Charter School Productivity Report: Red Flags or Red Herrings?
A new report, The Productivity of Public Charter Schools, released by my research team on July 22 has generated a lot of buzz from places like National Public Radio and Wall Street Journal Live. As I described in a previous post here, we find that charter school sectors in all 28 states that we study demonstrate higher productivity and/or return on investment than their traditional public school sectors. That finding has ruffled some feathers. One commentary that is attracting substantial attention is a statement by Jim Hull, an employee of the National School Boards Association, criticizing our report’s methodology. But does Mr. Hull raise red flags or red herrings? I think it is the latter and here is why.
Mr. Hull criticizes our report for not taking account of differences in expenditures across the charter and traditional public school (TPS) sectors. Our response is that our original study, upon which the productivity report builds, was a revenue study – a careful documentation of the different amounts of funding received by charters and TPS, based on audited “official statements of record”. Revenue studies, such as ours, should track revenues, not expenditures. His claim that our revenue study, and the treatment of revenues in our productivity report, are flawed because they don’t account for expenditures is like criticizing a horse for not being like a canary. We focused on revenues in our study because they represent the funding that schools receive and the reporting of education revenues follows consistent mandatory accounting practices whereas the reporting of school expenditures does not.
Hull says that a valid study of charter school productivity would focus on whether or not “charter schools can provide the same services with fewer funds than traditional public schools” (TPS). Actually, such an approach would produce an invalid productivity study. Most productivity gains come from one organization doing different things than the other organization. More productive organizations provide fewer services and programs that are not effective at accomplishing their core mission, which we propose is increasing student learning in the case of schools, and more services and programs that are effective in accomplishing their core mission. Productivity differences stem precisely from different organizations doing different things. Hull is arguing that TPS are as productive as charter schools once you factor out all the ways that TPS are unproductive, like allocating resources to high-cost extra-curricular activities and adding a lot of special programs that do not necessarily help kids learn. Yes, TPS have a lot more bells and whistles than charters. The problem, from a productivity standpoint, is that those extra bells and whistles don’t appear to provide TPS with much of an advantage regarding their core mission of student learning. Therefore, by definition, they make the organization less productive than other organizations that do not do those things.
Here is an example to prove my point. Our charter school productivity study is like a recently released national study of brain health that finds that adults living in Maryland tend to have much healthier brains than adults living in Mississippi. Just like that study, our study is a descriptive examination of differences across groups. According to Mr. Hull’s logic, we should object that people in Mississippi are being judged unfairly relative to people in Maryland, because people in Mississippi smoke more cigarettes, eat more fatty foods, and read less than people in Maryland. To compare them fairly on brain health, Hull would say, you have to control for all the ways that people in Mississippi are unhealthy. Actually, no, you don’t and you shouldn’t.
Now, Mr. Hull might object that adults in Mississippi choose to live unhealthier lives than people in Maryland whereas TPS are required to spend money on all kinds of unproductive things. Sure, that’s exactly the point. Those are policy and regulatory decisions that could be changed. But first you have to acknowledge that TPS spend more money than public charter schools on services and programs that do not significantly impact student learning. When schools, such as public charter schools, are liberated from the need to spend money unproductively, then they spend their revenues more productively. Perhaps we should grant more public schools that freedom.
Hull claims that we do not acknowledge that TPS enroll more special education students than charters. That is false. On pages 9-11 of our report we discuss the fact that TPS enroll 3 percent more special education students, on average, than public charter schools. We then explain that (1) it is implausible that this small gap in special education enrollments explains the lower levels of charter school funding we documented in our revenue study, even if many of the extra special education students in TPS are extraordinarily expensive; and, (2) we control for differences in student characteristics across charters and TPS throughout our productivity study. So, special education is not a big difference and we control for it in our analysis.
Hull claims that our accounting for TPS revenues is inflated by the fact that many TPS receive funds to transport public charter school students to school and therefore are assigned revenues that actually benefit charter, not TPS, students. States vary in their laws regarding the transportation of charter school students. Many states do not require charter schools to provide transportation to students and provide no money for them to do so. Transportation funds can’t be generating bias in those states. Many other states require charter schools to provide transportation to students and also classify charters as their own Local Educational Agency (LEA). In those states, transportation funds either pass through the local TPS to the charter school or they go directly to charters. In either case we count them as charter revenues. Again, no bias.
A very small number of states in our study both require that charters provide transportation AND do not treat charters as their own LEA. For those states, charter transportation funds are sent to the TPS and, if documented, we credit them to charters (so again no bias) but if not sufficiently documented, they might be improperly credited to TPS. In our revenue study, we urge states not to adopt such policies specifically because they interfere with fiscal transparency, but they are not common and involve far less than 1 percent of total revenue in our study. Thus, transportation funds could not come anywhere near explaining more than a tiny fraction of the large productivity advantage for charters.
Finally, Mr. Hull claims that many public charter schools co-locate in TPS buildings for free and that these facility “gifts” to charters explain much of the funding gap. This is a clear case of generalizing from outliers. Charter co-location with TPS is uncommon. When it does occur charters usually have to pay a facility fee to the TPS. We account for that in our revenue analysis so there is no way that charter co-locations, rare as they are, explain even a fraction of the productivity advantage of charter schools that we uncover in our study.
Whew, that was exhausting. I think I’ll take a vacation to Alaska’s Bering Sea, where the herring are a lot tastier and more satisfying than the many red ones that populate Mr. Hull’s critique.
-Patrick J. Wolf
Patrick Wolf is Distinguished Professor of Education Policy and 21st Century Endowed Chair in School Choice in the Department of Education Reform at the University of Arkansas in Fayetteville.