Home » Press Releases » Statement on Charter School Achievement

Statement on Charter School Achievement

CER Press Release
Washington, D.C.
June 19, 2012

CER President Jeanne Allen made the following statement about a story on charter schools that aired on Minnesota Public Radio:

A story on Minnesota Public Radio gives us a taste of what is to come this week as the National Alliance of Public Charter Schools meets in Minneapolis.

If past is prologue, we will see many more stories citing “studies” that show “mixed results” on charter school performance.

While Minnesota Public Radio doesn’t offer a citation for the study purporting to analyze charter schools on a national basis, we can guess. It’s the CREDO study, which continues to be trotted out by the media despite its unsound methodology. It’s like a bad penny!

Here is a resource that debunks the methodology used in the CREDO study. The study is not remotely “national.” It examined fifteen states, whereas charters are in forty-one states plus the District of Columbia. But that’s only the first problem. The reason no one else has tried to do a real national study is that it cannot be done. There is no way to do apples to apples comparisons across state lines, so no one else has pretended to try.

The only way to truly measure charter school performance is at the state level. And in study after study, where apples are measured against apples and oranges against oranges, we see charter schools consistently outperforming traditional public schools.

And when they do not? That brings up another issue raised in the Minnesota Public Radio story, one that is often used as a data point against charter schools: closures. But this makes no sense. The closure of a charter school proves the accountability measures built into the system for charter schools are working. How often do public schools close for underperforming? Please see this study for hard data about school closures and accountability.

It’s fine to put charter schools under the microscope. But conclusions should only be drawn from reliable data.