2002.
COSTS OF VOUCHERS
Vouchers end up costing taxpayers more - for administration and to pay the costs of students not formerly served in public schools.
By 1998-99, about 6,000 Milwaukee students received vouchers worth about $5,000 each for a total cost of about $29 million. This created a net loss of $22 million to the public schools. ("Tax Funding for Private School Alternatives: The Financial Impact on Milwaukee Public Schools and Taxpayers," Institute for Wisconsin's Future, 1998)
For 2000-02, 3,900 Cleveland students received vouchers worth about $2,250 each at a cost of about $9 million. Additional transportation and administrative costs bring the total up to more than $10 million.
Even private scholarship programs can reduce funds available for public schools, especially when students use such scholarships initially but then return to public schools mid-year. A September 2000 study by Paul Peterson found that about half the students who received private scholarships in Dayton, Ohio, New York City, and Washington, D.C., were back in the public schools by the second year of the program.
Edgewood School District in San Antonio, Texas, provides an illustration of the financial impact of private scholarships. More than 800 students used $4,000 scholarships to attend private and parochial schools - and the Edgewood district lost $5,800 for each student who left. Since students left from a wide array of schools and grades - and given the large number who returned to public schools the loss of $4.8 million caused numerous disruptions and dimunitions in the quality of education for public school students.
A study of the California voucher proposal, Proposition 38 have found that vouchers would cost $3 billion annually by the fourth year - just to pay the costs of students who already attended private schools, regardless of income. (Public Analysis for California Education, University of California at Berkeley and Stanford University, September 2000)
The Michigan voucher proposal would have cost up to $35 million a year to pay for vouchers in the seven school districts deemed eligible because of a high dropout rate. If a school district were determined to qualify for vouchers, the law made no distinction between students in public schools and those in private schools, and approximately 10,500 students attending private schools in those seven districts would have been eligible to receive vouchers.
VOUCHERS AND SCHOOL IMPROVEMENT
The effects of competition on public schools are purely conjectural.
Milwaukee took steps to raise standards for schools and expectations for students before vouchers were established. Alex Molnar of the University of Wisconsin at Milwaukee conducted a study of student achievement among regular public schools, voucher schools, and public schools that reduced class size and enhanced professional development (as part of the Student Achievement Guarantee in Education or SAGE program). Students in the SAGE schools outperformed their peers in private schools and other Milwaukee public schools.
Florida established a mechanism for determining low performing schools before the voucher plan was created. Under the School Improvement and Educational Accountability Act, 33 schools statewide were identified as "critically low achieving." After two years of concerted effort, including redesigning programs and increasing community involvement, all of those 33 schools came off the list.
Public schools throughout the nation are engaged in improvement efforts - all without the threat of vouchers or other competition. For teachers and others who work in the schools to meet the needs of children, school improvement has intrinsic rewards - beyond the financial punishments and rewards used in profit-making enterprises.
The public rejects vouchers and prefers improving public schools.
Many of the polls that show large support for vouchers use loaded terms, such as 'parental choice' and 'opportunity scholarships.' The more specific one is about a real voucher proposal, the more support goes down.
A Public Agenda opinion data research project conducted in the spring of 1999 found that "Most Americans know very little about vouchers, charter schools or for-profit schools . . . Experts and advocates may hold carefully thought-out positions, but the public has barely begun to learn how these proposals work." Public Agenda found that 63 percent of the general public say they know "very little" or "nothing" about school vouchers. Awareness of vouchers was not much higher in Milwaukee and Cleveland where school voucher sites exist.
In September 2000, Phi Delta Kappan magazine found that three-fourths of Americans would rather invest in improving existing private schools rather than paying for vouchers.
Vouchers and Voters
Since 1972, there have been 10 attempts at vouchers and tuition tax credits, and all 10 have been overwhelmingly defeated. In fact, only one of these has been able to garner even 40 percent of the vote.
|Location||Year||Yes %||No %|
Tuition Tax Credits
|Location||Year||Yes %||No %|
|District of Columbia||1981||11%||89%|
IMPROVING PUBLIC SCHOOLS
Some voucher advocates argue that vouchers will prompt public schools to change. Public schools nationwide have been making significant progress over the past two decades although there is much more that can be done.
The National Education Association and its affiliates support direct efforts to improve schools through programs proven to work, including raising expectations for students through high standards - and giving students the help they need to meet those standards, through reducing class size and enhancing teachers' knowledge and skills.
Following are some of the indicators of public school improvement—absent "competition through vouchers."
America's public schools have made significant progress in every area over the past decade.
- The proportion of high school graduates ages 25 to 29 with a bachelor's or higher degree has grown significantly, from 27 percent in 1994 to 32 percent in 1999.
- More U.S. students are getting high scores on their SAT college entrance examinations. In just the last 7 years, the proportion of graduating seniors getting high scores on the Math SAT has increased by 21 percent. (http://www.collegeboard.org/sat/cbsenior/yr2001/2001reports.html)
- Despite the fact that many more students are taking SAT exams than ever before—as more and more students go on to college—the average SAT scores continue to rise. Verbal SAT scores have risen 6 points since 1990, while Math SAT scores have risen 22 points since 1980 and 13 points since 1990. (http://www.collegeboard.org/sat/cbsenior/yr2001/2001reports.html)
- The average national ACT college entrance examination score made solid gains from 1992 to 1997 and has been steady ever since. The average ACT score has not declined in 12 years, even as the number of students taking the examination has increased by nearly 30 percent in the same period. Today, nearly half of college bound seniors who take the ACT examination achieve a score within the range required for admission to most selective and highly selective colleges. (http://www.act.org/news/releases/2001/08-15-01.html)
- Some 66 percent of the nation's public high schools offer Advanced Placement courses, through which students can earn college credit. Just 45 percent of the nation's private high schools offer these advanced courses. (http://www.collegeboard.org/sat/cbsenior/yr2001/2001reports.html)
- Increasing numbers of 12th and 11th grade students are taking Advanced Placement exams, through which they can get college credit. The number of AP exams per 1000 students has surged from 50 in 1984 to 100 in 1990 to 178 in 2000 -- an increase of 356 percent in just 16 years. (http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2001451)
- The proportion of high school dropouts among persons 16 to 24 years of age has declined by over 20 percent since 1980. Over the past few years, the percentage of students completing high school is the highest it has ever been in American history. (http://nces.ed.gov/pubs2002/droppub_2001/)
- The proportion of high school graduates who have gone on to complete a bachelors degree or higher by the age of 29 has risen by 25 percent since 1987. The proportion completing college today is the highest it has ever been. (http://nces.ed.gov/pubs2001/digest/)
- Forty-seven percent of public school teachers have completed at least a master's degree, compared with only 34 percent of private school teachers. Furthermore, 65 percent of public school teachers have at least 10 years of teaching experience, while only 45 percent of private school teachers have that much experience. (http://nces.ed.gov/pubs2001/digest/ch2.html#2)
- Research shows that teachers are much more likely to be stable in public schools than in private schools. Nearly one-quarter (23 percent) of teachers in small private schools turn over every year, while 13 percent of public school teachers and 14 percent of urban public school teachers turn over each year. Thus, public school teachers are more likely to be stable and experienced than are private school teachers. (Ingersoll, R. M., Teacher Turnover, Teacher Shortages, and the Organization of Schools. Center for the Study of Teaching and Policy. January 2001. For more information, including state by state information about progress in America's public schools, see NEA's website—http://www.nea.org/
For more information on what steps states and localities are taking to enhance educational accountability and improve quality and access, see Education Week's Quality Counts 2002 website - http://www.edweek.com/