by AFT President Sandra Feldman
May 1998

Evidence shows voucher
schools don't cost less
-- or work better.

Voucher supporters often claim that private schools are the best bargain around. The figures they cite usually suggest that private schools spend only about half the money per pupil that public schools do. Such comparisons make a big impression on taxpayers. Even some people who have always supported public schools wonder whether a voucher system might not be a good deal--and perhaps on a statewide or national basis.

But people in Cleveland, the home of an experimental voucher program, were brought up short this year. In March, the Cleveland program, which serves 3,000 low-income students, was found to be wildly over budget. There was a 41 percent cost overrun, and the state had to bail out the operation with an extra $2.9 million drawn from money earmarked for the public schools. And next year's projected budget of $8.7 million may have to be adjusted upward another 55 percent.

Officials had plenty of warning, too. An independent audit had already flagged a number of questionable practices and expenses, among them an extraordinary bill for transportation, which resulted from children in the voucher program taking taxis to school. That alone ran up costs an extra $1.5 million.

While these problems may be the result of particular bugs in a particular program, they suggest that running even a small voucher program can't be done on the cheap. If you don't spend money to create good procedures, and then monitor them, which public boards of education are required to do--you're likely to spend it on things like taxi-rides, the way they're doing in Cleveland.

How much would a voucher program that established sound procedures and financial and academic accountability cost? Professor Henry Levin of Stanford University has done an analysis showing that adequate procedures and mechanisms to support a national voucher system (together with the politically inevitable costs of paying for children already in private school) would come to approximately $73 billion per year--on top of the costs of the vouchers themselves.

Comparing Apples to Oranges

But even without Cleveland-style problems, the idea that private schools are half the price of public schools is a lot of hokum.

Private-school tuition pays for a basic program for a particular child. And the tuition quoted is usually for elementary schools, which are considerably less expensive than middle or high schools. Per-pupil expenditure, which is the way public school costs are usually presented, is an average of all the services the school system offers to students at every level.

For example, the special education program, which is federally mandated and very expensive, is, properly, included in quotes of per-pupil expenditures. Private schools are not required to accept special education students, so nearly all of them--some 12 percent of the student population--are educated in the public schools. In addition to special education, the per-pupil expenditure for public schools includes other mandated--and important--services like transportation, English-as-a-second-language, and breakfast and lunch programs. These are costly, and even if a private school offers them, they are not included in the tuition fee usually cited in comparisons with public school per-pupil expenditures.

When private schools offer the same combination of services as public schools, they cost as much, if not more. That's what Professor Levin found when he looked at the Milwaukee voucher program, and a similar analysis confirms that is true in Cleveland--even without the cost overruns.

Disappointing Results

Further, despite what voucher advocates say, there is no solid proof that private schools do a better job academically. People in Cleveland were understandably dissatisfied with their long-neglected public schools. However, the official evaluation of the Cleveland voucher program indicates that its students are not doing any better than comparable public school children. The evidence about the Milwaukee voucher program, which serves a similar group of children, also fails to support the claim that kids in voucher schools are moving ahead of public school children.

Despite what voucher advocates say, vouchers will not give children a better education--or a cheaper one. But we know things that do work. The Cleveland program spent $10 million this year on vouchers for 3,000 children. The same money could have brought two proven reforms strategies into all of Cleveland's elementary schools: Success for All, a reading program, would have cost $4 million; and the $6 million left could have been spent reducing class size in the same schools. Now I'd call that a real bargain--for the kids as well as the taxpayers.