I recently posted a link to a Brookings Brief by Mark Dynarski, which warned that vouchers had not been successful in two states, Louisiana and Indiana. About the same time, the University of Arkansas released a research review that lauded vouchers. Although I did not know Dynarski, I contacted him and asked if he would explain the discrepancy for the readers of the blog. He graciously agreed.
In a recent article for Brookings, I highlighted recent research on vouchers to attend private schools that had found negative effects on student achievement. The same day, May 26, the Wall Street Journal published an editorial pointing to positive effects of school vouchers on student achievement, citing a review of studies published by researchers at the University of Arkansas. You asked if I could help readers understand the discrepancy.
In its reading of the University of Arkansas review, the Wall Street Journal included the review’s findings for voucher programs that operated in the US and programs that operated in Colombia and India. The largest positive effects of vouchers were from the program in Colombia. Education systems are quite different in other countries, however, and findings from Colombia and India have little relevance to debates about vouchers in the US today. If we ask about voucher programs that have operated in the US, the review reports that average effects of those programs is about zero.
The Louisiana and Indiana programs I focused on operated statewide. The negative effects reported for these programs could be a result of private schools being compared to higher-quality public schools in suburban and rural areas. Earlier voucher programs that reported positive results often operated in single cities—Milwaukee, New York City, Dayton, DC—which means studies of them essentially are comparing private schools only to urban public schools.
The Louisiana and Indiana programs also are recent, and my piece notes another possible explanation for their negative effects. Public schools have been under pressure for the last fifteen years to improve student achievement, which may have caused them to up their game. Recent research I cited concluded that public schools have substantially caught up with private schools. The National Assessment of Education Progress reports that private schools still have higher test scores than public schools, but those score differences could arise because of differences between private school students and public school students. The research approaches used in the Louisiana and Indiana studies allow for ‘apples to apples’ comparisons. Essentially the same students are compared in public and private schools and the test-score results favor public schools.
Vouchers will continue to be an important topic for discussion and debate, and we need to be open to new evidence and let our understanding of the world and of education be affected by it. I emphasized in my piece that our historical understanding that private schools perform better than public schools may be flawed. The University of Arkansas review is valuable for synthesizing a large amount of research on vouchers since the nineties into quantitative findings. The recent studies in Louisiana and Indiana are valuable for asking what the effects of vouchers might be today if a state were to begin a program or continue one. That the findings are negative means policymakers should proceed with caution—the relative positions of public and private schools may have changed.
I hope the discussion is useful for your readers, who rightly might feel a sense of whiplash from having different findings about vouchers released on the same day.