More On Vouchers (and on how to shift the debate to the right)

Here's a good example of how the public debate gets pushed steadily rightward. I wrote my column early this week (my editor wanted to take a well-deserved vacation for Thanksgiving, and put the Lack-of-Perspective section together on Wednesday). I got paired up with a conservative arguing that choice absolutely works, positively, lots of studies (but none cited in such a way that you could check them) have shown that scores improve for everybody! That's balance--a conservative ideologue, and a moderate (that's me, in case you were confused) saying hey, the evidence is mixed. Any room for a voice saying vouchers are evil? Not this week. Not this year.

COMPETITION: BOOST OR BANE?
Vouchers Don't Guarantee Success
East Valley Tribune, Dec. 1, 2002

Do school vouchers work? Despite the firmness of opinions on the issue, there have been surprisingly few valid empirical studies. In this case, “few” means about zero -- at least until recently.

Most voucher studies cited in this paper and elsewhere are really opinion pieces masquerading as research. Usually produced, then promoted, by groups with entrenched positions, these “studies” are merely junk science that unsurprisingly support proponents’ preexisting beliefs.

In fairness, it’s tough to study fairly whether vouchers actually improve student achievement, as measured by standardized tests. Students who apply for vouchers may have more motivation to learn than similarly situated peers, or have more involved, committed parents.

Not only would a fair study need to figure out which factors affect achievement, but those we do know about are awfully difficult to gauge. (For example, measure just how committed a parent is, on a ten-point scale.) You also could dispute whether tests accurately measure student achievement, but that’s an argument for another day.

In The Education Gap: Vouchers and Urban Schools, academic researchers William Howell and Paul Peterson got around these “control group” problems by studying, over several years, thousands of students who applied for voucher programs in New York, Dayton, and Washington, D.C. More students applied for each program than could get vouchers, so recipients were chosen by lottery -- and a surprising number of lottery winners declined the voucher. Howell and Peterson thus could study what appear to be comparable groups.

Their study kept track of other characteristics of the students, their schools, their parents, and their environment, confirming the assumption of comparability)

So, do vouchers work? Howell and Peterson’s research says the answer is yes -- and no.

In Dayton and New York, but less so in D.C., black children who used vouchers got significantly better test scores than peers who did not. However, white and Hispanic students using vouchers showed no improvement. For them, vouchers had no effect on achievement. Surprisingly, whether vouchers work seems to depend on race; vouchers work for blacks, but don’t work for Hispanic and white kids.

Frustratingly, Howell and Peterson have no good explanation for the difference. The white and Hispanic children in the study came from low-income families in inner cities with poorly performing public schools. In fact, that’s how those students qualified to apply for vouchers.

The study also ruled out language as a cause for the Hispanic students’ results. While private schools as a rule provide less assistance for non-English-speakers, an analysis of English-proficient Hispanic students also found no gains in achievement from vouchers.

The explanation Howell and Peterson offer is a “differential theory of school choice,” which assumes that the public school systems in the test cities offer black students schools significantly worse than similarly situated white and Hispanic students, or that black parents are better “shoppers” finding better private schools than do non-black parents.

Maybe black students face worse public schools than Hispanic or white students nationally, but that doesn’t explain the difference for the students in the study, all of whom attended the same school districts (and often the same schools) and came from similar low-income families.

As Sara Mead of the Progressive Policy Institute noted in her review of The Education Gap, Howell and Peterson don’t answer the most important question raised by their surprising findings: What’s different for black students in private schools, and can we bring it to public schools -- without basing educational opportunities on race?

The facts about vouchers thus are perplexing and nuanced -- exactly the opposite of the debate about vouchers. School choice works, and yet it doesn’t. So let’s drop the certainty, and may a thousand practical approaches bloom.