No evidence that choice students outperform public school students, Mary Burke says

One way Democratic gubernatorial candidate Mary Burke contrasts herself with Gov. Scott Walker is with the state's school voucher program, which uses public money to pay for lower-income children to attend private schools.

With his 2013-’15 state budget, Walker expanded the voucher program, also known as school choice, by opening it to a limited number of students outside of Milwaukee and Racine.

Burke, a Madison School Board member, opposes the expansion. Speaking to the Milwaukee Rotary Club on May 6, 2014, she elaborated on why, saying:

"What I want to see is initiatives and programs that are actually based on research and success that’s going to improve education, instead of saying that, well, our answer is just to roll out a voucher program that actually has no research that shows that it’s going to improve student learning."

This debate has been engaged many times before. So let's see what we know about research into the performance of public vs. voucher school students.

Vouchers and their expansion

Vouchers in Wisconsin date to 1989, when Republican Gov. Tommy Thompson and a Democratic-majority Legislature created the Milwaukee Parental Choice Program, the nation's first urban school voucher program of its kind in the country.

Under the original law, the number of vouchers was limited to 1 percent of the enrollment in Milwaukee Public Schools. And only non-sectarian private schools in the city were eligible to take choice students. Choice started with about 300 students in the 1990-’91 school year.

The program has since been expanded several times -- more than 25,000 students now participate -- with the two most recent expansions occurring under Walker:

Expansion of school vouchers in Wisconsin

1995: Religious schools can participate; voucher limit raised to 15 percent of Milwaukee Public Schools enrollment

2005: Voucher limit raised to 22,500 MPS students

2011: Voucher limit eliminated; income eligibility raised; private schools outside Milwaukee can accept Milwaukee choice students; Racine-area students and schools are made eligible

2013: On limited basis, vouchers available to students outside of Milwaukee and Racine

Now let’s look at data on the performance of public vs voucher students.

Test scores and Arkansas research

In assessing Burke’s claim, it’s important to note there are limited ways to compare the performance of voucher vs public school students. Wisconsin does not require voucher schools to publicly disclose measures such as retention rates, graduation rates or college-entrance exam scores.

As we reported in an April 2013 article on competing claims about voucher schools, and as we found in doing additional research for this article, there are two primary measures of vouchers in Wisconsin:

1. Annual statewide test scores reported in 2011 through 2014; and

2. A study by researchers based at the University of Arkansas that tracked the test scores of particular choice and Milwaukee Public Schools students over five years, ending in 2011.

1. Here’s how public and voucher students compared on statewide test scores:

• 2011: In first direct comparison involving statewide tests, students in Milwaukee Public Schools performed better than or about the same as voucher students in reading and math

• 2012 and 2013: Students attending public schools were more proficient, as a group, than their private school-choice counterparts in Milwaukee and Racine in reading and math

• 2014: On average, MPS students performed higher than Milwaukee choice students in reading and math. Racine public school students also scored higher on math and reading proficiency than Racine voucher students.

An April 2014 report from the Public Policy Forum, a nonpartisan research group in Milwaukee, shows that from 2010-’11 through 2012-’13, Milwaukee Public Schools students outperformed voucher students at every grade level (third through 10th grades) on the statewide reading and math tests.

So, overall, four years worth of publicly available, annual state test scores give a clear edge to public school students over voucher students.

Featured Fact-check

Terry Gorman

stated on April 24, 2012 in a House committee hearing

"When these [undocumented] students graduate from college, they're still illegal aliens. They cannot get a job."

2. Arkansas research

The Arkansas research, meanwhile, took a different approach. Its "longitudinal" study tracked representative samples of 2,727 Milwaukee choice students and 2,727 Milwaukee Public Schools for five years, from the 2006-’07 through 2010-’11 school years.

Key findings:

• Growth in math achievement was similar in all years, while voucher students made gains in reading in 2010-'11 that were higher than students in the Milwaukee Public Schools sample.

• Milwaukee students enrolling in a private high school in the choice program "increases the likelihood of a student graduating from high school, enrolling in a four-year college and persisting in college by 4 to 7 percentage points."

It’s important to note that the reading advantage was for just one year, and the Arkansas researchers cautioned that some evidence suggested it was due to the introduction of a new state testing requirement for private choice schools.

University of Arkansas education reform professor Patrick Wolf told us there's no way to know whether the reading improvement was due to the voucher schools themselves or other factors, such as teachers preparing their students more for the annual test because 2011 was the first year that voucher schools had to publicly disclose their test scores.

On the high school finding, a majority of students who were enrolled in private choice schools in the ninth grade were no longer in a choice school by the time they reached 12th grade. The results, therefore, should be interpreted as the effect of "exposure" to choice schools rather than long-term persistence in that sector, the Arkansas study found.

Wolf said that while there is no overall evidence that voucher schools improved learning in Milwaukee, for Burke to say there is no evidence "is a bit of a stretch," given what the Arkansas research found.

Before we close, we’ll note two other studies that indicate voucher schools have an indirect impact on student learning.

One study was done in 2003 by an economics professor then at Harvard University and one was done in 2008 by an economist with the Federal Reserve Bank of New York. The studies said competition from voucher schools were linked with improved performance of students in Milwaukee Public Schools.

For example, the 2003 study found that test scores in math, science, social studies, language and reading increased more in MPS schools where at least two-thirds of students were eligible for vouchers.

Our rating

Burke said the Wisconsin voucher program "has no research that shows that it’s going to improve student learning."

Public school students consistently outperformed voucher students on statewide test scores from 2011 to 2014. The one study held up by choice advocates shows only limited instances in which voucher students performed better, and the study couldn't determine whether the voucher schools themselves or other factors were the reason.

Burke goes too far in saying there is no evidence that vouchers improve student learning, but the evidence is thin. We rate her statement Mostly True.

To comment on this article, go to the Milwaukee Journal Sentinel website.