Earlier this month, a research institute at Stanford University affiliated with the Hoover Institution reported that students at independent charter schools in Los Angeles performed a lot better than their peers in traditional Los Angeles Unified District schools. The Center for Research on Education Outcomes (CREDO) has now released a report for California as a whole, and the results are mixed.

Data for six years of students in grades 2 to 11 in nearly 1,000 charter schools showed that overall they performed better in reading but did worse in math. In reading, 32 percent of charter students outperformed their peers at traditional schools while 21 percent underperformed. But in math, 37 percent underperformed while 29 percent outperformed traditional schools with similar students.

CREDO matched students with similar demographics in charters and nearby district schools, and, as it did with its other charter school studies, translated the differences in achievement into roughly equivalent additional days of learning (with the caveat that they should be “interpreted cautiously”). The result: charter school students gained 14 learning days~~ in reading but lost the same number of days in math, based on a 180-day school year.~~

However, there were marked differences among charter schools. Students in urban locations, poor students and African-Americans who attended charters gained learning days compared to their peers at traditional schools. Students in rural areas, and Asian and white students enrolled in charters lost learning days (72 days lost in math learning for white students).

Results for Hispanic students were mixed, with seven learning days gained in reading and 14 days lost in math. However, low-income Hispanics in charter schools gained 22 days in reading and 29 days in math. A smaller proportion of English learners attend charter schools than traditional public schools (17 percent versus 24 percent statewide), but those who do attend charters perform better academically than their peers, with 36 learning days gained in reading and 50 days in math.

Elementary and middle school charters outperformed traditional schools, but charter high schools overall performed worse. At multi-level charter schools, serving elementary and middle grades or middle and high school students, students lost more than 100 learning days in math. “Unfortunately, more than half of the charter students in California attend high schools or multi-level schools, so their lack of growth has a large impact on the overall math results,” the study said.

As in Los Angeles, charter schools that are part of a charter management organization, ~~such as KIPP, Rocketship Education and Aspire Public Schools, excelled compared with unaffiliated charters usually started by parent groups, teachers or local non-profits. For charter management-connected schools, the gain was 36 days in reading and 28 days in math; for unaffiliated charters, the reading gain was 7 days; the loss in math jumped to 29 days.~~

CREDO published its first report on charters in California in 2009. In that study, charter school students on average had a seven-day learning gain in reading and a 22-day loss in math. The new report, using achievement scores in the same schools operating then and now, shows a doubling of the gain, to 14 days, in reading and a narrowing of the gap in math, to seven days.