Student Vouchers Aren't Working. Here's Why
Do school voucher programs get less effective as they’re scaled up?
While vouchers appear to be enjoying a higher profile with Betsy DeVos as the U.S. secretary of education, the research on outcomes from these programs has taken a dramatic turn, one at odds with the direction DeVos and other policymakers are pursuing.
For years, voucher advocates have pointed to a series of more than a dozen reports—usually funded or conducted by voucher proponents—that used randomized approaches, similar to those used in medical research, to isolate the effects of vouchers on treatment groups in citywide programs. While other researchers have questioned those reports over the last decade and a half, voucher advocates have claimed that these "gold standard" studies showed vouchers boosting achievement significantly for some students. Furthermore, they liked to point out, no students were harmed by school voucher programs.
But that has all changed. In April, the Institute of Education Sciences released a rigorous study showing that the congressionally mandated Opportunity Scholarship Program in the nation's capital caused significant negative effects on student learning. Students who used vouchers through the program to attend private schools in Washington experienced a 7-percentile-point decline in mathematics and an almost 5-percentile-point decline in reading compared with students who applied to, but were randomly rejected from, the program.
This report follows recent research on voucher programs in Louisiana, Ohio, and Indiana, all producing large, negative effects on learning for voucher students. In Louisiana, an average student using a voucher would end the first year of the program falling from the 50th to the 34th percentile in math. If the student was in 3rd through 5th grade, he or she would end the year even lower, at the 26th percentile. The impact of participation in Ohio's EdChoice program was "unambiguously negative across a variety of model specifications, for both reading and mathematics," according to a study from the Thomas B. Fordham Institute last year. Similar negative findings are reported for Indiana's statewide voucher program, the largest in the nation.
Vouchers remain a favored program, however, for Secretary DeVos and other school choice enthusiasts. How are they dealing with the results from these new studies? Rather than addressing this research, DeVos continues to defend the $250 million voucher-grant program in President Donald Trump's proposed fiscal 2018 budget. But it's been even more fascinating to watch professional voucher advocates, who have spent their careers arguing that their studies of voucher-program test scores trump all other approaches to measuring the impact of vouchers, now twist themselves into knots to belittle the importance of test scores.
Some have tried to attribute the negative results to regulations that discourage "better" private schools in certain states from accepting vouchers that would then require their students to take tests. This claim does not hold water when we are also seeing large, negative effects of vouchers in the other states as well. Another possible explanation is that most of the research of years past that supported the success of vouchers was funded and conducted by voucher advocates who sought a particular result. However, some of these new, negative findings were also produced by pro-voucher organizations and researchers, to their credit.
Perhaps a likelier explanation for these poor results has to do with the actual students and schools themselves, including how students were grouped in private and public schools. Prior to the recent batch of research that has cast doubt on vouchers, studies lauding vouchers tended to be based on local and more targeted programs involving relatively small, nonrepresentative sets of students and schools. Yet, overall, private schools are actually no more effective, and often less so, than public schools. Indeed, our own research indicates that any apparent advantages for students in private schools are actually more a reflection of the fact that private schools do a better job of attracting—not producing—high-scoring students.
For our book, The Public School Advantage, we examined two nationally representative data sets to determine whether private schools really offer superior educational programs and outcomes, or whether higher test scores in private schools are simply a reflection of the fact that they serve more advantaged students. Those analyses revealed that, after accounting for differences in demographics, public schools are more effective, particularly in teaching mathematics.
Research as far as back as the Coleman Report in 1966 indicates that private school students enjoy the beneficial "peer effect" of being around affluent classmates who have abundant educational resources at home and parents who have firsthand experience with school success. These students benefit from the experience of having teachers who are able to focus on solely academic content, rather than the nonacademic needs of some students.
This peer effect is a significant factor in student learning, but frankly, there are only a limited number of academically advantaged peers to go around. And so, as choice programs expand, the private-school peer effect is diluted. Hence, despite benefits of greater socioeconomic integration for students from low-income families, the benefits may not be scalable in expanding voucher programs that are based on self-selection.
It makes sense, then, that negative results are now appearing as researchers carefully examine larger-scale programs. Earlier studies looked only at students leaving small groups of (presumably failing) public schools for small groups of private schools that self-selected into the voucher program. Those studies were therefore not representative of the wider populations of public and private schools.
Yet the newer, larger-scale studies are starting to more closely approximate the nationally representative samples we previously analyzed when coming to our conclusion that public schools, in fact, have an edge over private schools in student learning.
There is a disturbing disconnect between the predictable, negative effects that vouchers are having on students, and the continued enthusiasm policymakers show for these programs despite the growing consensus that they are causing harm. Do we, as parents, taxpayers, and voters, want to fund programs that elevate choice, but lead to detrimental outcomes for children? Is choice a means or an end? Do we want choice for its own sake, or do we want it to improve achievement for all children?
Vol. 36, Issue 36, Page 28Published in Print: June 16, 2017, as Why School Vouchers Aren't Working