« What Easton Would Do If He Ruled Education Research | Main | Research Advisory Board Takes a Hiatus »

On Second Thought: D.C. Voucher Findings Reanalyzed


An independent review out today takes issue with the federal study of the District of Columbia's private school voucher experiment.

Published in March by the Institute of Education Sciences, the study found that after three years, students who nabbed a tuition voucher through the D.C. Opportunity Scholarship Program were doing modestly better in reading, and no better in math, than public school peers who tried for, but failed to win, a scholarship.

You'll recall that interpretations of the findings were all over the map in the news media, with some writers pronouncing the program a success and others seeing them as an indictment of it. The results were reason enough, though, for the Obama administration to suggest allowing program to sunset, and possibly grandfathering in the 1,600 students who now receive the vouchers, worth up to $7,500 a year.

In his review of the federal study, though, Stanford University economist Martin Carnoy contends the researchers were not as nuanced as they could be. For instance, he says, they did not emphasize the fact that most of the gains came among students who were "more academically adept before they were offered the voucher."

But, as the study's lead researcher, Patrick Wolf of the University of Arkansas, points out, he didn't exactly hide that information, either. "What's more important," he asked. "the average effect for all participants or on those who seemed to benefit the most?"

Carnoy has other quibbles as well. For one, there was a lot of school-switching going on among voucher students, and some never used their vouchers at all. (Wolf says those patterns are typical of big-city voucher programs.) For another, much of the academic gain in reading came among the first cohort of students. (Even so, Wolf says, differences between the two cohorts were not statistically significant.)

The review is part of the Think Tank Review Project based at Arizona State University, which was formed to set the record straight on reports issued by think tanks looking to advance their point of view. The project, it has to be said, is financed by the Great Lakes Center for Education Research and Practice, which is made up of the National Education Association and its state affiliates in that region—presumably no fans of private school vouchers.

Extraneous note to the Think Tank Review Project from Wolf: Since when is the federal government a think tank?


First, I am a Vermont school superintendent but truth in packaging says I should note my affiliation with the ASU group. The important part in Ms. Viadero's reprise is that the reading findings were modest and the math gains were non-existent. Math is considered the better school performance indicator as reading scores reflect outside influences to a greater degree. Ms. Viadero should not lightly dismiss the fact that 80% of the control group and 40% of the experimental group disappeared. That's enough to wreck any study. Finally, Mr. Wolf, the study's lead investigator, has made numerous prior pro-voucher advocacy statements. Even if the government is not a think-tank, it makes him a strange choice to head an "independent" evaluation.

I agree with Mr. Mathis' comments regarding the telling of the whole story about the DC voucher experiment. One other fact left out by the coverage of the study is the fact that the capacity for enrolling new students was dramatically shrinking apparently because they have too few seats available. To me, this underscores the issue that this is more about ideology than about helping students. But, read the report yourself. Also, by way of full disclosure, I, too am indirectly associated with the think tank group and believe that having a healthy, robust discussion about improving EVERY student's education is a beneficial exercise.

Comments are now closed for this post.

Follow This Blog


Most Viewed on Education Week



Recent Comments

  • kathleen Gebhardt: I agree with Mr. Mathis' comments regarding the telling of read more
  • William J. Mathis: First, I am a Vermont school superintendent but truth in read more