A new national study on the effectiveness of networks that operate charter schools finds overall that their middle school students’ test scores in reading, mathematics, science, and social studies aren’t significantly better than those of students in regular public schools.
The average results varied widely: Students in some charter networks managed three years of growth in two years; in others, students tested a year behind grade level after a year or two in the program.
The findings from the research group Mathematica and the Center on Reinventing Public Education at the University of Washington Bothell underscore the point that being run by a charter-management organization, or CMO, isn’t a predictor of an individual school’s or student’s success, and that CMOs cannot be lumped together as being effective or ineffective. Previous studies have shown the same about individual charters.
“I don’t think there’s any doubt some of the CMOs have done a great job,” said Thomas Toch, a Washington writer and policy expert who writes extensively about CMOs. “But it doesn’t mean that every CMO is going to be successful. That’s the clear message here. It suggests just how hard creating good new schools is, and how hard it is to scale networks with even very good schools.”
The study made public last week is part of a long-running project by Mathematica of Princeton, N.J., and the Center on Reinventing Public Education. It involved 40 CMOs with 292 schools in 14 states; all the management groups were nonprofits that controlled at least four schools and had at least four schools open in fall 2007.
The researchers focused on charter-management organizations to explore whether that model could be effective for scaling up the successes of individual charter schools. Charters are publicly funded but free of many rules governing regular public schools.
CMOs exist in part to address the unevenness in quality from charter to charter, said Robin Lake, the associate director of the Center on Reinventing Public Education. “There was a real question about, ‘Are CMOs helping to improve the quality of charters overall?’ ” And the answer, she said, is that they haven’t had a significant positive effect as a group.
The study also found that some practices associated with charter schools run by management organizations were particularly effective. Comprehensive behavior policies—including zero-tolerance, specific behavior codes with rewards and sanctions, and “contracts” with students or their parents about behavior—were identified as having a positive effect on students’ math and reading scores.
Also, charter-management organizations that provide intensive coaching of teachers, including frequent reviews of lesson plans and observation, appeared to boost student achievement.
Researchers also explored how quickly those organizations grew, whom the schools served, the resources they used, and what influenced their growth.
The report does not name the networks involved in the study. They were guaranteed anonymity for participating.
One finding from the 3½ -year-long project is that the CMOs serve a disproportionately large number of black, Hispanic, and low-income students—even more so than the districts in which they operate—but fewer students with disabilities and English-language learners.
The study was funded by the Bill & Melinda Gates Foundation and the Walton Family Foundation, with project-management assistance from the nonprofit NewSchools Venture Fund, which invests in charter schools and other educational innovations. (Gates also provides support for organizational capacity-building to Editorial Projects in Education, the nonprofit publisher of Education Week.)
Nationwide, about 130 CMOs serve some 250,000 students. They account for the operation of about one in five of the 5,000 charter schools in the country, an increase from 12 percent in 1999.
Joshua Haimson of Mathematica, the project director of the CMO studies, said examining the work of CMOs “allows us to answer two important questions: To what extent have CMOs been effective at expanding educational models, and how have they done that?”
Looking Under the Hood
To gauge charter networks’ success at promoting student achievement, the researchers compared their students’ performance with that of students at nearby district-run schools, and in some cases, with independently run charters. The researchers looked at test-score gains for individual students from a year before they entered the CMO schools to up to three years later and compared them with data from students who resembled them in nearby districts.
Of the 40 CMOs in the study, data from 22 were complete enough to be used in this portion of the report. Two years after enrolling, students at 11 of the 22 did significantly better in math while a third did significantly worse. In 10, students experienced positive effects in reading, while at six, there were negative results.
Because some CMOs were able to advance students three grade levels in two years, their methods have the potential to close achievement gaps, Mr. Haimson said. Larger networks of charter schools generally did better at improving student achievement than smaller ones.
In addition to behavior policy and teacher coaching, the researchers examined features often found in network charter schools, including their use of additional instructional time, performance-based pay for teachers, and frequent formative assessments.
At first, additional time offered at some CMO-run schools appeared to influence student performance, but digging deeper, the researchers determined it was the teacher coaching and behavior policies that were the actual drivers, Mr. Haimson said.
Schoolwide behavior strategies by some CMO schools include setting behavior standards and signed responsibility agreements, but schools also said they had more flexibility than district principals in defining the details of all behavior policies.
Teacher coaching included more-frequent observation of teachers and more feedback to teachers from those observations, as well as frequent reviews of teachers’ lesson plans. CMO schools were more likely than nearby public school systems to base teachers’ pay on student test scores and observations than on seniority and education.
While behavior policy and teacher coaching emerged as definitive ways to improve achievement, the other practices shouldn’t be discounted, said Ms. Lake. Nor should those strategies be considered surefire.
“The takeaway shouldn’t be, if you just plug in a good behavior policy, you’re going to see good results,” Ms. Lake said.
The study also looked at how much CMOs spend per student compared with regular schools. Spending in the charters studied ranged from $5,000 to $20,000 per student a year, including public and private money.
The mixed results lead to a larger conclusion about school reform, said Mr. Toch, also a former reporter and editor for Education Week in the 1980s. “CMOs as a strategy are only one piece of the school improvement puzzle,” he said. “I don’t think we can expect to see thousands of truly game-changing schools from the CMO movement,” even though some such networks have shown success.