But broad political support for No Child Left Behind proved short-lived. Teachers unions opposed the reforms from the left, while libertarians and states-rights conservatives denounced them from the right. When Republicans took control of more governor’s mansions and state legislatures in the 2000s, they expanded vouchers to an unprecedented degree. Three of the largest programs sprang up in Indiana, Louisiana and Ohio, which collectively enroll more than a third of the 178,000 voucher students nationwide.
Most of the new programs heeded Mr. Friedman’s original call for the government to enforce “minimum standards” by requiring private schools that accept vouchers to administer standardized state tests. Researchers have used this data to compare voucher students with similar children who took the same tests in public school. Many of the results were released over the last 18 months, while Donald J. Trump was advocating school choice on the campaign trail.
The first results came in late 2015. Researchers examined an Indiana voucher program that had quickly grown to serve tens of thousands of students under Mike Pence, then the state’s governor. “In mathematics,” they found, “voucher students who transfer to private schools experienced significant losses in achievement.” They also saw no improvement in reading.
The next results came a few months later, in February, when researchers published a major study of Louisiana’s voucher program. Students in the program were predominantly black and from low-income families, and they came from public schools that had received poor ratings from the state department of education, based on test scores. For private schools receiving more applicants than they could enroll, the law required that they admit students via lottery, which allowed the researchers to compare lottery winners with those who stayed in public school.
They found large negative results in both reading and math. Public elementary school students who started at the 50th percentile in math and then used a voucher to transfer to a private school dropped to the 26th percentile in a single year. Results were somewhat better in the second year, but were still well below the starting point.
This is very unusual. When people try to improve education, sometimes they succeed and sometimes they fail. The successes usually register as modest improvements, while the failures generally have no effect at all. It’s rare to see efforts to improve test scores having the opposite result. Martin West, a professor at the Harvard Graduate School of Education, calls the negative effects in Louisiana “as large as any I’ve seen in the literature” — not just compared with other voucher studies, but in the history of American education research.
There’s always the chance that a single study, no matter how well designed, is an outlier. Studies of older voucher programs in Milwaukee and elsewhere have generally produced mixed results, sometimes finding modest improvements in test scores, but only for some subjects and student groups. Until about a year ago, however, few if any studies had shown vouchers causing test scores to decline drastically.
In June, a third voucher study was released by the Thomas B. Fordham Institute, a conservative think tank and proponent of school choice. The study, which was financed by the pro-voucher Walton Family Foundation, focused on a large voucher program in Ohio. “Students who use vouchers to attend private schools have fared worse academically compared to their closely matched peers attending public schools,” the researchers found. Once again, results were worse in math.
Three consecutive reports, each studying one of the largest new state voucher programs, found that vouchers hurt student learning. Researchers and advocates began a spirited debate about what, exactly, was going on.
Mark Dynarski of the Brookings Institution noted that the performance gap between private and public school students had narrowed significantly over time. He argued that the standards, testing and accountability movement, for all its political shortcomings, was effective. The assumed superiority of private schools may no longer hold.
Some voucher supporters observed that many private schools in Louisiana chose not to accept voucher students, and those that did had recently experienced declining enrollment. Perhaps the participating schools were unusually bad and eager for revenue. But this is another way of saying that exposing young children to the vagaries of private-sector competition is inherently risky. The free market often does a terrible job of providing basic services to the poor — see, for instance, the lack of grocery stores and banks in many low-income neighborhoods. This may also hold for education.
Others have argued that standardized test scores are the wrong measure of school success. It’s true that voucher programs in Washington and some others elsewhere, which produced no improvements in test scores, increased the likelihood of students’ advancement and graduation from high school. One study of a privately financed voucher program in New York found positive results for college attendance among African-Americans.
But research has also linked higher test scores to a host of positive outcomes later in life. And voucher advocates often cite poor test scores in public schools to justify creating private school vouchers in the first place.
The new voucher studies stand in marked contrast to research findings that well-regulated charter schools in Massachusetts and elsewhere have a strong, positive impact on test scores. But while vouchers and charters are often grouped under the umbrella of “school choice,” the best charters tend to be nonprofit public schools, open to all and accountable to public authorities. The less “private” that school choice programs are, the better they seem to work.
The new evidence on vouchers does not seem to have deterred the Trump administration, which has proposed a new $20 billion voucher program. Secretary DeVos’s enthusiasm for vouchers, which have been the primary focus of her philanthropic spending and advocacy, appears to be undiminished.
