Of the 40 CMOs that were selected for inclusion in the study for various reasons, including having a minimum of four member schools, 22 networks had sufficient data for the student-achievement analysis, which looked at three years of middle-school performance. The study found that, in general, students at charter-network schools outperform similar students at traditional public schools, although sometimes not by very much. But that overall average masks an enormous variation among different CMOs. High-performing CMOs are so effective they are providing the equivalent of three years of schooling for students every two years. But CMOs at the low end are so bad they are effectively costing students a year of learning every two years. Bottom line: 10 of the 22 CMOs are outperforming their public-school peers in math and reading, in some cases substantially; eight are middling; and four are serious laggards.
If you follow the issue closely don’t miss the – cliche alert – treasure trove of descriptive information in this study about how CMOs are operating. A few pro and con commenters have opined to the effect that this data must be either “forged by public schools and teacher unions” or not valid because it doesn’t use a pure RCT or randomized model as, for instance, Caroline Hoxby does in her research on charter school effects. The methods are solid, learn about them yourself at the link above. The focus on middle schools stems from a data availability issue. It’s harder to do longitudinal studies for elementary school students because most states don’t assess in the early grades and at the high school level the assessment policies are very mixed, creating data issues there and necessitating the use of other measures – eg graduation, college-going, etc…in addition to this ongoing research effort the new Broad Prize for CMOs will also shed some light on those issues.
How you interpret the performance data probably has a lot to do with that you think about charters and CMOs in the first place. My take is two-fold. One, given where a lot of CMOs operate I’m not surprised by the quality issues. A problematic mix of poor quality charter authorizing and badly designed state policies create an environment where school replication is not always a function of quality. But, while we can certainly do better there, even with improved laws and better authorizing no one should expect 100 percent success. There is inherent risk in creating new entities. Worth noting that even the very good CMOs have some individual schools that struggle. In my view the question is how much risk are we willing to tolerate and given that we’re talking about schools, how much should we tolerate?