The Color of Vouchers
Do school vouchers help poor black children? This question has recently sparked an academic controversy over the results of a 2000 study suggesting that it did. Based on research on a private voucher program in New York City, Harvard professor Paul Peterson reported that black voucher students scored six percentile points higher than their public counterparts.[1]
Critics immediately questioned his results. Not only were gains were limited to one ethnicity at one grade level, they argued, but the methodology was dubious: only students with a black mother were considered black. Two Princeton economists named Alan Krueger and Pei Zhu recalculated Peterson’s results using a more encompassing definition of black. Krueger and Zhu defined as black any child who had a black parent, mother or father. Adding these students to the black sample made any academic gains statistically insignificant.
It is important that these criticisms be taken with a grain of salt. Race is not a mathematically precise state and reasonable people often differ on appropriate definitions. For example, it is only in the American experience that individuals with less than a majority of African heritage are considered “black.” Globally this strict classification of black is regarded as “extremely unusual."[2] Peterson’s black-mother criterion is therefore not unwarranted and Krueger’s expansive definition of race is by no means the one true way.
Regardless of what one thinks about the merits of the methodological challenge to Dr. Peterson’s work, it does not warrant the editorializing of New York Times education columnist Michael Winerip, who stated, “It is scary how many prominent thinkers were ready to make new policy from a single study…” This anti-voucher rhetoric ignores the facts (unreported in Mr. Winerip’s article) that five random-assignment voucher studies have found beneficial effects from vouchers. Even Krueger and Zhu’s study found that students benefited from receiving vouchers just not to the level of statistical significance found by Peterson.[3]
In fact, the strong reaction to a study finding a positive, but not statistically significant, voucher effect is strong confirmation of the weight of voucher research. After all, if the best empirical argument you can make against vouchers is that, under certain racial definitions, they no longer benefit students in a “statistically significant” way, then the balance of the evidence must be on the pro-voucher side. To this day, no study has shown that voucher programs hinder a students’ academic performance while numerous ones have.
Regardless of the current methodological controversy, voucher policy is not “new policy from a single study.” Vouchers and their supporters existed long before researchers began studying their effects. The ideas that initially propelled voucher policy were moral and economic. The idea that children, regardless of income, should have educational opportunity, played, and continues to play, a very large role in voucher policy and that is something no empirical study is going to change.
Notes
[1] Michael Winerip, “School voucher study under scrutiny,” Dayton Daily News, 14 May 2003.
[2] F. James Davis, “Who is Black? One Nation’s Definition,” PBS Online, 2000. Available at: http://www.pbs.org/wgbh/pages/frontline/shows/jefferson/mixed/onedrop.html.
[3] Jay P. Green, “An Unfair Grade for Vouchers,” Wall Street Journal, 16 May 2003.
Danielle Mendola is a former research intern with The Buckeye Institute. Joshua C. Hall is a senior fellow with The Buckeye Institute.
