This blog originally appeared on Fordham’s Flypaper on June 4, 2014.
Mike Petrilli over at Fordham raises a question that I get all the time from policymakers: what explains the pretty extreme variation we see in charter school outcomes across states? The easy answer is that it’s policy, and by changing policy we can ensure quality.
But it’s not that simple. Policy guarantees nothing, and good state laws don’t necessarily result in good schools. Instead, charter quality depends mainly on implementation, school design development, and the talent pipeline.
Macke Raymond and her colleagues at CREDO have done some initial work with their massive data set to see whether state caps, multiple authorizers, and other factors bear any relation to outcomes. They didn’t find much, and what they did find was sometimes counterintuitive: for example, charter caps were associated with worse quality. In looking over the CREDO outcomes by state, there are also no obvious patterns related to state funding levels—for instance, Pennsylvania charters underperform other public schools at a cost of $12,000 per pupil, while California charters outperform their peers for more like $8,000 per pupil.
Does all this mean that policy doesn’t matter—that charter outcomes are random? Absolutely not. It just means that we haven’t yet amassed the right evidence to know precisely which policies matter most. It’s past time we did so.
Here are the hypotheses I’d want to test in a comprehensive study of how different components of charter laws and implementation relate to quality:
- That certain aspects of charter laws create a minimum floor for quality. For example, in states where only school districts can authorize, like Oregon and Arkansas, we see few charter schools and with very low quality.
- That too many authorizers are a problem. There are places, like Ohio, where this is clear, but the number of authorizers probably doesn’t matter as much as what the authorizer does and how state law sets up the conditions for authorizing. We need to know much more about what makes certain authorizers more effective than others. NACSA has a thoughtful set of principles and standards, but no one knows whether those recommendations result in high quality or just more bureaucracy.
- That the type of schools approved matters. As Michael Goldstein argued here, “no excuses” schools account for much of the positive charter outcomes. Why, then, do some cities have so few of them? At least part of the explanation is probably investments in new school development and support structures. States like Rhode Island and Tennessee are weird outliers: they are getting strong results but are low in the state law rankings. Is this because of strong recruiting and screening of “no excuses” schools by folks at the Rhode Island Mayoral Academies, the Tennessee Charter School Incubator, and the Tennessee Achievement School District?
- That charter policies and implementation strategies matter for talent recruitment. We know that most high-performing, no-excuses CMOs only locate where Teach for America has an active presence and where they can find strong leaders. But we also know that talent can be recruited to cities where there is a reform “buzz” and investments in new schools and supports—as well as a welcoming regulatory environment. That New Orleans is now a magnet for extraordinary talent could not have happened if the law required charters to hire from a union pool.
Policy matters. But implementation matters, too: whether government agencies have the savvy to choose strong providers and the guts to close low-performing schools, and what kind of talent, school designs, and investments are rolled out in a given state. As my colleagues have found with studies of SEA policy, you can get a lot done with suboptimal policy if you’ve got great people, and all the policy in the world won’t save you if you don’t.
We need much better evidence, pronto, about which policies and authorizer actions are the most powerful drivers of success. That’s doable. CRPE, in partnership with Public Impact, is designing a study right now to connect those dots. Interested funders, please give me a ring…
