Charter Schools, Special Education Students, And Test-Based Accountability

Opponents often argue that charter schools tend to serve a disproportionately low number of special education students. And, while there may be exceptions and certainly a great deal of variation, that argument is essentially accurate. Regardless of why this is the case (and there is plenty of contentious debate about that), some charter school supporters have acknowledged that it may be a problem insofar as charters are viewed as a large scale alternative to regular public schools.

For example, Robin Lake, writing for the Center for Reinventing Public Education, takes issue with her fellow charter supporters who assert that “we cannot expect every school to be all things to every child.” She argues instead that schools, regardless of their governance structures, should never “send the soft message that kids with significant differences are not welcome,” or treat them as if “they are somebody else’s problem.” Rather, Ms. Lake calls upon charter school operators to take up the banner of serving the most vulnerable and challenging students and “work for systemic special education solutions.”

These are, needless to say, noble thoughts, with which many charter opponents and supporters can agree. Still, there is a somewhat more technocratic but perhaps more actionable issue lurking beneath the surface here: Put simply, until test-based accountability systems in the U.S. are redesigned such that they stop penalizing schools for the students they serve, rather than their effectiveness in serving those students, there will be a rather strong disincentive for charters to focus aggressively on serving special education students. Moreover, whatever accountability disadvantage may be faced by regular public schools that serve higher proportions of special education students pales in comparison with that faced by all schools, charter and regular public, located in higher-poverty areas. In this sense, then, addressing this problem is something that charter supporters and opponents should be doing together.

It is well-established that test-based accountability in the U.S. is dominated by measures that tell you far more about the students a school serves (e.g., proficiency rates) than about schools’ contributions to those students’ progress (e.g., growth model estimates). Even in states and districts that do employ growth models, which are designed to isolate, however approximately, schools’ contributions to their students’ testing progress, status measures such as proficiency rates still dominate overall ratings. And, of course, NCLB is based almost entirely on absolute proficiency, and recent attempts to reauthorize this law appear to be perpetuating this flawed interpretation of testing data.

As a result of the heavy reliance on these (misinterpreted) measures, any schools serving large proportions of traditionally lower-scoring subgroups are at a severe, often fatal disadvantage. And that of course includes special education students. In fact, among the schools that were held accountable for their special education students' testing results under NCLB between 2003 and 2005, over half failed to meet proficiency targets for this subgroup (Davidson et al. 2013). 

Now, to be clear, to whatever degree the fact that charters serve disproportionately fewer special education students is due to deliberate recruitment practices, or so-called “counseling out,” it has to stop. But there is also evidence that this "service gap" may be driven largely by classification, resources/facilities, or simple parental preferences (e.g., Winters 2013).

Regardless of why it occurs, this gap does seem to exist in most places (and it may be even larger if one looks at students with the most severe disabilities), and some charter school supporters are calling on operators to address it. I have no idea if charter schools could be effective in serving this student population if they made more a more concerted effort to do so.

What I do know is the following: Given the (flawed) design of virtually all test-based accountability systems in the U.S., coupled with the enormous “do or die” pressure on urban schools to produce testing results, there is a rather serious disincentive for charters even to try. Testing outcomes are the coin of the realm when it comes to charter schools’ reputations. Any charter school that invests in personnel and facilties to serve special education students, or otherwise focuses actively on increasing their special education student population, risks stigma and failure. This incentive structure may prove a serious obstacle to addressing the special education "service gap."

But our reliance on these inept, unfair, poorly conceived accountability measures has consequences that go way beyond differences in special education populations between charter and regular public schools. Whatever accountability disadvantages may be levied on regular public schools as a result of their serving proportionally more special education students are completely swamped by the accountability disadvantages that all schools in low-income areas, both regular public and charter, face in comparison with schools in more affluent districts.

There is a reason why, according to NCLB and virtually every state-specific school rating system, schools serving more disadvantaged populations are virtually guaranteed to receive low marks: It is due to the dominance of indicators that represent the conflation of student and school performance.

Yet, ironically, charter opponents and supporters are far more likely to use these measures selectively to attack each other than they are to acknowledge that these simple rates aren't telling us much of anything about school performance, and that improving how we measure school performance is actually an area in which the interests of both camps converge.