Charter schools inconsistent but improving, national study finds

An update of a landmark 2009 study showed charter schools on the whole have turned their performance around and are serving poor and minority students especially well.

Charter schools are improving, according to a new national study, though there is still a wide quality range between the best and the worst.

The study, conducted by the Center for Research on Education Outcomes (CREDO) at Stanford University, is an update of the center’s landmark 2009 study, which compared students’ performance at charter schools in 16 states with that of their traditional public school counterparts and found that, in many instances, charter schools fell short or made no difference in students’ learning.

This time, the news for charters was much better: Overall, in reading, charter students in the 26 states studied had the equivalent of eight extra days of learning each year beyond their peers in traditional public schools, compared with a seven-day loss in learning that the 2009 study found.

In mathematics, the 2013 study found no significant difference in learning, whereas the 2009 study found that charter school students had the equivalent of 22 fewer days of learning. The gains for poor and minority students and English language learners were even greater.

“The general reaction to the 2009 study was shock and disbelief,” says Margaret Raymond, the director of CREDO. “It was the first time we as researchers could provide enough of a wide-angle view for people to understand they’d gotten their doctors’ report back and the news wasn’t all that great.

“The real drive here was to see if in fact the attention to quality that appeared to have followed … had actually paid off,” she says. “We were intrigued to find out that in fact in both reading and math there has been an increase in performance.”

The CREDO study, the most comprehensive study done of charter school performance, uses what it calls “virtual twins” to gauge student learning – a methodology in which researchers compare students with students, rather than schools with schools, while controlling for as many demographics as possible.

The virtual “twin” students are composites of students that go to a traditional public school that the charter school student might otherwise have attended, and that in all other respects look like the charter student.

While the charter schools, on average, performed better in this study than in the 2009 one – and fared particularly well against traditional public schools among the most disadvantaged learners – there was still a wide variety in quality, and disparity from state to state. All of which, say some charter school advocates, underlines why it’s so important to create stronger charter-school laws, with clear expectations for academic performance and clear mechanisms in place for shutting down poorly performing schools.

In this study, 19 percent of charter schools posted gains in reading that were significantly weaker than their traditional public counterparts, and 31 percent were weaker in math. Greg Richmond, president and CEO of the National Association of Charter School Authorizers, notes that these are similar to his organization’s finding last fall that between 900 and 1300 charter schools were performing in the bottom 15 percent of their state’s accountability system.

“This confirms what we’ve seen in our experience, and our belief that we have to close the charter schools that aren’t performing,” says Mr. Richmond, who believes that closing about 1,000 of the worst charter schools would go a long way toward improving the sector, and the overall reputation of charters.

In the CREDO study, about 8 percent of charters in the original 16 states had been closed since the 2009 study – one reason for the progress there. And Richmond notes that some of the states in which charter schools come out the worst in this study – like Texas and Nevada – have already made big improvements to their laws that he hopes will help them improve accountability for charter schools. Texas, for instance, will now automatically close charter schools that repeatedly fail state standards.

“We’re not talking about so-so schools being closed, these are the bottom of the barrel for a couple years running,” says Richmond. “That’s very helpful to have the law to back up those decision-makers and say this is a requirement. And it’s helpful for the folks starting the schools to know what the performance expectations are.”

Shutting down poor performers was always a basic principle of the charter movement, notes Richmond, but back in the 1990s when many charters were beginning, few people anticipated how tricky that might be to do without clear guidelines.

“It’s very hard for public officials who have to vote on these things to look at that murky [academic performance] data in the face of crying children, and decide to close a school,” Richmond says.

Nina Rees, president and CEO of the National Alliance for Public Charter Schools, agrees that states need to do a better job at closing poorly performing charters, and also holding charter school authorizers accountable for the job they do, and removing the ones who aren’t doing well.

But she hopes, too, that this new report will change the minds of people who used the 2009 report to criticize the whole charter movement. Charter skeptics, she notes, seized onto the 2009 findings that just 17 percent of charters were outperforming traditional public schools in math, with 46 percent at the same level, “to say that two-thirds of charters are doing worse or no better” than other public schools. “Now,” says Ms. Rees, “you can say two-thirds of charter schools are doing better or the same” in math. “The fact that we’ve turned the tide, and turned that data point around is compelling.”

The best news for the charter movement out of the CREDO study is the gains that charter schools are making with some of the most disadvantaged populations of students.

The charter sector has a greater proportion of students in poverty and a greater percentage of minority students than it did in 2009, notes Dr. Raymond, and those students are coming into schools at a lower achievement level than they used to. “What’s interesting is that it’s exactly those students who are benefiting the most from enrolling in charter schools,” Raymond says.

Students in poverty, for instance, receive the equivalent of an extra 14 days in reading and an extra 21 days in math compared with their traditional public school counterparts. Black students across the board get an extra 14 days of learning, and black students in poverty get an extra 29 days of reading and 36 of math. English language learners get an extra 36 days of both reading and math. 

“It’s a pretty strong story to be told,” says Raymond.