Philip Gleason, Melissa Clark, Christina Clark Tuttle, and Emily Dwoyer
Institute for Education Sciences
June 2010
Before you dismiss this report as another in the line of frequent—and often contradictory—charter effectiveness studies, take note: This is the first large-scale randomized trial examining charter effectiveness across multiple states and communities (hard to believe, perhaps, but true!). Using the same technique as other randomized (but smaller) charter studies—comparing students “lotteried in” or “lotteried out” of charters—researchers analyzed the academic outcomes for students who applied to thirty-six charter middle schools in fifteen states that held lotteries for the 2004-05 and 2005-06 school years. The students had to have taken the state assessment the year before applying (at a traditional school, usually) and their academic impacts for “lotteried in” and “lotteried out” students were averaged over the next two years. What do we learn? On average, charter middle schools perform roughly the same as their district counterparts in reading and math, and their performance varies widely. But there’s an important caveat: Charter middle schools serving low-income and/or low-performing (pre-admission) students had significant positive effects on math, while charters serving more affluent and/or higher-achieving (pre-admission) students had significant negative effects on math. In reading, there were no effects for low-income and/or low-performing students and negative effects for affluent and/or higher-performing students. Still, the positive impact on math performance adds weight to what we already know: While charter schools overall may not outperform traditional schools, they often provide better alternatives for urban and low-income children. Peruse the findings here.