Don’t Believe the Spin: Milwaukee Voucher Program a Success
The teachers' unions and the media can't change the facts, try as they might.
March 30, 2009 - 12:08 am
A new set of empirical studies was recently released on the Milwaukee voucher program, and they’re being spun in a misleading way. The data clearly show that vouchers deliver an improved education to Milwaukee students. But the usual suspects are misleadingly spinning the study as proof that vouchers don’t work, and the media is buying the spin.
The battery of empirical studies, conducted by the School Choice Demonstration Project, examined a broad array of questions. One study found that the Milwaukee program saves major taxpayer money, confirming what we knew from previous studies: private schools are more efficient than the bloated government bureaucracy. Another found that private schools in the voucher program are more racially diverse, also confirming a large body of earlier research unanimously showing that vouchers provide a more racially integrated school environment. Yet another study found that vouchers have not affected home prices.
The two studies that are getting all the attention, though, are the ones on test scores. The first of these studies examined whether voucher students have better academic outcomes than a comparison group of public school students. The students in the comparison group were selected by the researchers because they had similar demographics to those of the voucher students, similar test scores at the start of the study, and came from the same neighborhoods as the voucher students.
The second study examined whether competitive incentives from the voucher program improve outcomes in Milwaukee public schools, as the government school monopoly can no longer take students for granted but must serve them better to keep them. For each individual public school student in a randomly selected sample, it counted how many voucher-participating private schools in his or her grade level were within a normal school-commuting distance, then analyzed whether the presence of more voucher options improved outcomes.
The first study found that voucher students had similar outcomes to public school students in the comparison group. The second study found that more voucher options do improve outcomes for public schools.
Amazingly, that’s being spun as a negative finding for the program. The teachers’ unions are touting only the result of the first study and implying that it shows the program produces no improvement.
It’s bad enough that everyone seems to be ignoring the program’s positive impact on public schools. About four-fifths of the students are still in public schools. Why look only at the results for the voucher students, only one-fifth of the total? If you had a medical treatment that would help four-fifths of all patients suffering from some horrible disease — and what else can you call the present state of our education system but a horrible disease? — that would be considered a fantastic result.
But it gets worse. These results don’t just show that the program improves education for students in public schools. They also indicate that the program improves education for the students who are using vouchers.
Think about it. If the voucher program improves public schools — as the second study finds and the previous research on this question also consistently shows — and the voucher students are doing the same as the public school students, what does that seem to be telling us? That the voucher program did improve outcomes for the students using vouchers! The first study just didn’t see the improvement because vouchers deliver the same academic benefit to participants and to students remaining in public schools.
It’s also worth bearing in mind that the matched-sample method has a lot of drawbacks that may conceal any differences between the groups. A much better approach is the one known as “random assignment,” the gold-standard method that’s used in medical trials. Voucher programs in Milwaukee and elsewhere have frequently been studied with random assignment, and those studies consistently show that vouchers produce better academic results.
Also, in those studies there was no possibility that improvement in public schools would mask the performance of the voucher group being studied, as there was in the new Milwaukee study. Moreover, this is only the first year of results from an ongoing study that will track the students for four years. In the past, it has been quite common for studies of vouchers to find no statistically certain improvement among students using vouchers in the first year, but the improvements consistently emerged as more years of data were collected.
The media and the education elite have made it a long-term habit to spin every piece of news that comes along as a loss for vouchers. But their “see no good, hear no good, speak no good” approach to the evidence can’t change the facts.