Tierney misrepresented study on Milwaukee school vouchersMarch 7, 2006 1:00 PM EST ››› SIMON MALOY
New York Times columnist John Tierney, in his March 7 column (subscription required), misrepresented the findings of a study conducted by Harvard researcher Rajashri Chakrabarti on school vouchers in Milwaukee, claiming that Chakrabarti's study showed "that as the voucher program expanded in Milwaukee, there was a marked improvement in test scores at the public schools most threatened by the program." In fact, Chakrabarti's 2005 study, which compared school voucher programs in Milwaukee and Florida, questioned whether the Milwaukee voucher program actually had an effect on public schools.
School voucher programs allow parents to withdraw their children from public schools and enroll them in private schools with the help of vouchers for a set amount of public funding, money that is then transferred to the private school. Such programs are usually promoted under the concept of "school choice."
From Tierney's March 7 New York Times column:
In fact, the students in public schools have benefited from the competition. Two studies by Harvard researchers, one by Caroline Hoxby and another by Rajashri Chakrabarti, have shown that as the voucher program expanded in Milwaukee, there was a marked improvement in test scores at the public schools most threatened by the program (the ones with large numbers of low-income students eligible for the vouchers).
The competition spurred the public system to shift power from the central administration to individual schools, allowing councils of parents and teachers to decide who should teach there, instead of forcing the schools to accept incompetent teachers just because they had seniority.
Chakrabarti's study, however, found that the Milwaukee program was deficient to voucher programs in Florida, and that "the results in Milwaukee are mixed" in terms of improving performance in nearby public schools. Chakrabarti concluded:
This study contributes in this direction by comparing the effects of two U.S. voucher programs -- Florida and Milwaukee -- that differ fundamentally in their designs. The Florida program is a "threat of voucher" program that first threatens the failing schools with vouchers and vouchers are introduced only if they fail to meet a certain government designated quality cutoff. The Milwaukee program, on the other hand, is a "voucher shock" program with a sudden government announcement that all low income public school students would be eligible for vouchers. In the context of an equilibrium theory of public school and household behavior, this paper argues that the Florida-type program should bring about an unambiguous improvement in public school performance and this improvement should exceed the improvement (if any) in the Milwaukee-type program.
Moreover, the findings of Harvard economist Caroline Hoxby regarding the Milwaukee voucher program have garnered significant criticism. As Media Matters for America previously noted. Helen F. Ladd, Duke University professor of economics and professor of public policy, has criticized the findings of Hoxby's studies -- one of which was a study of school voucher programs in Milwaukee, Arizona, and Michigan. Ladd claimed that Hoxby's conclusion that competition created by school voucher programs "significantly raised the productivity of the traditional public schools" in Milwaukee cannot be accepted as definitive.
According to Ladd:
In particular, she presents evidence that the threat to public schools in Milwaukee of losing students to voucher schools and the threat to public schools in Arizona and Michigan of losing students to nearby charter schools significantly raised the productivity of the traditional public schools in that city and those states. If she is correct, these results are potentially very important for education policy makers.
Without further corroboration from other researchers, however, it is premature to accept Hoxby's results as definitive. While her research is extremely sophisticated in many ways, it is flawed in that the unit of analysis is the school rather than the individual student.