Statistics From Cleveland Add Fuel to the Voucher Debate
The politically charged battle over statistics from school voucher programs, fiercely waged last year in Milwaukee, has shifted to Cleveland.
Two studies released this summer paint dramatically different portraits of the Cleveland program, which is the first in the nation to provide publicly financed vouchers that allow poor students to attend religious schools.
Coming as a court challenge to the constitutionality of the program continued and as the Ohio legislature voted to continue its funding, the studies provided fresh ammunition to both sides in the debate.
One analysis, by researchers from three universities, found gains in test scores among students at two of the 55 schools in the program, while a report from the American Federation of Teachers called the program "a cruel hoax."
The AFT report, released last month, was accompanied by a scorching critique of the university professors' study that set off a round of bitter accusations between the two camps.
Education experts cautioned, however, that neither study offered the last word on Cleveland's voucher program.
"People who do this kind of research pretty much have their minds made up already, and they're not likely to produce surprising conclusions," said Richard Elmore, a professor at Harvard University's graduate school of education, who was not involved in either study.
Test Gains Cited
The first study, conducted by Jay P. Greene of the University of Texas at Austin, Paul E. Peterson of Harvard, and Stanford University's William Howell, looked at test scores for voucher students at two private schools. The schools--Hope Central and Hope Ohio City--were created especially for the program by a businessman who is a voucher proponent.
From fall of last year to spring of this year, the researchers found, Hope students in grades K-3 gained an average of 5.5 percentile points on reading tests and 15 points in math. At a June press conference held to release the results, the principal of one of the Hope schools termed the improvements "a miracle."
Mr. Peterson and Mr. Greene were the authors of a report last year that fueled a similar debate over Wisconsin's voucher program for Milwaukee, which does not include religious schools. The report found learning gains among students who had stayed in the program three years.
A separate, state-funded evaluation of that program, however, concluded that it had produced no significant gains in test scores. (See Education Week, Sept. 4, 1996.)
The Cleveland program, which began last year with 1,994 students in grades K-3, is scheduled to expand this fall to 3,000 children, including pupils in 4th grade.
If, that is, the program survives in the courts. In May, a state appeals court said the $5.25 million program violates federal and state bans on government aid to religious institutions. That ruling has been appealed to the Ohio Supreme Court.
State officials petitioned the court for permission to keep the program running while that review continues, and the high court last month granted a stay that will allow them to do so. (See story, this page.)
In looking at the university professors' Cleveland study, the AFT and others say the problem is that testing students from fall to spring exaggerates achievement gains.
"The practice of looking at gains in terms of changes from fall to spring in percentile ranks was once used a lot in evaluating Title I programs," said Robert Linn, an assessment professor at the University of Colorado at Boulder. But, he said, the practice was stopped in assessing the federal program for disadvantaged students because it led to an inflated view of learning gains.
But other researchers say fall-to-spring testing is not out of line--particularly since better data are not yet available. "You've got to start somewhere," Mr. Elmore said.
For their part, the researchers acknowledged that their report was just a start. Over the next couple of years, they plan a more thorough study that will include voucher students from many more schools and that will compare them to a control group.
Both sides also quibble over the professors' interpretation of the students' gains. Mr. Greene and Mr. Peterson said the gains were large, since disadvantaged students tend to lose 1 to 2 percentile points a year in test scores.
But the AFT, citing data from the federal Title I program, said poor students normally improve their test scores from year to year.
The teachers' union also questioned why the researchers had focused on just two schools when 55 private schools accepted voucher students. But Mr. Greene said the Hope schools, which took all comers, served 15 percent of all students in the program and that no other schools had tested students twice.
Union officials also contended that the Hope schools were bankrolling the researchers--a charge the professors flatly deny. The conservative John M. Olin Foundation is their primary source of financial support for the ongoing evaluation, they said.
"I think the facts contradict what the AFT says in their comment on our report," Mr. Greene said.
Access an Issue
As for its analysis, the AFT's biggest criticism of the program is that the students at existing private schools who entered the voucher program had crowded out some public school students who wanted to attend those schools.
Of the 1,994 students in the voucher program, only 664 came from public schools. Another 496--or 25 percent--were students who had already been enrolled in private schools. They took up 56 percent of the 830 seats available to voucher students in the established private schools.
The remaining 834 students were kindergartners. At the time, few of the city's public schools offered full-day kindergarten. More full-day kindergartens are scheduled to be opened this year.
"So much for giving poor parents the same choices as rich parents," said Sandra Feldman, the president of the 940,000-member teachers' union.
But the legislation authorizing the program allowed 50 percent of voucher students to come from private schools. That percentage was later whittled down to 25.
More Research Needed
The union also says that only half the public school students who had gotten a "green light" to obtain vouchers chose to do so.
But Bert L. Holt, who administers the voucher program, said there were other reasons those students did not follow through. Program administrators were unable to find some of the parents who applied, and the program was expanded late in 1996.
"You can't tell parents, 'Your child has a scholarship,' in August of 1996 and school starts two weeks later," Ms. Holt said.
One thing nearly everyone agrees on is that more research is needed. If the voucher students are eventually shown to fare better in the wide range of private schools they attend, no analysis currently under way will likely explain why.
"You can't treat these schools like a black box that students go into," Mr. Elmore said. "Unless you open up the box and look inside you're not producing research that has much relevance to how students learn.''
Vol. 16, Issue 41, Page 9Published in Print: August 6, 1997, as Statistics From Cleveland Add Fuel to the Voucher Debate